I reviewed all games of ASL and saw Protoss wasn't the worst race.
Protoss arrived round of 8 in all season, it's similar Terran and Zerg. Protoss was absent 2 times in round of 4 (ASL9,10). Terran was absent 3 times in round of 4 (ASL5,10,11). Zerg was absent 1 time in round of 4 (ASL1). Difference about the number of race in semifinal and final is not much.
Through 13 seasons of ASL, there are so much players, maps have played/used but Protoss, Terran and Zerg still fine. We have different maps but we never have winrate 100% in all matchup, except mirror matchup. It means any race can win/lose in all maps. ASL is Player vs Players, it's not Race vs Race. Skill of player will define who will win.
Flash said that Everything is skill. It means the all of rest is not important (race, map, luck, matchup...). Winner is always who played better. It doesn't depend on race, map, luck, matchup...
I don't think Protoss is the worst race but Protoss have so much bad players, that is why Protoss usually lose in the final match.
If Terran is the best race, why Terran just won 6 times in 13 seasons. It would be 2 if we minus Flash. Maybe I can know what race is the worst after this topic.
There is no worst race, there is no best race. Protoss is strong, Zerg is weak, Terran is slow. Map and luck definitely matter, but in general greater skill wins, yes.
So what you did here, I think, is something like: (1) presented data that is consistent with protoss being the worst race but also consistent with other explanations (2) made the highly contentious claim that nothing but skill matters in determining who wins (3) arrived, therefore, at the conclusion that the data is explained by protoss being the least skilled
But there are at least two issues here: (a) the claim that only skill matters is doing all the work for you. If your reader comes in with the belief that protoss is disadvantaged, so race matters, you need a really convincing argument, perhaps with some data, to make them change their belief. But you don't do this. (b) your explanation has the problem that it presents a further problem: why are the elite protoss players less skilled than the elite terran and the elite zergs? What explains this?
You cannot make definite conclusions regarding which race is the best and the worst, even with that much data. The performance of each race ist always dependent on maps and on the player pool. But if you assume that over the years the maps and the player pools balanced themselves out, you can conclude that protoss players can perform on a similar level as the other races.
In the ASL specifically, Protoss won when the maps were Protoss favored. ASL does a good job saying map balance can be more important then the player pool. See the Rain vs Snow season.
Imba race : Protoss Best race when played to perfection:Terran shit race that doesnt matter how good you play u still loss to some shitty protoss sending zealots to your base: Zerg
At the highest level when players have incredible macro and balance that with great micro, what it comes down to is decision making and scouting. All the pros know what decisions need to be made if they have the intel since they have played thousands of games, but scouting for protoss at a higher level is harder. Terran has scans and mines, zerg has zerglings and overlords, protoss has zealots and obs. Now observers at a lower level are great since they are cloaked, but at a high level where players are actively trying to deny vision and have detection sitting an observer outside your opponents base is no longer viable. Also obs cost gas. Sure there are zealots, but that is 100 minerals to check for an expansion compared to 25 minerals or straight up free that the other races pay to scout. So although the game is very balanced I believe it is vision that makes the difference at a professional level. The next time you watch a PvZ or PvT check out the mini map. Terran v Protoss will have mines everywhere and scans, while protoss vision is limited to their army. In Zerg vs Protoss, the zerg will have lings sitting on expansions outside of the base, and sometimes if the protoss is not dominating the air the zerg will have overlords splitting the map gaining vision of anything past the halfway mark
On July 12 2022 01:41 Kare wrote: Hardest race to play: 1.Terran 2. Zerg 3. Protoss
Best race if played close to perfection: 1. Zerg 2. Terran 3. Protoss
My opinion
It's always been like this.
Even going to 17 years ago when pros were just figuring stuff out. Generally speaking I think Terran has outperformed expectations. (and not just because of Flash).
Those 'standard' maps (4p) that everyone applauds are specifically very kind and safe for Terran. I think this is the main reason for their dominance.
The minute mapmakers stray away from the classic formula, Protoss starts winning or Zerg suffers a lot against Terran.
Historically, results wise, it was always like this: T > Z >> P. It is a much stronger argument that P is the worst performing race than that T is the best performing race. KLM results also point very strong in that direction.
On July 12 2022 04:18 Mutaller wrote: At the highest level when players have incredible macro and balance that with great micro, what it comes down to is decision making and scouting. All the pros know what decisions need to be made if they have the intel since they have played thousands of games, but scouting for protoss at a higher level is harder. Terran has scans and mines, zerg has zerglings and overlords, protoss has zealots and obs. Now observers at a lower level are great since they are cloaked, but at a high level where players are actively trying to deny vision and have detection sitting an observer outside your opponents base is no longer viable. Also obs cost gas. Sure there are zealots, but that is 100 minerals to check for an expansion compared to 25 minerals or straight up free that the other races pay to scout. So although the game is very balanced I believe it is vision that makes the difference at a professional level. The next time you watch a PvZ or PvT check out the mini map. Terran v Protoss will have mines everywhere and scans, while protoss vision is limited to their army. In Zerg vs Protoss, the zerg will have lings sitting on expansions outside of the base, and sometimes if the protoss is not dominating the air the zerg will have overlords splitting the map gaining vision of anything past the halfway mark
Protoss should always have map control against terran and zerg other than a few all-in openings or 973 hydras. Having map control and roaming the map while denying scouting is FREE scouting. Pro protoss players with map control never sit their army at their rally point when they have map control, they are always pressuring and engaging/retreating. I always felt it was T>P>Z rather than T>Z>P overall.
On July 12 2022 02:36 [sc1f]eonzerg wrote: Imba race : Protoss Best race when played to perfection:Terran shit race that doesnt matter how good you play u still loss to some shitty protoss sending zealots to your base: Zerg
shit race that doesnt matter how good you play u still loss to some shitty protoss sending zealots to your base: Zerg & Terran
On July 12 2022 01:41 Kare wrote: Hardest race to play: 1.Terran 2. Zerg 3. Protoss
Best race if played close to perfection: 1. Zerg 2. Terran 3. Protoss
My opinion
This is correct
Protoss is the worst race when played at top tier pro level, and I’m so happy about it because I hate all the cheating/cheesing/BMing apes I see on the ladder
On July 12 2022 02:36 [sc1f]eonzerg wrote: Imba race : Protoss Best race when played to perfection:Terran shit race that doesnt matter how good you play u still loss to some shitty protoss sending zealots to your base: Zerg
shit race that doesnt matter how good you play u still loss to some shitty protoss sending zealots to your base: Zerg & Terran
Spoken like someone who's never lost a PvP to zealots in the base.
This blog is such a gem. I am so in agreement with the following statement :
In YGOSU, it is mostly agreed upon that when three or more players are playing on team melee, protoss becomes the strongest race due to the likelihood of strong play-making potential from units such as early game harassment from probes having a single player dedicate all his focus on it, corsairs that never die throughout the entire game versus the zergs, and game changing late-game spellcasters such as arbiters.
I was thinking the same about Protoss before, and then I saw a few team melee games between Bisu+Best vs Light+Rush, and man oh man did the Protoss team totally destroy Terran.
The thing is, the Terran army during a fight always gets some sort of value even if you set your units up badly. If Protoss units are in range, they fire. Same thing can't be said for Protoss: + Dragoons hitting Depots or a floating Barrack. + Zealots getting on top of each other, and worse, eating a mine together. + High Templars evaporating before casting any storms + Shuttles dying mid-air because frankly, by the time you have the time to grab them they're already dead. All of the above examples can be mitigated by a great player but only to some extent. Chances are, if you are able to finish sorting out the targeting of your ground army, your Shuttles are likely on auto pilot and you won't be able to cast all the storms you'd like. But if you have two or three players controlling the same battle. The extra values Protoss can get is huge.
Same thing can be said for PvZ. No more scouting Probe dying early. No more High Templar full of energy dying before storm can get off. No more Corsair wasting.
Obviously Terran and Zerg can benefit a lot from team melee as well. But I feel like the extra values are nowhere near Protoss'.
It is so absurd that people in 2022 STILL dont understand the most important concept of ANY Balance Discussion:
Maps
Why even bother with statements without looking at maps.
That is precisely the reason why any "general" statement regarding balance is useless. You simply can not say T>Z>P. Or Terrans are more succesfull than Zergs and Protoss ect.. ect.. ect..
The only thing Id be willing to agree on OVERALL without map balance stuff is that
Protoss is the easiest to learn and has the lowest skill ceiling whereas Terran and Zerg are the hardest to learn and masters with the highest skill ceiling.
But that statement tells us nothing else than what it says.
One more thing that also clearly make any comparisons in terms of "RACE BALANCE" extremely tough is this:
We simply do not know how good (lets say the top10) of each Race in particular is. It could be that for 8 years the best 5 terrans simply are better than the best 5 protosses. So you can never just assume that Terran is "broken" or imbalanced. It simply could be that those particular TOP5 Terran players are just BETTER than the Top5 Protoss players.
Actually it probably would be easier to look a massive overall population stats and come to conclusions. But that is not helpful as well because the skill range will be between D to S+ and this also skews things.
TL;DR:
Talking about BALANCE of the RACES without including MAPS and a certain timeframe/map pool (ect..) is a waste of time. It does NOT MAKE ANY SENSE <3
@Peeano Sure, it belongs to style of player who likes how play/enjoy their games
@Puosu Protoss has units to beat Terran and Zerg. PvT, Protoss has Arbiter/High Templar/Carrier, of course Zealot and Dragoon. PvZ, Protoss' units even are very good.
I think StarCraft is perfect balance game. Terran is passive in TvP, active in TvZ. Protoss is passive in PvZ, active in PvT. Zerg is passive in ZvT, active in ZvP.
Personally, I am Terran (C rank), my TvZ is 1-2% winrate in 100 games latest. I did everything but I can't win D-B Zerg, Zerg is just sure win from all build order, maps, apm, knowledge, take advantage or get disadvantage...
I remember a game, Zerg took 4 pool, only killed 1 my scv. After 15 minutes, Zerg has 5 bases and I still sit in my base (1 base) but I has 1500-1600 mmr, I don't think I am noob.
I just think I lose TvZ so much because my skill is so bad in TvZ. It's so hard to micro to defend from 9 pool, mass lings early game, lurker rush...
Anyway, 20 years history of StarCraft, there is not much elite players (T, P, Z), I think, maybe we will have some good Protoss players who have enough skill to beat Terran, Zerg in the finals. Another reason, it's usually a player good at specific matchup and bad at the others matchup. Example, Mini, Bisu (P) is good at PvZ, bad at PvT. Larva (Z) is good at ZvP, bad at ZvT... When you reviewed all games in a league, you can know why champion won in the final (except Flash).
On July 11 2022 21:49 BisuDagger wrote: In the ASL specifically, Protoss won when the maps were Protoss favored. ASL does a good job saying map balance can be more important then the player pool. See the Rain vs Snow season.
I agree with this. Maps are more of a factor than the race itself. The maps truly are the balancing factor.
i used to hate zerg during the era when queen and soma looked unstoppable in pvz. before queen lost to mini twice in ASL. but i never insulted their skill or called their race broken, etc. some of you are absolutely seething.
Protoss at the top level is 100% the worst race and it's down to the maps and the fact it's almost impossible to make a map that's really good for PvZ without completely breaking all the other matchups. (see Central Plains and how that became a PvP map)
On July 13 2022 02:21 whylessness wrote: i used to hate zerg during the era when queen and soma looked unstoppable in pvz. before queen lost to mini twice in ASL. but i never insulted their skill or called their race broken, etc. some of you are absolutely seething.
If you play BW and you don't play toss there is a very high chance you are a salty, excuse searching ragekid. Protoss players in general are just the better human beings. Always have been, always will be.
I think all balance discussions under the BW forum should be deleted and banned. It's complete nonsense and there is always someone trying to bring something to the table, while having no idea how to play properly.
It's all about YOU. The races are just the tool that YOU use.
Other than that maps are important. Balance whiners should be muted in the forums. It's over 20 years after the last balance patch. Get over it and start learning to play ffs.
On July 13 2022 15:43 BlueStar wrote: I think all balance discussions under the BW forum should be deleted and banned. It's complete nonsense and there is always someone trying to bring something to the table, while having no idea how to play properly.
It's all about YOU. The races are just the tool that YOU use.
Other than that maps are important. Balance whiners should be muted in the forums. It's over 20 years after the last balance patch. Get over it and start learning to play ffs.
Agree 100% ! I never understood the discussion about this topic. If you think zerg is better then pick zerg as your main race. Or terran or protoss. You are free to choose whatever you like...
I don't think so, Rain (P) vs Snow (P) in ASL 5 because they played better than the others players. ASL 5 has 4 maps: Sparkle Transistor Gladiator Third World.
Theory: Sparkle and Third World favors P, Transistor favors Z, Gladiator favors T but Third World just was used from round of 16. That means, players have 3 maps favor 3 race at round 24. This season, T players were so bad from round of 24 (10P, 7T, 7Z) to round of 16 (7P, 2T, 7Z).
At round of 16 (7P, 2T, 7Z) with 2 maps favors P went to round of 8 (4P, 2T, 2Z). P was disqualified 3 players, Z was disqualified 5 players, T was disqualified 0 player.
What did you see? When there are 3 balance maps, T was disqualified 5 players, P was disqualified 4 players. When there are 2/4 maps favor P, T was disqualified 0 player, P was disqualified 3 players. What a balance map you said?
ASL 2 has 4 maps: Benzene, Circuit Breaker, Eye of the Storm, Demian
Theory: Circuit Breaker, Eye of the Storm, Demian favor T, and Benzene favors T too statistically. What a season with all maps favor Terran. Sure the final: Flash (T) vs Sea (T)
But do you know round of 8: 4P, 2T, 2Z? Why 4P (Bisu, BeSt, GuemChi, Stork), 2Z (hero, Jaedong) can arrived a season that has all maps favor Terran?
But do you know 2 match semi-final: Best, Jaedong almost won with similar 2-3 (Best vs Sea; Jaedong vs Flash). In turning point of the match, both they decided wrong, if they had played again, they would have crossed their opponents and the final would be Best (P) vs Jaedong (Z) in a season with all maps favor Terran. Anyway, Best 2-3 Sea, Jaedong 2-3 Flash. It's not 0-3.
I think you can check yourself ASL 10 with Soma (Z) vs ZerO (Z) in the final. It must be ASL 10 has so much maps favor Zerg, huh.
Maps just look like nicknames on ladder. You can change map/nickname but you can't change your skill, and you can't change result a BO5 by maps.
On July 13 2022 15:43 BlueStar wrote: I think all balance discussions under the BW forum should be deleted and banned. It's complete nonsense and there is always someone trying to bring something to the table, while having no idea how to play properly.
It's all about YOU. The races are just the tool that YOU use.
Other than that maps are important. Balance whiners should be muted in the forums. It's over 20 years after the last balance patch. Get over it and start learning to play ffs.
No, the kind of post like yours should be banned instead. Any discussion, as long as it doesn't descend into chaos and insult, is worthwhile. That is what forums are for.
Every time a thread like this appears, there are people like you who come and be like "this is nonsense, stop whining, get good". Well, I haven't seen any whining in this thread. If anything, people are discussing balance at the highest level, not complaining about their own experiences of the game. So why are you telling them to learn to play?
It's like we were saying the Ballon d'Or favors strikers rather than defenders and goalkeepers, and you came in and tell everyone to learn to play as strikers??? The only nonsense comes from you.
In YGOSU, it is mostly agreed upon that when three or more players are playing on team melee, protoss becomes the strongest race due to the likelihood of strong play-making potential from units such as early game harassment from probes having a single player dedicate all his focus on it, corsairs that never die throughout the entire game versus the zergs, and game changing late-game spellcasters such as arbiters.
I was thinking the same about Protoss before, and then I saw a few team melee games between Bisu+Best vs Light+Rush, and man oh man did the Protoss team totally destroy Terran.
The thing is, the Terran army during a fight always gets some sort of value even if you set your units up badly. If Protoss units are in range, they fire. Same thing can't be said for Protoss: + Dragoons hitting Depots or a floating Barrack. + Zealots getting on top of each other, and worse, eating a mine together. + High Templars evaporating before casting any storms + Shuttles dying mid-air because frankly, by the time you have the time to grab them they're already dead. All of the above examples can be mitigated by a great player but only to some extent. Chances are, if you are able to finish sorting out the targeting of your ground army, your Shuttles are likely on auto pilot and you won't be able to cast all the storms you'd like. But if you have two or three players controlling the same battle. The extra values Protoss can get is huge.
Same thing can be said for PvZ. No more scouting Probe dying early. No more High Templar full of energy dying before storm can get off. No more Corsair wasting.
Obviously Terran and Zerg can benefit a lot from team melee as well. But I feel like the extra values are nowhere near Protoss'.
I think zerg and terran both benefit a lot more from team melee in tvz/zvt than they do in tvp / zvp. In particular in tvp, terran's opportunity for harassment are really limited by how well protoss defends, and terran isn't able to really do anything if protoss doesn't leave any opening. Pvt is kiinda the same, except that shuttle+reaver combo is much more capable of exploiting the tiniest holes.
Tvz/zvt however you'll see resource banks, unused defiler and vessel energy, miscontrolled scourge, big piles of units standing around, all that kind of stuff where team melee would visibly impact the performance. Whereas watching the best players play TvP there's just not all that much room for improvement in the first place.
PvZ is really the only matchup that was ever considered disadvantageous for protoss. But there are examples contradicting this view. Bisu achieved a PvZ winrate of 71.51% which is greater than that of Flash's 69.7% in TvP (the historically worst matchup for terran), and only slightly behind Flash's TvZ winrate of 72%, with both players having played a similar number of games. Bisu also had a 9-5 record against Jaedong (which is of course a small sample), while Flash achieved 20-20. This even though Jaedong had a better record against protoss with 67.38% compared to 63% vs terran. Furthermore, Bisu's best matchup has always clearly been PvZ, which puts a big question mark on the claim of racial imbalance.
These three players are typically considered the best of their respective race, and it's quite clear that their winrates contradict the idea that protoss players have it the worst, even in the allegedly worst matchup of PvZ.
I think that protoss players have never figured out what sets Bisu apart so much. He clearly has a far superior understanding of the matchup than anyone else. Why can't other protoss players study his game and apply their findings? I think that's the real question that needs to be asked. It's like the answer is right there in front of people, but they're not picking it up. We have hundreds of vods of Bisu playing PvZ that can be studied.
Instead people resort to complaining about balance, which is the lazy option.
On July 14 2022 00:52 Magic Powers wrote: PvZ is really the only matchup that was ever considered disadvantageous for protoss. But there are examples contradicting this view. Bisu achieved a PvZ winrate of 71.51% which is greater than that of Flash's 69.7% in TvP (the historically worst matchup for terran), and only slightly behind Flash's TvZ winrate of 72%, with both players having played a similar number of games. Bisu also had a 9-5 record against Jaedong (which is of course a small sample), while Flash achieved 20-20. This even though Jaedong had a better record against protoss with 67.38% compared to 63% vs terran.
Aren't these records from way back to the Kespa era, and only counted from official matches? Bisu and Jaedong certainly played more than 14 matches in their career.
Anyway, map and meta have changed dramatically since then. Players rise and fall as well. Bisu is not that dominant against Zerg anymore. And Jaedong is not even top 3 Zerg. I don't think these records mean much nowadays. We need to look at modern data.
On July 14 2022 00:52 Magic Powers wrote: PvZ is really the only matchup that was ever considered disadvantageous for protoss. But there are examples contradicting this view. Bisu achieved a PvZ winrate of 71.51% which is greater than that of Flash's 69.7% in TvP (the historically worst matchup for terran), and only slightly behind Flash's TvZ winrate of 72%, with both players having played a similar number of games. Bisu also had a 9-5 record against Jaedong (which is of course a small sample), while Flash achieved 20-20. This even though Jaedong had a better record against protoss with 67.38% compared to 63% vs terran.
Aren't these records from way back to the Kespa era, and only counted from official matches? Bisu and Jaedong certainly played more than 14 matches in their career.
Anyway, map and meta have changed dramatically since then. Players rise and fall as well. Bisu is not that dominant against Zerg anymore. And Jaedong is not even top 3 Zerg. I don't think these records mean much nowadays. We need to look at modern data.
These records don't mean much today? Well okay, I can't really argue with that. I can only say I completely disagree.
On July 14 2022 00:52 Magic Powers wrote: PvZ is really the only matchup that was ever considered disadvantageous for protoss. But there are examples contradicting this view. Bisu achieved a PvZ winrate of 71.51% which is greater than that of Flash's 69.7% in TvP (the historically worst matchup for terran), and only slightly behind Flash's TvZ winrate of 72%, with both players having played a similar number of games. Bisu also had a 9-5 record against Jaedong (which is of course a small sample), while Flash achieved 20-20. This even though Jaedong had a better record against protoss with 67.38% compared to 63% vs terran.
Aren't these records from way back to the Kespa era, and only counted from official matches? Bisu and Jaedong certainly played more than 14 matches in their career.
Anyway, map and meta have changed dramatically since then. Players rise and fall as well. Bisu is not that dominant against Zerg anymore. And Jaedong is not even top 3 Zerg. I don't think these records mean much nowadays. We need to look at modern data.
These records don't mean much today? Well okay, I can't really argue with that. I can only say I completely disagree.
Well, first of all, a quick check on liquipedia says Bisu vs Jaedong record back in the Kespa era is 10-10, not 9-5, so it kinda invalidates your original point though.
Secondly, looking at overall record of a top player also tells you nothing about balance. Obviously if player A is the top 1/2/3 of his race, he'll beat the majority of players from another race, making his overall stats > 50/60/70% - so that percentage only tells you how many times said player bullies people below his level.
I failed to filter for any league, but after I do that it comes out 12-8 in favor of Bisu vs Jaedong. Secondly I have no clue how the record of a top player wouldn't say something about balance. Flash, Bisu and Jaedong are named as the top three players. If they have even or uneven records against one another, that's one of the best data points to use for balance talks, because they're the greatest masters of their respective race. They make the fewest mistakes, they select the best of the known strategies, and they execute them most optimally in their current environment. Arguing that their records against one another shouldn't carry greater weight than other player matchups seems completely absurd. On what grounds?
Dont know how you get 12-8 but here's what liquipedia says: Bisu vs Jaedong
For the other point. I was talking about how you used the example of Bisu 71.5% win rate in PvZ to disprove the notion that P have it worst. I wasn't talking about their (Bisu/Flash/Jaedong) records directly against each other.
But then again, if we ignore the super small sample size and go by their records against each other, then it proves exactly that T>Z>P (Kespa era only): - Flash 14-10 Bisu - Flash 26-25 Jaedong - Jaedong 10-10 Bisu Same if you go for number of medals. But I must add that this is no good way to determine balance.
I know the importance of scouting, vision in this game and LOL game. But almost games these days are marco, decide to attack/defend and skilling will definte who is winner.
I think as soon as you are B rank at least, you can guess your opponents easily. In the elite matchup, you see both players playing looks like hack map, they don't have to scout to play well.
I remember Shuttle (P) vs Beast (Z), Shuttle had built Canons so much without saw Beast's Zerglings. This is called intuition of gosu, they play many games enough to create their intuition.
On July 14 2022 08:42 TMNT wrote: Dont know how you get 12-8 but here's what liquipedia says: Bisu vs Jaedong
For the other point. I was talking about how you used the example of Bisu 71.5% win rate in PvZ to disprove the notion that P have it worst. I wasn't talking about their (Bisu/Flash/Jaedong) records directly against each other.
But then again, if we ignore the super small sample size and go by their records against each other, then it proves exactly that T>Z>P (Kespa era only): - Flash 14-10 Bisu - Flash 26-25 Jaedong - Jaedong 10-10 Bisu Same if you go for number of medals. But I must add that this is no good way to determine balance.
Furthermore, Bisu has a much better record in PvZ than any zerg player has in ZvP. Neither JD nor Effort were able to break 70%. I will stress this point. The highest level protoss player has proven that PvZ is not zerg favored, and that cannot be attributed to Bisu's general skill, because in the other two matchups he performed significantly worse than in PvZ. JD's best matchup by far was ZvZ, even outmatching Flash's incredible TvT record. He also performed significantly worse in the other two matchups. This type of information proves that these were in fact the best players, and that their general skill doesn't explain why they dominated certain matchups more than others. Therefore Bisu's PvZ record does stand as a valid counter-argument against the claim that PvZ is zerg favored.
On July 14 2022 19:08 HOLYBATS wrote: Bisu and Mini is outlier regarding the pvz. Other high level pros can not show the same success.
How is that a valid argument? Bisu had roughly equal competition during his peak. Despite that, he was only able to push his PvZ winrate to the limit, but not PvT and PvP. Meanwhile zerg players were unable to push their ZvP winrate to the limit despite only Bisu being able to put up maximum resistance, and they all fell way short of Bisu's winrate. The balance argument falls apart when you realize how many contradictions come with it.
I also think so but it's just right at semi-final and final. Anyway, I still believe on players' skill than their race. Terran is not the best but Boxer, Nada, Flash, iloveoov is the best. That's why the finals usually has TvP, TvZ more than TvT, just only Terran player was good enough (skill) came to the final. It's not 2 Terran players.
Maps, race and anything have 99-100% balance but players' level (skill) is unique and this parameter can be down and up from every single games.
@TentativePanda
All race even have cheating/cheesing, all of us get often upset about it. I just think those stuff is build order/strategy/style (so strong but not sure win) and they appear everywhere, everytime (play for fun, ladder, tournament, round of 24, the final).
Terran has 8 Barrack (TvZ), fake-double factory (TvP). Zerg has 4 pool, 9 pool speed lings, 2 hatchery to muta (ZvT), 973 (ZvP)...
On July 14 2022 08:42 TMNT wrote: Dont know how you get 12-8 but here's what liquipedia says: Bisu vs Jaedong
For the other point. I was talking about how you used the example of Bisu 71.5% win rate in PvZ to disprove the notion that P have it worst. I wasn't talking about their (Bisu/Flash/Jaedong) records directly against each other.
But then again, if we ignore the super small sample size and go by their records against each other, then it proves exactly that T>Z>P (Kespa era only): - Flash 14-10 Bisu - Flash 26-25 Jaedong - Jaedong 10-10 Bisu Same if you go for number of medals. But I must add that this is no good way to determine balance.
Furthermore, Bisu has a much better record in PvZ than any zerg player has in ZvP. Neither JD nor Effort were able to break 70%. I will stress this point. The highest level protoss player has proven that PvZ is not zerg favored, and that cannot be attributed to Bisu's general skill, because in the other two matchups he performed significantly worse than in PvZ. JD's best matchup by far was ZvZ, even outmatching Flash's incredible TvT record. He also performed significantly worse in the other two matchups. This type of information proves that these were in fact the best players, and that their general skill doesn't explain why they dominated certain matchups more than others. Therefore Bisu's PvZ record does stand as a valid counter-argument against the claim that PvZ is zerg favored.
Bisu performing better in PvZ than PvP or PvT and having a 70% win rate shows that he is better at PvZ, it does not show that there is no PvZ imbalance. Watching Bisu, there's no surprise that PvZ is his best matchup because of what he excels at (staggeringly good multitask) - something which is hugely beneficial in PvZ but which is much less of a factor in PvP or PvT (matchups that are to a greater degree defined by how well you control smaller goon reaver armies or how you perform in huge macro slugfests, respectively).
On July 14 2022 08:42 TMNT wrote: Dont know how you get 12-8 but here's what liquipedia says: Bisu vs Jaedong
For the other point. I was talking about how you used the example of Bisu 71.5% win rate in PvZ to disprove the notion that P have it worst. I wasn't talking about their (Bisu/Flash/Jaedong) records directly against each other.
But then again, if we ignore the super small sample size and go by their records against each other, then it proves exactly that T>Z>P (Kespa era only): - Flash 14-10 Bisu - Flash 26-25 Jaedong - Jaedong 10-10 Bisu Same if you go for number of medals. But I must add that this is no good way to determine balance.
Furthermore, Bisu has a much better record in PvZ than any zerg player has in ZvP. Neither JD nor Effort were able to break 70%. I will stress this point. The highest level protoss player has proven that PvZ is not zerg favored, and that cannot be attributed to Bisu's general skill, because in the other two matchups he performed significantly worse than in PvZ. JD's best matchup by far was ZvZ, even outmatching Flash's incredible TvT record. He also performed significantly worse in the other two matchups. This type of information proves that these were in fact the best players, and that their general skill doesn't explain why they dominated certain matchups more than others. Therefore Bisu's PvZ record does stand as a valid counter-argument against the claim that PvZ is zerg favored.
Bisu performing better in PvZ than PvP or PvT and having a 70% win rate shows that he is better at PvZ, it does not show that there is no PvZ imbalance. Watching Bisu, there's no surprise that PvZ is his best matchup because of what he excels at (staggeringly good multitask) - something which is hugely beneficial in PvZ but which is much less of a factor in PvP or PvT (matchups that are to a greater degree defined by how well you control smaller goon reaver armies or how you perform in huge macro slugfests, respectively).
I'm not arguing that Bisu's >70% PvZ winrate shows that there's no PvZ imbalance. I'm arguing that his winrate, when combined with several other facts, contradicts the claim of PvZ imbalance. No one other than Bisu has a winrate >70%, not a protoss player and not a zerg player. Everyone is at least 4% short, which is very significant. This is not explained by Bisu's overall skill, since, if PvZ were the worst protoss matchup, it would be expected that Bisu also perform worse in PvZ and better in PvT, but the opposite is the case. Or it would be expected that Bisu perform significantly better in PvT than he does, but that also isn't the case. Out of all of Bisu's skills, the most exceptional one is his PvZ and the least exceptional is his PvT, so Bisu's overall skill is an invalid argument. Instead a better hypothesis is that he understands PvZ on a level that is much greater than that of the top zerg players, and of course also of every protoss player. This shows that it's an understanding specific to PvZ, not overall skill, that pushes the PvZ winrate. Therefore the only remaining valid explanation for Bisu's PvZ winrate that would still allow for PvZ being imbalanced would be that the matchup is more difficult to figure out for protoss than it is for zerg. But there's no evidence for such difficulty, so it's pure speculation. In fact Bisu himself has left a long trail of information behind to be picked up and used by protoss players. An explanation that doesn't allow for PvZ imbalance is that protoss players are, as a group, performing further below the true potential of their race than zerg players, with Bisu being one of the few exceptions demonstrating where the true potential lies. Therefore the claim of PvZ favoring zerg remains speculative, as there is at least one equally valid or equally speculative explanation that supports the claim of no imbalance.
It's not just maps, it's also the metagame. FS and CB were considered to be balanced during Kespa era, but not anymore. Also certain leagues, (OSL in the past, ASL now) prefer exciting maps more than "balanced maps", (which I personally prefer). ASL maps been pretty anti-Terran , since they want to see people take down Flash.
And even though T>Z>P>T (or T>Z>>P>=T, if you will), holds over thousands of pro games. even large sample sizes can't avoid bias. Who knows what happens without Boxer/Savior/Bisu , zergs never find muta stacking, islands/semi-islands were more common, or meta-shifts that happened in more recent times (like TvZ with late mechanic/1-1-1, gate-expand PvZ, 9-7-3 ZvP, mass shuttle PvT, rax expand TvP, 2.5 hatch/late game queens ZvT, and so many more smaller ones I don't even remember).
On July 14 2022 08:42 TMNT wrote: Dont know how you get 12-8 but here's what liquipedia says: Bisu vs Jaedong
For the other point. I was talking about how you used the example of Bisu 71.5% win rate in PvZ to disprove the notion that P have it worst. I wasn't talking about their (Bisu/Flash/Jaedong) records directly against each other.
But then again, if we ignore the super small sample size and go by their records against each other, then it proves exactly that T>Z>P (Kespa era only): - Flash 14-10 Bisu - Flash 26-25 Jaedong - Jaedong 10-10 Bisu Same if you go for number of medals. But I must add that this is no good way to determine balance.
Furthermore, Bisu has a much better record in PvZ than any zerg player has in ZvP. Neither JD nor Effort were able to break 70%. I will stress this point. The highest level protoss player has proven that PvZ is not zerg favored, and that cannot be attributed to Bisu's general skill, because in the other two matchups he performed significantly worse than in PvZ. JD's best matchup by far was ZvZ, even outmatching Flash's incredible TvT record. He also performed significantly worse in the other two matchups. This type of information proves that these were in fact the best players, and that their general skill doesn't explain why they dominated certain matchups more than others. Therefore Bisu's PvZ record does stand as a valid counter-argument against the claim that PvZ is zerg favored.
Bisu performing better in PvZ than PvP or PvT and having a 70% win rate shows that he is better at PvZ, it does not show that there is no PvZ imbalance. Watching Bisu, there's no surprise that PvZ is his best matchup because of what he excels at (staggeringly good multitask) - something which is hugely beneficial in PvZ but which is much less of a factor in PvP or PvT (matchups that are to a greater degree defined by how well you control smaller goon reaver armies or how you perform in huge macro slugfests, respectively).
I'm not arguing that Bisu's >70% PvZ winrate shows that there's no PvZ imbalance. I'm arguing that his winrate, when combined with several other facts, contradicts the claim of PvZ imbalance. No one other than Bisu has a winrate >70%, not a protoss player and not a zerg player. Everyone is at least 4% short, which is very significant. This is not explained by Bisu's overall skill, since, if PvZ were the worst protoss matchup, it would be expected that Bisu also perform worse in PvZ and better in PvT, but the opposite is the case. Or it would be expected that Bisu perform significantly better in PvT than he does, but that also isn't the case. Out of all of Bisu's skills, the most exceptional one is his PvZ and the least exceptional is his PvT, so Bisu's overall skill is an invalid argument. Instead a better hypothesis is that he understands PvZ on a level that is much greater than that of the top zerg players, and of course also of every protoss player. This shows that it's an understanding specific to PvZ, not overall skill, that pushes the PvZ winrate. Therefore the only remaining valid explanation for Bisu's PvZ winrate that would still allow for PvZ being imbalanced would be that the matchup is more difficult to figure out for protoss than it is for zerg. But there's no evidence for such difficulty, so it's pure speculation. In fact Bisu himself has left a long trail of information behind to be picked up and used by protoss players. An explanation that doesn't allow for PvZ imbalance is that protoss players are, as a group, performing further below the true potential of their race than zerg players, with Bisu being one of the few exceptions demonstrating where the true potential lies. Therefore the claim of PvZ favoring zerg remains speculative, as there is at least one equally valid or equally speculative explanation that supports the claim of no imbalance.
I don't understand why you can't grasp the idea that a player could be so good at a matchup that his winrate is 70%+ despite that matchup being disadvantageous to him.
Here's an example to make it clearer: not so long ago, Snow played a showmatch vs a Chinese Zerg. At the start of the game, he was asked to kill 2 probes. That's no doubt imbalance, but Snow went on to win 2-0. Obviously if he played more his win rate would be like 80%. So does that 80% contradicts the idea that the games played was imbalance?
Mind that I'm not trying to prove PvZ is imbalance. I'm just trying to say that the two things (Bisu is good at PvZ and PvZ is imbalance) can exist without being contradictory.
"I don't understand why you can't grasp the idea that a player could be so good at a matchup that his winrate is 70%+ despite that matchup being disadvantageous to him."
Because Bisu only excelled this much in PvZ, not PvT and PvP. It wasn't just "Bisu being so much better". It was Bisu being so much better specifically in PvZ. In the other matchups he wasn't slacking, but he also wasn't shining anywhere near as much, which is in line with most other players who typically have one strong matchup, maybe two, but not three. The only exception to that rule is Flash. Not Bisu. Therefore it makes no sense to argue that Bisu was so good in PvZ simply because he was just such a strong player overall. He had the typical weaknesses. His PvT winrate reached 63%, not 71.5%, that's a huge difference, far worse than his PvZ, despite also being a great achievement.
You keep forgetting the context. No zerg player came close in ZvP to Bisu's PvZ winrate. The two best ones were at least 4% removed. This doesn't make sense unless Bisu figured out a way to play PvZ that not only erases every notion of imbalance, but even far surpasses it. He was not performing a few % below JD and Effort, as would be expected, but instead above. He overshot the goal by, I'd say, around 10% or so. That's not something someone just does without having a much deeper understanding of the matchup. Because we're talking about the most competitive era of BW, everyone was optimizing the game around the clock.
In YGOSU, it is mostly agreed upon that when three or more players are playing on team melee, protoss becomes the strongest race due to the likelihood of strong play-making potential from units such as early game harassment from probes having a single player dedicate all his focus on it, corsairs that never die throughout the entire game versus the zergs, and game changing late-game spellcasters such as arbiters.
I was thinking the same about Protoss before, and then I saw a few team melee games between Bisu+Best vs Light+Rush, and man oh man did the Protoss team totally destroy Terran.
The thing is, the Terran army during a fight always gets some sort of value even if you set your units up badly. If Protoss units are in range, they fire. Same thing can't be said for Protoss: + Dragoons hitting Depots or a floating Barrack. + Zealots getting on top of each other, and worse, eating a mine together. + High Templars evaporating before casting any storms + Shuttles dying mid-air because frankly, by the time you have the time to grab them they're already dead. All of the above examples can be mitigated by a great player but only to some extent. Chances are, if you are able to finish sorting out the targeting of your ground army, your Shuttles are likely on auto pilot and you won't be able to cast all the storms you'd like. But if you have two or three players controlling the same battle. The extra values Protoss can get is huge.
Same thing can be said for PvZ. No more scouting Probe dying early. No more High Templar full of energy dying before storm can get off. No more Corsair wasting.
Obviously Terran and Zerg can benefit a lot from team melee as well. But I feel like the extra values are nowhere near Protoss'.
Sure, Protoss players can just "A move" to win with range E-A rank in the ladder. It means Protoss is so strong enough suicide still win game.
Maybe, StarCraft lack of Protoss players has good skill after 20 years. I feel about strong Protoss when I have to fight in early game: Marine + Tank vs Zealot + Dragoon, sometime I lose even I have Tanks equal Dragoons.
Interesting discussion. I also don't fully believe Protoss is the worst race, it's heavily map-dependent and we have seen the best of all three races take games from each other. I also saw someone mention why other Protoss players don't copy Bisu for PvZ. If I recall properly, I think it was Best or another player who stated that they tried, but they can't seem to replicate what Bisu can do in that match-up.
This is misleading. When Flash is around, terran places 1st 63%, but only 40% when he isn't. When Flash is around, protoss places 2nd 80% of the time, but only 9% when he isn't. The data is so absurdly unbalanced in player skill that there's no way the results can be used for claims of matchup imbalance. Not to mention Bisu's presence/absence, which I didn't account for. Or perhaps Jaedong's. And so forth.
Edit: looked into how Bisu affects the results, and unsurprisingly protoss wins once and has a bunch of second places when he's there, while terran drops off after Flash leaves. As soon as Bisu leaves, protoss has no more hopes of first place, but still places second a few times, and Flash is still missing so terran keeps placing third.
The explanation for this is that KCM allows for winning streaks where one player goes undefeated. The best players play a disproportionate amount of games.
Again, that list of titles isn't very useful, especially not the ones from post-Kespa. The WCG Korea titles are particularly useless, because protoss was almost always hugely underrepresented (even in the prelims).
Edit: This also leads to one of the more relevant points regarding racial balance: representation. It's expected that a race that is represented by fewer players as opposed to more players would have lower odds of survival until the finals. Here's the distribution of Korean progamers with a TLPD page:
We can see a few interesting things. Terran representation is the most balanced, while zerg is very overrepresented and protoss is very underrepresented. People who remember the early days of SC may also remember the hype around zerg, as it was considered the race that benefits the most from raw APM, which is a very straight forward skill to succeed with in a video game. Koreans loved playing zerg and representation soared. Terran was the second most popular race, I'm guessing, mostly because of Boxer. But with it being considered the most tactically complex race in the early days, it wasn't equally appealing to the masses. Meanwhile protoss received almost no hype at all, except from Nal_rA perhaps.
Edit2: The more I look, the more cases of severe protoss underrepresentation I discover in pretty much all of the individual leagues. Only rare cases of protoss overrepresentation. And it gets even worse towards the early 2000's.
This also leads to one of the more relevant points regarding racial balance: representation. It's expected that a race that is represented by fewer players as opposed to more players would have lower odds of survival until the finals. Here's the distribution of Korean progamers with a TLPD page:
Dude I have several counters to your points but dont really have time to elaborate them for now. But it seems like your methodologies are always flawed, even starting from the data collection. I will quickly point out this one in this post:
Look at the distribution you found. You'd think there are more than 38 Protoss players in the history of Korean BW? That list from liquipedia is seriously lacking and all over the place: there are female BJs from the modern era listed there, but Snow, Mini, Shuttle, Rain,... etc. (all player during Kespa), for examples, are nowhere to be seen. So that distribution is probably way off the mark.
Edit: okay after further checking, you forgot to click "Further" to expand to the full list. But keep in mind that it's literally the list of every player ever, including amateurs, fast map players, female BJs and whatnot.
yeah Best said that he can't do PvZ like Bisu and had "given up". I was hearing Bisu can play it with a lot of correct guessing without seeing what's in the fog of war for Z.
I don't know about Bisu understanding PvZ so much more than other Protosses that even after looking at the replays or playing in the same team with him or talking with him that it didn't even shed some light on the others.
What sets him apart is mainly his skill set, notably the multi-tasking. That makes his corsairs never die which is vital in PvZ. That makes him able to pressure with zealots at different Zerg's bases (while at the same time still chasing lings in his main). A lot of Bisu's dominant PvZ comes down to that. His zealots got so much value in the early game and his corsairs just keep killing overlords that the Zerg can't recover economically and will just die to a midgame push.
On the other hand, Bisu isn't as good as Snow with reavers, and isn't as good as Best at expanding and macroing like a mad man, hence his PvT is a little bit behind those two.
PvZ and PvT are really different animals. An "overall skill" doesn't exist that makes a player good at a matchup also equally good at the other matchup.
This also leads to one of the more relevant points regarding racial balance: representation. It's expected that a race that is represented by fewer players as opposed to more players would have lower odds of survival until the finals. Here's the distribution of Korean progamers with a TLPD page:
Dude I have several counters to your points but dont really have time to elaborate them for now. But it seems like your methodologies are always flawed, even starting from the data collection. I will quickly point out this one in this post:
Look at the distribution you found. You'd think there are more than 38 Protoss players in the history of Korean BW? That list from liquipedia is seriously lacking and all over the place: there are female BJs from the modern era listed there, but Snow, Mini, Shuttle, Rain,... etc. (all player during Kespa), for examples, are nowhere to be seen. So that distribution is probably way off the mark.
Edit: okay after further checking, you forgot to click "Further" to expand to the full list. But keep in mind that it's literally the list of every player ever, including amateurs, fast map players, female BJs and whatnot.
You can point out when I get something wrong, that's more than alright. But handwaving away my findings without providing or guiding to corrected data is not going to impress me.
So what is the true ratio of T/Z/P players among Korean pros during the Kespa era, or any other era? Can you help me find the numbers?
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
On July 16 2022 03:43 TMNT wrote: I don't know about Bisu understanding PvZ so much more than other Protosses that even after looking at the replays or playing in the same team with him or talking with him that it didn't even shed some light on the others.
What sets him apart is mainly his skill set, notably the multi-tasking. That makes his corsairs never die which is vital in PvZ. That makes him able to pressure with zealots at different Zerg's bases (while at the same time still chasing lings in his main). A lot of Bisu's dominant PvZ comes down to that. His zealots got so much value in the early game and his corsairs just keep killing overlords that the Zerg can't recover economically and will just die to a midgame push.
On the other hand, Bisu isn't as good as Snow with reavers, and isn't as good as Best at expanding and macroing like a mad man, hence his PvT is a little bit behind those two.
PvZ and PvT are really different animals. An "overall skill" doesn't exist that makes a player good at a matchup also equally good at the other matchup.
I think your argument isn't perfectly sound. On the one hand you're saying "What sets him apart is mainly his skill set, notably the multi-tasking."
On the other hand you're also saying "An "overall skill" doesn't exist that makes a player good at a matchup also equally good at the other matchup."
If mainly his skill set sets him apart, and it's not a skill specific to PvZ, then it fails to explain his PvZ winrate, since it doesn't also explain his winrate in the other two matchups. You haven't pointed to a PvZ-specific skill, but instead to multi-tasking. Such a skill is roughly equally useful in every protoss matchup. The argument doesn't work.
On July 16 2022 03:43 TMNT wrote: I don't know about Bisu understanding PvZ so much more than other Protosses that even after looking at the replays or playing in the same team with him or talking with him that it didn't even shed some light on the others.
Don't you think as a very top player you'd want to tell nobody at all about some of the keys of your plays in some match up? It's the only way to ensure the opponents don't know for sure how you approach the match up and etc.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Now, it's hard to reach a consensus as to whether P is the worst race (at pro level). But I'm more inclined to that opinion because there are too many evidences, spanning over the course of 20+ odd years, pointing toward that:
- P is the least successful race in Starleagues (all of them). See picture above.
- There is no Protoss bonjwa
- KCM created a league where the 3 races compete directly with each other. No mirrors. Only top players were selected to play (like, players in the top 5 of each race). Sometimes there were the odd appearances of a weaker player (like top 10/15), but it's for all 3, not just P. The result: after 20 seasons, P is the worst, by a HUGE margin. For me this is the strongest evidence.
- Sponbbang: since the Afreeca era, results of thousands of games from every active player were recorded. And again, PvZ and PvT win rate are all less than 50% (both around 45% I think, which is really significant given the sample size is tremendous).
- There is the data that shows P represents ~50% of the players on the ladder but only ~40% in S rank.
(There is also the gameplay aspect. But let's not go down to that nightmare path)
The thing is, with that many evidences, over that long of time, and that big of sample size, all pointing towards it. P is the worst race is the most likely explanation than any other theories.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You're citing a ZvP winrate of 54% in a post where you're arguing against ZvP imbalance? Nobody (reasonable) is claiming that the imbalance is insurmountable or whatever, certain players excel at PvZ (making them better at PvZ than they are at PvT or PvP), certain map designs favor P over Z, but overall, across most top players and across most maps the imbalance exists, and it manifests through Zerg winning roughly 54% of the time instead of the 50% they'd win if it was entirely balanced.
https://sponbbang.com/race/ is a massive database of matches between progamers. Over slightly short of 43000 games of PvZ, we're left with these stats: 23087 wins for Zerg, 19577 wins for protoss, giving an overall win% of 54.1% for Zerg.
(TvP is 20434 - 18319 - 52.7% win for terran, TvZ is 26581 - 21165, 55.7% win for terran. Terran has an overall win rate of 54.4%, Protoss has an overall win rate of 46.5%, Zerg has an overall win rate of 48.9%)
https://sponbbang.com/race/ is a massive database of matches between progamers. Over slightly short of 43000 games of PvZ, we're left with these stats: 23087 wins for Zerg, 19577 wins for protoss, giving an overall win% of 54.1% for Zerg.
In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
In regards to the observed 54% ZvP winrate, people are using it as a gotcha, but this is fallacious. An observed winrate does not necessarily tell us something about balance. The reason is that the demonstration of skill in a given player pool is dependent on their knowledge of the game. As every veteran SC player knows, Bisu revolutionized PvZ back in 2007 or so. Without his findings, the ZvP winrate would likely be significantly >54% (for example it could be 57%). Demonstration of skill, even in a very large player pool over a large sample of games, is therefore not a perfect indicator of balance, because with hindsight we know that a 57% ZvP winrate would absolutely have to be a false representation of the true balance of the matchup. The same can (not must) therefore also be true for the 54% winrate. And since it's 4% removed from perfect balance, that means not too much improvement would be required to achieve a level of balance that is acceptable.
Not only that, but the observed TvZ winrate favors terran almost to the exact same degree (~54%), but I don't hear zerg players complain about it. (And my suspicion why zerg players aren't complaining is that, due to their historic overrepresentation in tournaments, they've had their fair share of titles, unlike protoss)
On July 16 2022 13:31 Avi-Love wrote: In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
Most people regard the years leading up to and including 2012 the most competitive era, because that's when the trio of Flash/JD/Bisu was dominating the scene. That era is almost always mentioned first. Today's standard of games is quite significantly lower, albeit still at a very high level.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
In regards to the observed 54% ZvP winrate, people are using it as a gotcha, but this is fallacious. An observed winrate does not necessarily tell us something about balance. The reason is that the demonstration of skill in a given player pool is dependent on their knowledge of the game. As every veteran SC player knows, Bisu revolutionized PvZ back in 2007 or so. Without his findings, the ZvP winrate would likely be significantly >54% (for example it could be 57%). Demonstration of skill, even in a very large player pool over a large sample of games, is therefore not a perfect indicator of balance, because with hindsight we know that a 57% ZvP winrate would absolutely have to be a false representation of the true balance of the matchup. The same can (not must) therefore also be true for the 54% winrate. And since it's 4% removed from perfect balance, that means not too much improvement would be required to achieve a level of balance that is acceptable.
This goes both ways. It could be the case that a demonstration of skill hasn't fully happened on the Z side, and the actual bslsnce is > 54%. This is all about what if. What really matters is what we witness, and historically, P is miles behind in terns of success.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
In regards to the observed 54% ZvP winrate, people are using it as a gotcha, but this is fallacious. An observed winrate does not necessarily tell us something about balance. The reason is that the demonstration of skill in a given player pool is dependent on their knowledge of the game. As every veteran SC player knows, Bisu revolutionized PvZ back in 2007 or so. Without his findings, the ZvP winrate would likely be significantly >54% (for example it could be 57%). Demonstration of skill, even in a very large player pool over a large sample of games, is therefore not a perfect indicator of balance, because with hindsight we know that a 57% ZvP winrate would absolutely have to be a false representation of the true balance of the matchup. The same can (not must) therefore also be true for the 54% winrate. And since it's 4% removed from perfect balance, that means not too much improvement would be required to achieve a level of balance that is acceptable.
This goes both ways. It could be the case that a demonstration of skill hasn't fully happened on the Z side, and the actual bslsnce is > 54%. This is all about what if. What really matters is what we witness, and historically, P is miles behind in terns of success.
Yes, of course it could be the other way around, too. Maybe zerg is underperforming. That's why I'm not making the claim that PvZ is in fact balanced, but only denying that the evidence for imbalance is sufficient. The winrate of a population can definitely be off by a few %, because there are still biases. And pointing to titles is among the worst ways of proving imbalance.
Chess is actually one of the best examples that proves that population performance (even at the top level) is not indicative of game balance. It's almost universally accepted that every perfectly played game of chess will end in a draw. However, that's not what we observe. The top players regularly win and lose games at a very high rate. This shows that observed winrates don't reflect the true potential. Even the world's best engines (performing at upwards of 3800 ELO) are unable to draw every single game.
In SC:R, we don't even have any engines that could help us improve our understanding, so the expectation is that we win or lose a lot more games than what could be the case.
On July 16 2022 13:31 Avi-Love wrote: In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
Most people regard the years leading up to and including 2012 the most competitive era, because that's when the trio of Flash/JD/Bisu was dominating the scene. That era is almost always mentioned first. Today's standard of games is quite significantly lower, albeit still at a very high level.
You don't know what you're talking about, please stop posting. Firstly, It's very evident that bw has evolved *a lot* since 2010, and while progamers admit that maybe the mechnical skill of 2010 was slightly higher, the knowledge and meta has improved leaps and bounds -- it's fairly obvious to any viewer above a certain mmr that the game is being played at a higher level now than it was at the time. Secondly, 2012 is a complete figment of your imagination, the 'competitiveness' of bw started declining the second progamers received access to the SC2 beta, which was in 2010. Unless you've deluded yourself into thinking progamers playing two games, and proleague having sc2 increased the skill level of bw...? Thirdly, who are "most" people? I'd love to see your data on this, everyone I know that plays this game understands that it has *naturally* evolved over time. I think your obsesssion with both Bisu and Jaedong is a clear sign that you're living in the past and your mental gymnastics are stuck in the mud. Jaedong is garbage and has been for years, and Bisu chokes so hard that it's fairly obvious he's not even the best protoss player in a competitive setting. Lastly, it is obvious that zvp is slightly zerg favoured. It is also fairly obvious that it has been for quite awhile -- the biggest change in balance over the last couple of years has been tvp swinging slightly towards being terran favoured.
So what does all of this mean? Does it mean bw needs a balance patch? Negative. Can it be fixed through maps? Absolutely. Is the goal of the ASL to have perfect map balance? Obviously not, they clearly value throwing in a couple of curveballs which is evident from historical data (Third world, Sparkle, Inner Coven etc). If you're a dumb foreigner losing x matchup at 1900 mmr on the ladder is it because of your race? No, you just suck. Stop having a shitty mentality and accept your shortcomings -- but it's a lot easier to whine and blame other factors than it is to look inward, hence this thread and the many replies in it, and why it's so pointless and should have been stopped straight away.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
What? Have you not read a word I wrote before that bold part? Or the example I gave after that bold part? What is this madness???
Your example (of 16 players) is literally the case where every single player entering the competition is equally good and have the same chance of winning (hence you have all of them at 50% win rate) - which is never the case. How could you not understand this?
On July 16 2022 13:31 Avi-Love wrote: In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
Most people regard the years leading up to and including 2012 the most competitive era, because that's when the trio of Flash/JD/Bisu was dominating the scene. That era is almost always mentioned first. Today's standard of games is quite significantly lower, albeit still at a very high level.
You don't know what you're talking about, please stop posting. Firstly, It's very evident that bw has evolved *a lot* since 2010, and while progamers admit that maybe the mechnical skill of 2010 was slightly higher, the knowledge and meta has improved leaps and bounds -- it's fairly obvious to any viewer above a certain mmr that the game is being played at a higher level now than it was at the time. Secondly, 2012 is a complete figment of your imagination, the 'competitiveness' of bw started declining the second progamers received access to the SC2 beta, which was in 2010. Unless you've deluded yourself into thinking progamers playing two games, and proleague having sc2 increased the skill level of bw...? Thirdly, who are "most" people? I'd love to see your data on this, everyone I know that plays this game understands that it has *naturally* evolved over time. I think your obsesssion with both Bisu and Jaedong is a clear sign that you're living in the past and your mental gymnastics are stuck in the mud. Jaedong is garbage and has been for years, and Bisu chokes so hard that it's fairly obvious he's not even the best protoss player in a competitive setting. Lastly, it is obvious that zvp is slightly zerg favoured. It is also fairly obvious that it has been for quite awhile -- the biggest change in balance over the last couple of years has been tvp swinging slightly towards being terran favoured.
So what does all of this mean? Does it mean bw needs a balance patch? Negative. Can it be fixed through maps? Absolutely. Is the goal of the ASL to have perfect map balance? Obviously not, they clearly value throwing in a couple of curveballs which is evident from historical data (Third world, Sparkle, Inner Coven etc). If you're a dumb foreigner losing x matchup at 1900 mmr on the ladder is it because of your race? No, you just suck. Stop having a shitty mentality and accept your shortcomings -- but it's a lot easier to whine and blame other factors than it is to look inward, hence this thread and the many replies in it, and why it's so pointless and should have been stopped straight away.
I'm not going to stop posting just because you ask me so nicely, tyvm. And I know very well what I'm talking about, because I've been around since the very early days of SC, I know the historical development of most of the strategies, discoveries and trends, I've followed the pro scene and a decent portion of the amateur scene, and I have a good understanding of statistical fallacies. The balance debate has raged for decades and it was never settled. It's never been proven beyond doubt that PvZ is actually imbalanced, it's only been shown that protoss players are losing more than zerg players. That by itself is not proof of imbalance, it's only an indicator that needs supporting evidence, of which none has ever been found.
2012 marks the end of the most competitive era of BW because of the migration of much of the pro scene to SC2. The latter is not up to debate, it's a fact.
Furthermore, this observation "the biggest change in balance over the last couple of years has been tvp swinging slightly towards being terran favoured." - i.e. terran having a positive record against protoss post-Kespa - is a perfect example of the observed winrate changing over time by several percentage points. Several percentage points. Ring a bell? 54% ZvP winrate? If this is a natural development, then it can just as well happen in the other matchups without any changes to the balance. Somehow your argument is that ZvP can only be imbalanced as demonstrated by the 54% ZvP winrate, even though provably another matchup statistic in TvP has demonstrably shifted over several years, which can just as much happen in ZvP as well (in either direction).
You're lazer focusing on the ZvP winrate to support your hypothesis of imbalance, but you're ignoring other data points when they point to some unknown variables being in play that contradict your hypothesis.
Edit: Furthermore, you're contradicting yourself. On the one hand you're rejecting the claim that the late Kespa era was the most competitive era, on the other hand you're outright admitting that JD's and Bisu's skills have strongly declined.
"Jaedong is garbage and has been for years, and Bisu chokes so hard that it's fairly obvious he's not even the best protoss player in a competitive setting."
Please point to one player post-Kespa who has a winrate of >70% in any of the matchups, other than Flash. I'll wait.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
What? Have you not read a word I wrote before that bold part? Or the example I gave after that bold part? What is this madness???
Your example (of 16 players) is literally the case where every single player entering the competition is equally good and have the same chance of winning (hence you have all of them at 50% win rate) - which is never the case. How could you not understand this?
The effect is always the same regardless of the winrate, unless a player has a 100% winrate. If you flood a field with protoss players, even if they have a winrate that is lower than 50%, the likelihood will increase that the finals consist of one or more protoss players.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
What? Have you not read a word I wrote before that bold part? Or the example I gave after that bold part? What is this madness???
Your example (of 16 players) is literally the case where every single player entering the competition is equally good and have the same chance of winning (hence you have all of them at 50% win rate) - which is never the case. How could you not understand this?
The effect is always the same regardless of the winrate, unless a player has a 100% winrate. If you flood a field with protoss players, even if they have a winrate that is lower than 50%, the likelihood will increase that the finals consist of one or more protoss players.
Sure, but the effect isn't as strong as you think because the winner doesn't play every other entrant in series. Every time you double the entrants the winner plays 1 more game.
2 entrants: 1 game 4 entrants: 2 games 8 entrants: 3 games
At this point you can add another 8 PokJus to the tournament and Flash only has to beat one more. Then you have to add another 16 PokJus to reduce Flash's win chance again.
The reason you're getting a nice pretty number like 6.25% that sounds like it supports your argument is because you're making assumptions that lead to Flash having a 12.5% winrate in tournament vs 7 PokJus.
If Flash has an 75% TvP series winrate (against top pros) in title playoffs that receive his full focus you end up having to add a truly absurd amount of PokJus to make Protoss likely to win.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
What? Have you not read a word I wrote before that bold part? Or the example I gave after that bold part? What is this madness???
Your example (of 16 players) is literally the case where every single player entering the competition is equally good and have the same chance of winning (hence you have all of them at 50% win rate) - which is never the case. How could you not understand this?
The effect is always the same regardless of the winrate, unless a player has a 100% winrate. If you flood a field with protoss players, even if they have a winrate that is lower than 50%, the likelihood will increase that the finals consist of one or more protoss players.
Sure, but the effect isn't as strong as you think because the winner doesn't play every other entrant in series. Every time you double the entrants the winner plays 1 more game.
2 entrants: 1 game 4 entrants: 2 games 8 entrants: 3 games
At this point you can add another 8 PokJus to the tournament and Flash only has to beat one more. Then you have to add another 16 PokJus to reduce Flash's win chance again.
The reason you're getting a nice pretty number like 6.25% that sounds like it supports your argument is because you're making assumptions that lead to Flash having a 12.5% winrate in tournament vs 7 PokJus.
If Flash has an 75% TvP series winrate (against top pros) in title playoffs that receive his full focus you end up having to add a truly absurd amount of PokJus to make Protoss likely to win.
Of course the effect is smaller or greater depending on the exact field, but the effect is always the same. With each protoss player you add to the field, the likelihood of a protoss player in the finals goes up, and thus also the likelihood for more protoss titles. You can actually do the math for yourself with various field sizes and using only two races for simplification sake, adjust the winrates, adjust the racial representation, do it as much as you want. You'll see a superior player's chance of survival go down incrementally with each additional player in the field. And as a collective the same thing happens. The fact remains that a severe underrepresentation of protoss players has been observed in the tournaments, and so a smaller number of titles is perfectly expected due to underrepresentation.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
What? Have you not read a word I wrote before that bold part? Or the example I gave after that bold part? What is this madness???
Your example (of 16 players) is literally the case where every single player entering the competition is equally good and have the same chance of winning (hence you have all of them at 50% win rate) - which is never the case. How could you not understand this?
The effect is always the same regardless of the winrate, unless a player has a 100% winrate. If you flood a field with protoss players, even if they have a winrate that is lower than 50%, the likelihood will increase that the finals consist of one or more protoss players.
No it won't, unless you flood a field with Bisus and Storks. If you flood a field of Britneys, they will all get knocked out before the quarterfinals lol. In theory, yeah you'll get an increase of chance, but only marginally, like less than 1% or something. Have you heard of a thing call betting odds? It happens in every sport. If you add 100 other low level tennis players to the Roland Gaross, the finals will just consist of some guys from Spain/Serbia/Swiss anyway.
But you can't even flood a field with Protoss players anyway. It's just impossible. In reality, like I said, they got wiped off during the qualifiers already.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
You're missing the point.
I provided the data that protoss IS in fact underrepresented. From that underrepresentation it can only be concluded that a protoss winning a title is less likely, and therefore it's a good explanation for a lack of protoss titles.
Whatever causes the underrepresentation of protoss players is another story. I have ideas about that, and my hypothesis is that it's mainly due to a zerg overrepresentation from a historic preference for that race. That part of my argumentation is speculative and you can choose to ignore it, but you can't ignore the observation that zerg is in fact overrepresented in the tournaments.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
What? Have you not read a word I wrote before that bold part? Or the example I gave after that bold part? What is this madness???
Your example (of 16 players) is literally the case where every single player entering the competition is equally good and have the same chance of winning (hence you have all of them at 50% win rate) - which is never the case. How could you not understand this?
The effect is always the same regardless of the winrate, unless a player has a 100% winrate. If you flood a field with protoss players, even if they have a winrate that is lower than 50%, the likelihood will increase that the finals consist of one or more protoss players.
No it won't, unless you flood a field with Bisus and Storks. If you flood a field of Britneys, they will all get knocked out before the quarterfinals lol. In theory, yeah you'll get an increase of chance, but only marginally, like less than 1% or something. Have you heard of a thing call betting odds? It happens in every sport. If you add 100 other low level tennis players to the Roland Gaross, the finals will just consist of some guys from Spain/Serbia/Swiss anyway.
But you can't even flood a field with Protoss players anyway. It's just impossible. In reality, like I said, they got wiped off during the qualifiers already.
No, you're wrong. And it's very easy to prove this with mathematics.
I'm going to assume that there's a field of players with Flash in it, who has a 95% winrate (per match) against each individual player. The winrate of the other players against one another is irrelevant for this example, because they act as a collective. We're searching for Flash's survival rate against all players, i.e. his odds of winning the tournament, which is single-elimination.
1) There are four players in the field (Flash has three competitors)
Flash plays a match and moves into the finals with a 95% chance. Flash plays the finals and wins with a 95% chance. Total chance of Flash winning the tournament = 95% * 0.95 95% * 0.95 = 90.25%
2) There are eight players in the field (Flash has seven competitors)
Flash plays a match and moves into the semi-finals with a 95% chance. Flash plays the semi-finals and wins with a 95% chance. Flash plays the finals and wins with a 95% chance. Total chance of Flash winning the tournament = 95% * 0.95 * 0.95 95% * 0.95 * 0.95 = 85.7375%
Observation: doubling the field size decreases Flash's odds of survival by ~4.51%
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
What? Have you not read a word I wrote before that bold part? Or the example I gave after that bold part? What is this madness???
Your example (of 16 players) is literally the case where every single player entering the competition is equally good and have the same chance of winning (hence you have all of them at 50% win rate) - which is never the case. How could you not understand this?
The effect is always the same regardless of the winrate, unless a player has a 100% winrate. If you flood a field with protoss players, even if they have a winrate that is lower than 50%, the likelihood will increase that the finals consist of one or more protoss players.
Sure, but the effect isn't as strong as you think because the winner doesn't play every other entrant in series. Every time you double the entrants the winner plays 1 more game.
2 entrants: 1 game 4 entrants: 2 games 8 entrants: 3 games
At this point you can add another 8 PokJus to the tournament and Flash only has to beat one more. Then you have to add another 16 PokJus to reduce Flash's win chance again.
The reason you're getting a nice pretty number like 6.25% that sounds like it supports your argument is because you're making assumptions that lead to Flash having a 12.5% winrate in tournament vs 7 PokJus.
If Flash has an 75% TvP series winrate (against top pros) in title playoffs that receive his full focus you end up having to add a truly absurd amount of PokJus to make Protoss likely to win.
Of course the effect is smaller or greater depending on the exact field, but the effect is always the same. With each protoss player you add to the field, the likelihood of a protoss player in the finals goes up, and thus also the likelihood for more protoss titles. You can actually do the math for yourself with various field sizes and using only two races for simplification sake, adjust the winrates, adjust the racial representation, do it as much as you want. You'll see a superior player's chance of survival go down incrementally with each additional player in the field. And as a collective the same thing happens. The fact remains that a severe underrepresentation of protoss players has been observed in the tournaments, and so a smaller number of titles is perfectly expected due to underrepresentation.
It goes down every time you double the field. This is because every time you flood the field with Protoss players they end up mostly playing and eliminating other Protoss players so the value you get for adding another Protoss player goes down with every Protoss player you add.
If you already have the best 31 Protoss players in the world you have to add another 32 Protoss players to add another round to the tournament. What exactly do you think is the bo3 winrate of the 32nd-63rd best Protoss player in the world? If you go again the game you're adding to Flash's tournament is the 64th-127th best Protoss player in the world. At this point we're probably actually starting to get into the 100% winrate area where your admit your assumptions no longer hold.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
First of all, this way of reasoning is fundamentally wrong. The bold part is only true if you consider every single player entering the competition is equally good and has the same chance of winning. You can add 100 another Britneys to the Protoss list and it'd increase 0 title for the race.
As always, correlation doesn't imply causation. Using your way of thinking, I can explain the reason Asian teams have never won a world cup is under-representation. I hope now you can see the flaw in your logic.
Secondly, this data is too raw. I performed the same filter that you probably did and got those numbers too - but this is literally a list of players who have an entry on liquipedia, regardless of era and level. Many have never played an "SL", and I still see some female BJs there. You might as well pull the player pool from ladder (in that case, Protoss overrepresenting at 50%). And the opposite also applies: there are progamers who played OSL/MSL who still don't have a liquipedia entry.
A better approach would be to count the number of players for each race in each SL, but I doubt any of us have time to do that. But then again, even if you find an under-representation of Protoss, it goes back to point one.
And tbh, I think if Protoss is under-represented at SLs, it actually supports the argument that Protoss is the worst race at the top level. You know why? Because a bigger proportion of them was wiped off in the qualifiers already. It also agrees with the fact that Protoss players account for like 50% of the ladder pool, but only like 40% in S rank.
Your claim (bold part) is wrong. To illustrate this, if you have a field of 16 players which consists of 15 A and 1 B, with both A and B having a 50% winrate, the odds of winning a single-elimination tournament are 6.25% for B. If you even out the number of A and B to 8 each, the odds of winning are 50% for both A and B. So you have a fundamental misunderstanding of how the representation of a race affects overall win probability in a tournament.
What? Have you not read a word I wrote before that bold part? Or the example I gave after that bold part? What is this madness???
Your example (of 16 players) is literally the case where every single player entering the competition is equally good and have the same chance of winning (hence you have all of them at 50% win rate) - which is never the case. How could you not understand this?
The effect is always the same regardless of the winrate, unless a player has a 100% winrate. If you flood a field with protoss players, even if they have a winrate that is lower than 50%, the likelihood will increase that the finals consist of one or more protoss players.
Sure, but the effect isn't as strong as you think because the winner doesn't play every other entrant in series. Every time you double the entrants the winner plays 1 more game.
2 entrants: 1 game 4 entrants: 2 games 8 entrants: 3 games
At this point you can add another 8 PokJus to the tournament and Flash only has to beat one more. Then you have to add another 16 PokJus to reduce Flash's win chance again.
The reason you're getting a nice pretty number like 6.25% that sounds like it supports your argument is because you're making assumptions that lead to Flash having a 12.5% winrate in tournament vs 7 PokJus.
If Flash has an 75% TvP series winrate (against top pros) in title playoffs that receive his full focus you end up having to add a truly absurd amount of PokJus to make Protoss likely to win.
Of course the effect is smaller or greater depending on the exact field, but the effect is always the same. With each protoss player you add to the field, the likelihood of a protoss player in the finals goes up, and thus also the likelihood for more protoss titles. You can actually do the math for yourself with various field sizes and using only two races for simplification sake, adjust the winrates, adjust the racial representation, do it as much as you want. You'll see a superior player's chance of survival go down incrementally with each additional player in the field. And as a collective the same thing happens. The fact remains that a severe underrepresentation of protoss players has been observed in the tournaments, and so a smaller number of titles is perfectly expected due to underrepresentation.
It goes down every time you double the field. This is because every time you flood the field with Protoss players they end up mostly playing and eliminating other Protoss players so the value you get for adding another Protoss player goes down with every Protoss player you add.
If you already have the best 31 Protoss players in the world you have to add another 32 Protoss players to add another round to the tournament. What exactly do you think is the bo3 winrate of the 32nd-63rd best Protoss player in the world? If you go again the game you're adding to Flash's tournament is the 64th-127th best Protoss player in the world. At this point we're probably actually starting to get into the 100% winrate area where your admit your assumptions no longer hold.
Of course some of the newly added players would eliminate one another, but it also increases their chance of survival to a later point in the tournament, for one because they either face one another more often (mirror matchups provide a 100% chance of a player of that race advancing to the next round), and also because there are a few more hurdles to overcome for the players of the opposing races. The fact remains that the total chance of survival for that race goes up very significantly. And every little bit matters when it comes to survival against a field, i.e. more titles for that race. If one race is represented with only 27% of players, those players cannot overcome that (as a collective) in the long run unless they demonstrate a superior winrate.
You can keep believing that leveling out the field would only have slightly improved the odds of protoss titles, but that's simply not right. The odds would've gone up a whole lot. Again, I can post the field sizes from individual tournaments.
2) 2011 ABC Mart MBCGame StarCraft League (slightly favoring zerg over protoss/terran) Zerg 12 Terran 11 Protoss 9
And this is an extreme example (there are quite a bunch like that):
3) 2010 Korean Air OnGameNet Starleague Season 1 Zerg 18 Terran 14 Protoss 8
And of course, there are also some examples when protoss was overrepresented, but those are quite rare, and it was never by a large margin. Also note that the cases of zerg overrepresentation are very numerous. Terran and protoss are usually underrepresented, in particular protoss. Fewer titles are obviously expected (for anyone who understand probabilities).
Oh and also, the argument that a flood of players would partially eliminate one another, that is of course true, but it also goes the other way. If they're less numerous and face one another, their odds of collective survival go down in that case as well. Only that they're less likely to face one another if they're less numerous, but the impact of eliminating one another also becomes greater.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
You're missing the point.
I provided the data that protoss IS in fact underrepresented. From that underrepresentation it can only be concluded that a protoss winning a title is less likely, and therefore it's a good explanation for a lack of protoss titles.
Less Protoss representation -> Less Protoss wins. This is sound, yes. Less Protoss wins -> Less Protoss representation. But this is also sound. And it leads to: [...] Less Protoss wins -> Less Protoss representation -> Less Protoss wins -> Less Protoss representation [...] This is a positive feedback loop. You can't just walk into it at an arbitrary point and call it the start and say it explains things.
Whatever causes the underrepresentation of protoss players is another story. I have ideas about that, and my hypothesis is that it's mainly due to a zerg overrepresentation from a historic preference for that race. That part of my argumentation is speculative and you can choose to ignore it, but you can't ignore the observation that zerg is in fact overrepresented in the tournaments.
The fact that you call this 'another story' and don't realise it is the whole story is the problem. Your headcannon that Protoss representation was repressed because of non-gameplay bias toward playing Zerg is extremely questionable and the fact that you imply this affected Protoss more than Terran already implies a race balance issue anyway. You don't see heroes with terrible winrates flood the pickrates at TI. You don't see shit guns flood the buy rates at a CSGO major. You don't see low tier characters flood top8 at Evo. There's no good reason to believe that Broodwar is somehow different.
Protoss has a lower % of champions (17 of 73 for osl msl asl, or 23%) than they have % of competitors. Not doing the math, but I'm guessing this difference roughly matches what a 46.5% win rate gets you.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
You're missing the point.
I provided the data that protoss IS in fact underrepresented. From that underrepresentation it can only be concluded that a protoss winning a title is less likely, and therefore it's a good explanation for a lack of protoss titles.
Less Protoss representation -> Less Protoss wins. This is sound, yes. Less Protoss wins -> Less Protoss representation. But this is also sound. And it leads to: [...] Less Protoss wins -> Less Protoss representation -> Less Protoss wins -> Less Protoss representation [...] This is a positive feedback loop. You can't just walk into it at an arbitrary point and call it the start and say it explains things.
Whatever causes the underrepresentation of protoss players is another story. I have ideas about that, and my hypothesis is that it's mainly due to a zerg overrepresentation from a historic preference for that race. That part of my argumentation is speculative and you can choose to ignore it, but you can't ignore the observation that zerg is in fact overrepresented in the tournaments.
The fact that you call this 'another story' and don't realise it is the whole story is the problem. Your headcannon that Protoss representation was repressed because of non-gameplay bias toward playing Zerg is extremely questionable and the fact that you imply this affected Protoss more than Terran already implies a race balance issue anyway. You don't see heroes with terrible winrates flood the pickrates at TI. You don't see shit guns flood the buy rates at a CSGO major. You don't see low tier characters flood top8 at Evo. There's no good reason to believe that Broodwar is somehow different.
If you want to explain the lacking protoss representation using the protoss winrate(s), then you'd have to also use the same reasoning to explain the zerg overrepresentation. Unfortunately this leads to a contradiction. Zerg players have the most balanced overall winrate of approximately 50% (or very slightly above). This would mean that, if the winrate dictated representation, then zerg players would be expected to make up approximately 33.3% of a tournament field on average. But this doesn't match the observation, as zerg is clearly very overrepresented, as much if not even more overrepresented than protoss is underrepresented. Furthermore, terran should be the most overrepresented race, but this doesn't match the observation either.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
You're missing the point.
I provided the data that protoss IS in fact underrepresented. From that underrepresentation it can only be concluded that a protoss winning a title is less likely, and therefore it's a good explanation for a lack of protoss titles.
Less Protoss representation -> Less Protoss wins. This is sound, yes. Less Protoss wins -> Less Protoss representation. But this is also sound. And it leads to: [...] Less Protoss wins -> Less Protoss representation -> Less Protoss wins -> Less Protoss representation [...] This is a positive feedback loop. You can't just walk into it at an arbitrary point and call it the start and say it explains things.
Whatever causes the underrepresentation of protoss players is another story. I have ideas about that, and my hypothesis is that it's mainly due to a zerg overrepresentation from a historic preference for that race. That part of my argumentation is speculative and you can choose to ignore it, but you can't ignore the observation that zerg is in fact overrepresented in the tournaments.
The fact that you call this 'another story' and don't realise it is the whole story is the problem. Your headcannon that Protoss representation was repressed because of non-gameplay bias toward playing Zerg is extremely questionable and the fact that you imply this affected Protoss more than Terran already implies a race balance issue anyway. You don't see heroes with terrible winrates flood the pickrates at TI. You don't see shit guns flood the buy rates at a CSGO major. You don't see low tier characters flood top8 at Evo. There's no good reason to believe that Broodwar is somehow different.
If you want to explain the lacking protoss representation using the protoss winrate(s), then you'd have to also use the same reasoning to explain the zerg overrepresentation. Unfortunately this leads to a contradiction. Zerg players have the most balanced overall winrate of approximately 50% (or very slightly above). This would mean that, if the winrate dictated representation, then zerg players would be expected to make up approximately 33.3% of a tournament field on average. But this doesn't match the observation, as zerg is clearly very overrepresented, as much if not even more overrepresented than protoss is underrepresented. Furthermore, terran should be the most overrepresented race, but this doesn't match the observation either.
No, because I think it is more complicated than you make out. I think the representation is more in line with the winrates at practice partner level. Titles are more in line with the winrates at Flash level.
Terran being hard to win with at low levels might suppress Terran representation: Few candidate pros. Protoss being hard to win with at high levels might suppress Protoss representation: Many candidate pros, few actual pros. Zerg having a good vP matchup and bad vT matchup would end up overrepresented in such an environment.
Maybe I am wrong, I don't claim to have all the answers, but I think it's a better theory than "it was just random/unrelated."
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
You're missing the point.
I provided the data that protoss IS in fact underrepresented. From that underrepresentation it can only be concluded that a protoss winning a title is less likely, and therefore it's a good explanation for a lack of protoss titles.
Less Protoss representation -> Less Protoss wins. This is sound, yes. Less Protoss wins -> Less Protoss representation. But this is also sound. And it leads to: [...] Less Protoss wins -> Less Protoss representation -> Less Protoss wins -> Less Protoss representation [...] This is a positive feedback loop. You can't just walk into it at an arbitrary point and call it the start and say it explains things.
Whatever causes the underrepresentation of protoss players is another story. I have ideas about that, and my hypothesis is that it's mainly due to a zerg overrepresentation from a historic preference for that race. That part of my argumentation is speculative and you can choose to ignore it, but you can't ignore the observation that zerg is in fact overrepresented in the tournaments.
The fact that you call this 'another story' and don't realise it is the whole story is the problem. Your headcannon that Protoss representation was repressed because of non-gameplay bias toward playing Zerg is extremely questionable and the fact that you imply this affected Protoss more than Terran already implies a race balance issue anyway. You don't see heroes with terrible winrates flood the pickrates at TI. You don't see shit guns flood the buy rates at a CSGO major. You don't see low tier characters flood top8 at Evo. There's no good reason to believe that Broodwar is somehow different.
If you want to explain the lacking protoss representation using the protoss winrate(s), then you'd have to also use the same reasoning to explain the zerg overrepresentation. Unfortunately this leads to a contradiction. Zerg players have the most balanced overall winrate of approximately 50% (or very slightly above). This would mean that, if the winrate dictated representation, then zerg players would be expected to make up approximately 33.3% of a tournament field on average. But this doesn't match the observation, as zerg is clearly very overrepresented, as much if not even more overrepresented than protoss is underrepresented. Furthermore, terran should be the most overrepresented race, but this doesn't match the observation either.
No, because I think it is more complicated than you make out. I think the representation is more in line with the winrates at practice partner level. Titles are more in line with the winrates at Flash level.
Terran being hard to win with at low levels might suppress Terran representation: Few candidate pros. Protoss being hard to win with at high levels might suppress Protoss representation: Many candidate pros, few actual pros. Zerg having a good vP matchup and bad vT matchup would end up overrepresented in such an environment.
Maybe I am wrong, I don't claim to have all the answers, but I think it's a better theory than "it was just random/unrelated."
Well I mean ok, the point I'm arguing is that it's difficult to figure out what actually causes the lack of titles, wins, representation, etc. and that pointing to imbalance results in many contradictions. And I do not agree that you can use a certain reasoning to explain protoss representation, but not use it to explain zerg representation. You have to be consistent.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
You're missing the point.
I provided the data that protoss IS in fact underrepresented. From that underrepresentation it can only be concluded that a protoss winning a title is less likely, and therefore it's a good explanation for a lack of protoss titles.
Less Protoss representation -> Less Protoss wins. This is sound, yes. Less Protoss wins -> Less Protoss representation. But this is also sound. And it leads to: [...] Less Protoss wins -> Less Protoss representation -> Less Protoss wins -> Less Protoss representation [...] This is a positive feedback loop. You can't just walk into it at an arbitrary point and call it the start and say it explains things.
Whatever causes the underrepresentation of protoss players is another story. I have ideas about that, and my hypothesis is that it's mainly due to a zerg overrepresentation from a historic preference for that race. That part of my argumentation is speculative and you can choose to ignore it, but you can't ignore the observation that zerg is in fact overrepresented in the tournaments.
The fact that you call this 'another story' and don't realise it is the whole story is the problem. Your headcannon that Protoss representation was repressed because of non-gameplay bias toward playing Zerg is extremely questionable and the fact that you imply this affected Protoss more than Terran already implies a race balance issue anyway. You don't see heroes with terrible winrates flood the pickrates at TI. You don't see shit guns flood the buy rates at a CSGO major. You don't see low tier characters flood top8 at Evo. There's no good reason to believe that Broodwar is somehow different.
If you want to explain the lacking protoss representation using the protoss winrate(s), then you'd have to also use the same reasoning to explain the zerg overrepresentation. Unfortunately this leads to a contradiction. Zerg players have the most balanced overall winrate of approximately 50% (or very slightly above). This would mean that, if the winrate dictated representation, then zerg players would be expected to make up approximately 33.3% of a tournament field on average. But this doesn't match the observation, as zerg is clearly very overrepresented, as much if not even more overrepresented than protoss is underrepresented. Furthermore, terran should be the most overrepresented race, but this doesn't match the observation either.
No, because I think it is more complicated than you make out. I think the representation is more in line with the winrates at practice partner level. Titles are more in line with the winrates at Flash level.
Terran being hard to win with at low levels might suppress Terran representation: Few candidate pros. Protoss being hard to win with at high levels might suppress Protoss representation: Many candidate pros, few actual pros. Zerg having a good vP matchup and bad vT matchup would end up overrepresented in such an environment.
Maybe I am wrong, I don't claim to have all the answers, but I think it's a better theory than "it was just random/unrelated."
Well I mean ok, the point I'm arguing is that it's difficult to figure out what actually causes the lack of titles, wins, representation, etc. and that pointing to imbalance results in many contradictions. And I do not agree that you can use a certain reasoning to explain protoss representation, but not use it to explain zerg representation. You have to be consistent.
Says the person claiming Protoss titles are caused by representation while also showing Zerg titles don't align with representation. I don't think what I said is inherently inconsistent, but your stance on representation:titles absolutely is.
From this he claimed because Protoss is under-represented, they won the least titles. However in that same data, Zerg's over-representation didn't lead to them winning the most titles.
On July 16 2022 04:17 Magic Powers wrote: Alright, I've added the remaining names from the list, and now the stats went from this: Terran: 44 (31.4%) Zerg: 58 (41.4%) Protoss: 38 (27.1%)
I'll also note that I recognize almost 100% of the players' usernames from the Kespa era. This is not a list including post-Kespa girl streamers or anything of that sort. (almost) every single one of these players is a legit competitor.
The conclusion remains the same. Protoss is underrepresented, which explains why a severe underrepresentation of protoss players has been observed in the individual leagues, which would obviously lead to fewer titles won. Especially when noting the observed Kespa era ZvP winrate of ~54%, it makes even more sense that the smaller pool of protoss players would end up winning a lot fewer titles.
I don't see how my argument lacks validity in any reasonable sense. You can choose to keep believing in ZvP imbalance, but the data is undeniably biased against protoss.
You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
You're missing the point.
I provided the data that protoss IS in fact underrepresented. From that underrepresentation it can only be concluded that a protoss winning a title is less likely, and therefore it's a good explanation for a lack of protoss titles.
Less Protoss representation -> Less Protoss wins. This is sound, yes. Less Protoss wins -> Less Protoss representation. But this is also sound. And it leads to: [...] Less Protoss wins -> Less Protoss representation -> Less Protoss wins -> Less Protoss representation [...] This is a positive feedback loop. You can't just walk into it at an arbitrary point and call it the start and say it explains things.
Whatever causes the underrepresentation of protoss players is another story. I have ideas about that, and my hypothesis is that it's mainly due to a zerg overrepresentation from a historic preference for that race. That part of my argumentation is speculative and you can choose to ignore it, but you can't ignore the observation that zerg is in fact overrepresented in the tournaments.
The fact that you call this 'another story' and don't realise it is the whole story is the problem. Your headcannon that Protoss representation was repressed because of non-gameplay bias toward playing Zerg is extremely questionable and the fact that you imply this affected Protoss more than Terran already implies a race balance issue anyway. You don't see heroes with terrible winrates flood the pickrates at TI. You don't see shit guns flood the buy rates at a CSGO major. You don't see low tier characters flood top8 at Evo. There's no good reason to believe that Broodwar is somehow different.
If you want to explain the lacking protoss representation using the protoss winrate(s), then you'd have to also use the same reasoning to explain the zerg overrepresentation. Unfortunately this leads to a contradiction. Zerg players have the most balanced overall winrate of approximately 50% (or very slightly above). This would mean that, if the winrate dictated representation, then zerg players would be expected to make up approximately 33.3% of a tournament field on average. But this doesn't match the observation, as zerg is clearly very overrepresented, as much if not even more overrepresented than protoss is underrepresented. Furthermore, terran should be the most overrepresented race, but this doesn't match the observation either.
No, because I think it is more complicated than you make out. I think the representation is more in line with the winrates at practice partner level. Titles are more in line with the winrates at Flash level.
Terran being hard to win with at low levels might suppress Terran representation: Few candidate pros. Protoss being hard to win with at high levels might suppress Protoss representation: Many candidate pros, few actual pros. Zerg having a good vP matchup and bad vT matchup would end up overrepresented in such an environment.
Maybe I am wrong, I don't claim to have all the answers, but I think it's a better theory than "it was just random/unrelated."
Well I mean ok, the point I'm arguing is that it's difficult to figure out what actually causes the lack of titles, wins, representation, etc. and that pointing to imbalance results in many contradictions. And I do not agree that you can use a certain reasoning to explain protoss representation, but not use it to explain zerg representation. You have to be consistent.
Says the person claiming Protoss titles are caused by representation while also showing Zerg titles don't align with representation. I don't think what I said is inherently inconsistent, but your stance on representation:titles absolutely is.
Huh? I'm offering a valid explanation, not claiming it to be definitive. I'm only saying it's a better explanation than "protoss is too weak". Lacking representation appears to explain the lack of titles better.
From this he claimed because Protoss is under-represented, they won the least titles. However in that same data, Zerg's over-representation didn't lead to them winning the most titles.
Now we're getting somewhere. Yes, zerg overrepresentation demonstrably didn't lead to them winning the most titles, that's exactly why winrates don't sufficiently explain the number of titles, because the title count defies expectations. You can't explain both observations (zerg and protoss lack of titles) with the same reasoning, and neither can I. I'm not claiming to have the correct explanation, I'm disputing the claim that matchup balance explains the lack of protoss titles, and offering alternative explanations that are at least equally valid. I'm arguing that pointing to titles for evidence of imbalance is generally flawed.
On July 16 2022 18:24 Lachrymose wrote: [quote] You can't just assume the relationship between Protoss professional population and Protoss titles is causative in the direction you want. You can't just assume the causative relationship isn't opposite (there's less Protoss population because Protoss is a less-winning race).
You can't just assume that less people chose to play Protoss randomly without any weighting. If you roll a three sided die and roll Green significantly less than you roll Blue or Red then the most likely explanation is something is wrong with your die.
If race choice were random it is much, much more likely that the populations of players are more equal at the casual and pre-amateur stages and some other pressure is causing Zerg and Terran players to be more likely to transition to professional and title winner.
In all likelihood we can go even further than this. Literally every single person agrees that Protoss is the easiest race and most successful at very low skill levels and so it makes sense that Protoss actually has the most candidate players at the pre-amateur level and yet they will go on to have the least pros and the least title winners.
In my opinion the simple and uncontroversial explanation that Protoss is easier but ultimately weaker is more likely than your explanation that randomly less people picked the green race.
I don't make the data. Zerg is clearly overrepresented in the pro scene, and that is despite zerg not having a greater overall winrate than terran. Protoss is underrepresented in the pro scene. Please go ahead and point out errors in the data I've collected, e.g. you could count the number of Korean players in the list that aren't progamers from the Kespa era and thus allow us to figure out a corrected distribution of the three races. So I'm not concluding from fewer protoss titles that protoss is underrepresented. I've straight up demonstrated that protoss is underrepresented. If you like I can post a random sample of individual tournaments so you can see how strongly biased against protoss the racial representation actually is.
An effect of racial overrepresentation in tournaments is that it increases the odds of a race to enter the finals and win the tournament. The math gets complicated with big numbers, but it's always the same effect.
The issue isn't the data. The issue is you assuming causation without proving it and then assuming the direction of the causation without proving it.
You collected data points A and B. You explicitly claimed A causes B without evidence or argument. You implicitly rejected B may instead be causing A without evidence or argument.
Your response to this being highlighted is only to reassert that the data points A and B exist and are valid.
You're missing the point.
I provided the data that protoss IS in fact underrepresented. From that underrepresentation it can only be concluded that a protoss winning a title is less likely, and therefore it's a good explanation for a lack of protoss titles.
Less Protoss representation -> Less Protoss wins. This is sound, yes. Less Protoss wins -> Less Protoss representation. But this is also sound. And it leads to: [...] Less Protoss wins -> Less Protoss representation -> Less Protoss wins -> Less Protoss representation [...] This is a positive feedback loop. You can't just walk into it at an arbitrary point and call it the start and say it explains things.
Whatever causes the underrepresentation of protoss players is another story. I have ideas about that, and my hypothesis is that it's mainly due to a zerg overrepresentation from a historic preference for that race. That part of my argumentation is speculative and you can choose to ignore it, but you can't ignore the observation that zerg is in fact overrepresented in the tournaments.
The fact that you call this 'another story' and don't realise it is the whole story is the problem. Your headcannon that Protoss representation was repressed because of non-gameplay bias toward playing Zerg is extremely questionable and the fact that you imply this affected Protoss more than Terran already implies a race balance issue anyway. You don't see heroes with terrible winrates flood the pickrates at TI. You don't see shit guns flood the buy rates at a CSGO major. You don't see low tier characters flood top8 at Evo. There's no good reason to believe that Broodwar is somehow different.
If you want to explain the lacking protoss representation using the protoss winrate(s), then you'd have to also use the same reasoning to explain the zerg overrepresentation. Unfortunately this leads to a contradiction. Zerg players have the most balanced overall winrate of approximately 50% (or very slightly above). This would mean that, if the winrate dictated representation, then zerg players would be expected to make up approximately 33.3% of a tournament field on average. But this doesn't match the observation, as zerg is clearly very overrepresented, as much if not even more overrepresented than protoss is underrepresented. Furthermore, terran should be the most overrepresented race, but this doesn't match the observation either.
No, because I think it is more complicated than you make out. I think the representation is more in line with the winrates at practice partner level. Titles are more in line with the winrates at Flash level.
Terran being hard to win with at low levels might suppress Terran representation: Few candidate pros. Protoss being hard to win with at high levels might suppress Protoss representation: Many candidate pros, few actual pros. Zerg having a good vP matchup and bad vT matchup would end up overrepresented in such an environment.
Maybe I am wrong, I don't claim to have all the answers, but I think it's a better theory than "it was just random/unrelated."
Well I mean ok, the point I'm arguing is that it's difficult to figure out what actually causes the lack of titles, wins, representation, etc. and that pointing to imbalance results in many contradictions. And I do not agree that you can use a certain reasoning to explain protoss representation, but not use it to explain zerg representation. You have to be consistent.
Says the person claiming Protoss titles are caused by representation while also showing Zerg titles don't align with representation. I don't think what I said is inherently inconsistent, but your stance on representation:titles absolutely is.
Huh? I'm offering a valid explanation, not claiming it to be definitive. I'm only saying it's a better explanation than "protoss is too weak". Lacking pepresentation appears to explain the lack of titles better.
You can't just say that. It doesn't "appear" to explain it at all per your own data. It doesn't apply to two of the three races. You claiming you P and P alone is explained by pickrate makes literally no sense. You have no argument whatsoever why that would be the case or how it even makes sense.
As an aside, here's the the race breakdown for the top48 of the 2008 TSL. This was an open ladder qualification for foreigners held on Iccup. The best foreigners in the world.
P: 17 Z: 15 T: 12 R (Random or race picker): 4
My memory is that if you go even lower on ladder back in the day it only gets even more Protoss-y and less Terran-y. I would appreciate the input of people who laddered in that era.
I'm pretty confident in my memory that P was over represented at the sub-pro level and the lack of pros is not explained by an overall lack of P players.
From this he claimed because Protoss is under-represented, they won the least titles. However in that same data, Zerg's over-representation didn't lead to them winning the most titles.
Now we're getting somewhere. Yes, zerg overrepresentation demonstrably didn't lead to them winning the most titles, that's exactly why winrates don't sufficiently explain the number of titles, because the title count defies expectations. You can't explain both observations (zerg and protoss lack of titles) with the same reasoning, and neither can I. I'm not claiming to have the correct explanation, I'm disputing the claim that matchup balance explains the lack of protoss titles, and offering alternative explanations that are at least equally valid. I'm arguing that pointing to titles for evidence of imbalance is generally flawed.
Yeah I can: at the absolute highest level Terran is the best race.
I just explained both observations with the same reasoning and it wasn't even hard. Wow.
From this he claimed because Protoss is under-represented, they won the least titles. However in that same data, Zerg's over-representation didn't lead to them winning the most titles.
Now we're getting somewhere. Yes, zerg overrepresentation demonstrably didn't lead to them winning the most titles, that's exactly why winrates don't sufficiently explain the number of titles, because the title count defies expectations. You can't explain both observations (zerg and protoss lack of titles) with the same reasoning, and neither can I. I'm not claiming to have the correct explanation, I'm disputing the claim that matchup balance explains the lack of protoss titles, and offering alternative explanations that are at least equally valid. I'm arguing that pointing to titles for evidence of imbalance is generally flawed.
It's very obvious: the number of titles is determined by the players at the highest level (let's say top 16), not the amount of players at qualifiers and preliminaries, or in other words, not the 100 Britneys you're trying to argue that would increase the likeliness of Protoss winning more titles.
Your reasoning is also fundametally wrong in that you automatically assume that somehow at lower levels, players just don't pick Protoss as much as Terran and Zerg. Yep. No reason at all, so it just happens to be like that huh? Despite the literal fact that at casual levels (the ladder) Protoss represents 50% of the player base.
I'm suspecting at this point the two of us are talking past each other. I've focused on the Kespa era, you're including post-Kespa and other data. That doesn't get us anywhere.
On July 16 2022 13:31 Avi-Love wrote: In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
Most people regard the years leading up to and including 2012 the most competitive era, because that's when the trio of Flash/JD/Bisu was dominating the scene. That era is almost always mentioned first. Today's standard of games is quite significantly lower, albeit still at a very high level.
You don't know what you're talking about, please stop posting. Firstly, It's very evident that bw has evolved *a lot* since 2010, and while progamers admit that maybe the mechnical skill of 2010 was slightly higher, the knowledge and meta has improved leaps and bounds -- it's fairly obvious to any viewer above a certain mmr that the game is being played at a higher level now than it was at the time. Secondly, 2012 is a complete figment of your imagination, the 'competitiveness' of bw started declining the second progamers received access to the SC2 beta, which was in 2010. Unless you've deluded yourself into thinking progamers playing two games, and proleague having sc2 increased the skill level of bw...? Thirdly, who are "most" people? I'd love to see your data on this, everyone I know that plays this game understands that it has *naturally* evolved over time. I think your obsesssion with both Bisu and Jaedong is a clear sign that you're living in the past and your mental gymnastics are stuck in the mud. Jaedong is garbage and has been for years, and Bisu chokes so hard that it's fairly obvious he's not even the best protoss player in a competitive setting. Lastly, it is obvious that zvp is slightly zerg favoured. It is also fairly obvious that it has been for quite awhile -- the biggest change in balance over the last couple of years has been tvp swinging slightly towards being terran favoured.
So what does all of this mean? Does it mean bw needs a balance patch? Negative. Can it be fixed through maps? Absolutely. Is the goal of the ASL to have perfect map balance? Obviously not, they clearly value throwing in a couple of curveballs which is evident from historical data (Third world, Sparkle, Inner Coven etc). If you're a dumb foreigner losing x matchup at 1900 mmr on the ladder is it because of your race? No, you just suck. Stop having a shitty mentality and accept your shortcomings -- but it's a lot easier to whine and blame other factors than it is to look inward, hence this thread and the many replies in it, and why it's so pointless and should have been stopped straight away.
I'm not going to stop posting just because you ask me so nicely, tyvm. And I know very well what I'm talking about, because I've been around since the very early days of SC, I know the historical development of most of the strategies, discoveries and trends, I've followed the pro scene and a decent portion of the amateur scene, and I have a good understanding of statistical fallacies. The balance debate has raged for decades and it was never settled. It's never been proven beyond doubt that PvZ is actually imbalanced, it's only been shown that protoss players are losing more than zerg players. That by itself is not proof of imbalance, it's only an indicator that needs supporting evidence, of which none has ever been found.
2012 marks the end of the most competitive era of BW because of the migration of much of the pro scene to SC2. The latter is not up to debate, it's a fact.
Furthermore, this observation "the biggest change in balance over the last couple of years has been tvp swinging slightly towards being terran favoured." - i.e. terran having a positive record against protoss post-Kespa - is a perfect example of the observed winrate changing over time by several percentage points. Several percentage points. Ring a bell? 54% ZvP winrate? If this is a natural development, then it can just as well happen in the other matchups without any changes to the balance. Somehow your argument is that ZvP can only be imbalanced as demonstrated by the 54% ZvP winrate, even though provably another matchup statistic in TvP has demonstrably shifted over several years, which can just as much happen in ZvP as well (in either direction).
You're lazer focusing on the ZvP winrate to support your hypothesis of imbalance, but you're ignoring other data points when they point to some unknown variables being in play that contradict your hypothesis.
Edit: Furthermore, you're contradicting yourself. On the one hand you're rejecting the claim that the late Kespa era was the most competitive era, on the other hand you're outright admitting that JD's and Bisu's skills have strongly declined.
"Jaedong is garbage and has been for years, and Bisu chokes so hard that it's fairly obvious he's not even the best protoss player in a competitive setting."
Please point to one player post-Kespa who has a winrate of >70% in any of the matchups, other than Flash. I'll wait.
This is a clueless foreigner just ranting for days. It reminds me of redditors talking about corporations pay no tax. Anyway, all the pros such as flash, best, and mini have all said their game knowledge and meta are far superior to anything that they had back in 2009~2011 (peak bw mechanical skills). The only thing that this era is missing peak mechanical skills. Nowadays, we have a concept of how efficiently you can mine at start of the gaming and that alone changes the trajectory of the game. We never had that in the past. No one knew about drone boosting nor probe boosting. Constantly drone boosting basically achives when you can get your lair to pop up by almost 8 seconds. That's the difference between if you're going to get busted by 2 rax MnM or not. Stop talking clueless foreigner.
He's not clueless. He's just superficial in his knowledge, especially regarding the modern era.
Like when he argued the KCM results, citing Bisu's availability as the reason Protoss didn't do well in that tournament, without ever digging deep enough to see that (1) Bisu of the modern era is not Bisu of the Kespa era, and (2) Bisu's record in KCM was not great at all - Protoss was carried by Snow and Best in KCM.
Also Flash wasn't a regular participant in KCM, certainly not enough to swing the results of 20 seasons.
On July 16 2022 13:31 Avi-Love wrote: In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
Most people regard the years leading up to and including 2012 the most competitive era, because that's when the trio of Flash/JD/Bisu was dominating the scene. That era is almost always mentioned first. Today's standard of games is quite significantly lower, albeit still at a very high level.
Today's gameplay is on another level, competitiveness doesn't matter as much when the game isn't fully discovered yet. Execution and game knowledge has a huge impact on balance.
All sports evolve with time, BW is no different. With more time people become better at decision making, macro, micro, multitask, builds get optimized and the game completely changes. It's kind of like you're unlocking new levels, players are able to do things they weren't able to do before, that in turn gives them more options b.o-wise.
It's the reason why 2h muta basically killed 3h muta for example. When high lvl pros started to mineral boost every game ~3 years ago, all the pros were forced to learn mining optimizations so they wouldn't fall behind. It's the same reason why as more time goes on builds become more and more greedy (players game knowledge evolves, their micro becomes good enough to defend with the bare minimum while maximizing their eco). It's the same reason why PvT is moving away from arbiter to mass speed shuttle storm style. Basically I'm giving you examples of why past games are meaningless when you're talking about todays balance, because the game wasn't played at the highest level execution-wise and b.o-wise.
You can think of it like this, if we take our modern day ladder ranks (1500/C being the default), how would you rank all the different eras, skill-wise? The earliest era would obviously be the lowest rank and eras would rank up as more time went on (due to mechanics/execution and game knowledge improvements).
It's no different than game knowledge/execution among players today, a D/C/B/A/S rank player doesn't have enough game knowledge or execution to understand the intricacies of higher level play, which ends up clouding their perspective. This applies to everyone up until the highest lvl of play.
TLDR: you over-think/over-analyze things when the issue is much more simple, just look at the stats of the highest lvl players in any era (compare them against their own era, not against all eras)
From this he claimed because Protoss is under-represented, they won the least titles. However in that same data, Zerg's over-representation didn't lead to them winning the most titles.
Now we're getting somewhere. Yes, zerg overrepresentation demonstrably didn't lead to them winning the most titles, that's exactly why winrates don't sufficiently explain the number of titles, because the title count defies expectations. You can't explain both observations (zerg and protoss lack of titles) with the same reasoning, and neither can I. I'm not claiming to have the correct explanation, I'm disputing the claim that matchup balance explains the lack of protoss titles, and offering alternative explanations that are at least equally valid. I'm arguing that pointing to titles for evidence of imbalance is generally flawed.
Yeah I can: at the absolute highest level Terran is the best race.
I just explained both observations with the same reasoning and it wasn't even hard. Wow.
The reasoning still doesn't suffice. If you explain terran's disproportionate number of titles with it being the best race, while also explaining zerg overrepresentation with zerg being a better race than protoss, then we should observe a terran overrepresentation that is significantly greater than zerg. This is why the explanation of representation being caused by winrate doesn't add up.
On July 16 2022 13:31 Avi-Love wrote: In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
Most people regard the years leading up to and including 2012 the most competitive era, because that's when the trio of Flash/JD/Bisu was dominating the scene. That era is almost always mentioned first. Today's standard of games is quite significantly lower, albeit still at a very high level.
Today's gameplay is on another level, competitiveness doesn't matter as much when the game isn't fully discovered yet. Execution and game knowledge has a huge impact on balance.
All sports evolve with time, BW is no different. With more time people become better at macro, micro, builds get optimized and the game completely changes. It's kind of like you're unlocking new levels, players are able to do things they weren't able to do before.
It's the reason why 2h muta basically killed 3h muta for example. When high lvl pros started to mineral boost every game ~3 years ago, all the pros were forced to learn mining optimization so they wouldn't fall behind. It's the same reason why as more time goes on builds end up becoming more and more greedy (players game knowledge evolves, their micro becomes good enough to defend pressure etc.). It's the same reason why PvT is moving away from arbiter to mass speed shuttle storm style. Basically I'm giving you examples of why past games are meaningless when you're talking about balance, because the game wasn't played at the highest level execution-wise and b.o-wise.
You can think of it like this, if we take our modern day ladder ranks (1500/C behind the default), how would you rank all the different eras, skill-wise? The earliest era would obviously be the lowest rank and eras would rank up as more time went on. It's no different than game knowledge among players today, a D/C/B/A rank player doesn't have enough game knowledge or execution to understand the intricacies of high level play, which ends up clouding their perspective.
TLDR: you over-think/over-analyze things when the issue is much more simple, just look at the stats of the highest lvl players in any era (compare them against their own era, not against all eras)
You're making valid points, but you're also presenting them in favor of today's competition.
I can argue that more difficult execution opens the door to new possibilities, and thus to different strategies altogether. Muta micro for example is highly execution-dependant. Post-Kespa a lot of the top zerg players failed to demonstrate the necessary skills to perform proper muta micro. Over the past few years some have improved it by a large margin, and top zerg players are now once again able to execute it much better.
During the Kespa era every single pro zerg was required to have strong muta micro, it was considered an essential skill. Things like that push the overall competitiveness of the field up.
Furthermore, I have a hard time buying into the idea that competitiveness goes up even though incentive goes down. Entertainment value is far more important for a streamer's income than being good at the game. The pros have admitted to this.
From this he claimed because Protoss is under-represented, they won the least titles. However in that same data, Zerg's over-representation didn't lead to them winning the most titles.
Now we're getting somewhere. Yes, zerg overrepresentation demonstrably didn't lead to them winning the most titles, that's exactly why winrates don't sufficiently explain the number of titles, because the title count defies expectations. You can't explain both observations (zerg and protoss lack of titles) with the same reasoning, and neither can I. I'm not claiming to have the correct explanation, I'm disputing the claim that matchup balance explains the lack of protoss titles, and offering alternative explanations that are at least equally valid. I'm arguing that pointing to titles for evidence of imbalance is generally flawed.
Yeah I can: at the absolute highest level Terran is the best race.
I just explained both observations with the same reasoning and it wasn't even hard. Wow.
The reasoning still doesn't suffice. If you explain terran's disproportionate number of titles with it being the best race, while also explaining zerg overrepresentation with zerg being a better race than protoss, then we should observe a terran overrepresentation that is significantly greater than zerg. This is why the explanation of representation being caused by winrate doesn't add up.
I already addressed this. Terran is only the best race at the absolute highest level on normal maps. Protoss is obviously the best race at casual levels for example.
The racial balance when people choose their race, when they transition to pro and when they win titles is not the same. Trying to use one number to represent them all is not right.
Terran is underrepresented at all skill levels but especially at low skill levels where their winrate is abysmal. Relatively speaking their representation improves at high skill levels where shockingly their winrate is stronger. High skill Terrans don't just appear out of the aether so they're limited by their smaller pool of candidate players. Since both their matchups get better at higher skill levels their title chances are good.
Protoss is overrepresented at low skill levels and underrepresented at high skill levels. Turns out Protoss has the best winrate at low high levels and the worst winrate at high skill levels. What an incredible coincidence, right? Since both their matchups get worse at higher skill levels their titles chances are not good.
From this he claimed because Protoss is under-represented, they won the least titles. However in that same data, Zerg's over-representation didn't lead to them winning the most titles.
Now we're getting somewhere. Yes, zerg overrepresentation demonstrably didn't lead to them winning the most titles, that's exactly why winrates don't sufficiently explain the number of titles, because the title count defies expectations. You can't explain both observations (zerg and protoss lack of titles) with the same reasoning, and neither can I. I'm not claiming to have the correct explanation, I'm disputing the claim that matchup balance explains the lack of protoss titles, and offering alternative explanations that are at least equally valid. I'm arguing that pointing to titles for evidence of imbalance is generally flawed.
Yeah I can: at the absolute highest level Terran is the best race.
I just explained both observations with the same reasoning and it wasn't even hard. Wow.
The reasoning still doesn't suffice. If you explain terran's disproportionate number of titles with it being the best race, while also explaining zerg overrepresentation with zerg being a better race than protoss, then we should observe a terran overrepresentation that is significantly greater than zerg. This is why the explanation of representation being caused by winrate doesn't add up.
He said "at the absolute highest level Terran is the best race". Let's say top 8/16. Of course Flash also has a say in this.
The over-representation of Zerg is at pro-level in general. Let's say top 32/40. *this arbitrary estimation is not true now, after actual counting.
I'm not sure about below-pro levels because there's no data, but you also have over-representation of Protoss at casual levels (ladder).
Of course everything is relative, but it kinda agrees with this graphic which was already posted many times:
Nothing is contradictory. It's you who keep mixing everything up. This also shows you that under-representation of Protoss at pro-level has nothing to do with them winning the least titles.
On July 16 2022 13:31 Avi-Love wrote: In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
Most people regard the years leading up to and including 2012 the most competitive era, because that's when the trio of Flash/JD/Bisu was dominating the scene. That era is almost always mentioned first. Today's standard of games is quite significantly lower, albeit still at a very high level.
Today's gameplay is on another level, competitiveness doesn't matter as much when the game isn't fully discovered yet. Execution and game knowledge has a huge impact on balance.
All sports evolve with time, BW is no different. With more time people become better at macro, micro, builds get optimized and the game completely changes. It's kind of like you're unlocking new levels, players are able to do things they weren't able to do before.
It's the reason why 2h muta basically killed 3h muta for example. When high lvl pros started to mineral boost every game ~3 years ago, all the pros were forced to learn mining optimization so they wouldn't fall behind. It's the same reason why as more time goes on builds end up becoming more and more greedy (players game knowledge evolves, their micro becomes good enough to defend pressure etc.). It's the same reason why PvT is moving away from arbiter to mass speed shuttle storm style. Basically I'm giving you examples of why past games are meaningless when you're talking about balance, because the game wasn't played at the highest level execution-wise and b.o-wise.
You can think of it like this, if we take our modern day ladder ranks (1500/C behind the default), how would you rank all the different eras, skill-wise? The earliest era would obviously be the lowest rank and eras would rank up as more time went on. It's no different than game knowledge among players today, a D/C/B/A rank player doesn't have enough game knowledge or execution to understand the intricacies of high level play, which ends up clouding their perspective.
TLDR: you over-think/over-analyze things when the issue is much more simple, just look at the stats of the highest lvl players in any era (compare them against their own era, not against all eras)
You're making valid points, but you're also presenting them in favor of today's competition.
I can argue that more difficult execution opens the door to new possibilities, and thus to different strategies altogether. Muta micro for example is highly execution-dependant. Post-Kespa a lot of the top zerg players failed to demonstrate the necessary skills to perform proper muta micro. Over the past few years some have improved it by a large margin, and top zerg players are now once again able to execute it much better.
During the Kespa era every single pro zerg was required to have strong muta micro, it was considered an essential skill. Things like that push the overall competitiveness of the field up.
Furthermore, I have a hard time buying into the idea that competitiveness goes up even though incentive goes down. Entertainment value is far more important for a streamer's income than being good at the game. The pros have admitted to this.
Muta micro isn't even close to a difficult execution nowadays. What's difficult is the intricate micro that is required ine arly part of the game, especially in zvp. No one wow's at muta micro. People wow at mini's 4 zlots killing 14 lings. Also, when did pros admit to entertainment value > skills? Higher skills literally raises viewership and donation money. Proleague players. Soma, sk, mini are constantly evolving and trying their absolute best to gain the slightest edge to beat their opponents). They live and die by their skills and the have admitted to this. If you're just a casual viewer who has a strong interest in the scene you wouldn't know of the intricacies that are involved in the pro scene. Flash has literally said he can beat his prime self 10/10 no problem. What TT1 said is 100% correct.
I think you guys are not getting the logical contradiction. Adding "at the highest level" doesn't suddenly make the contradiction disappear. Observation: Zerg is overrepresented (in tournaments and overall in the pro scene, at the highest level). Terran is a lot more appropriately represented (in tournaments and overall yada yada highest level). Protoss is underrepresented (in tournaments yada yada highest level).
Note that I'm talking about the Kespa era.
Possible causes: 1) zerg winrate (at the highest level) 2) zerg popularity (at the highest level) 3) ? (at the highest level)
If 1) then why is zerg overrepresented, but not terran? The terran winrate (at the highest level) is higher, therefore the winrate doesn't explain the zerg overrepresentation, since it's not leading to a terran overrepresentation.
Edit: I'm also not sure if it's clear how I'm using the term "representation" in this debate. Another term would be "racial distribution". It counts how many players who are main terran, zerg or protoss are represented in a given player pool or field. In the Kespa era progaming scene, the racial distribution of zerg is severely above expectation, terran is slightly below expectation, and protoss is severely below expectation. And expectation would be 1/3 (since virtually no one in the pro scene played random). This has nothing to do with wins, titles, winrate, or anything of that sort. It's the distribution of a race within a given population, in this case among all the progamers. That distribution shows that zerg is overrepresented both among the total progamer population and among participants of individual progamer tournaments.
On July 17 2022 06:14 Magic Powers wrote: I think you guys are not getting the logical contradiction.
Dude, your reasoning is wrong on so many levels it's absurd. You used the wrong data, a baseless assumption, and a flawed logic to come to your conclusion. It's all over the place. Lachrymose already spent a page explaining the flaws in your logic (the positive feedback loop) and other things already. But I'll point out a few more here:
In the Kespa era progaming scene, the racial distribution of zerg is severely above expectation, terran is slightly below expectation, and protoss is severely below expectation. And expectation would be 1/3 (since virtually no one in the pro scene played random).
See, this is a baseless assumption. You just automatically assumed expectation would be 1/3. Why? In fact it depends on a number of factors, including win rate at sub-pro levels (before the players were drafted into teamhouses), or historical factor (like, players tend to pick a race more because of their idols or something). To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data.
Observation: Zerg is overrepresented (in tournaments and overall in the pro scene, at the highest level). Terran is a lot more appropriately represented (in tournaments and overall yada yada highest level). Protoss is underrepresented (in tournaments yada yada highest level).
I told you before that the piece of data you used is just a list of players who have a liquipedia entry. It's not accurate. Here, I took the time to collect a more accurate one on for you: P: 235 (28.5%) T: 314 (38.1%) Z: 275 (33.4%) This is the number of participants of each race in the last 13 OSLs + last 13 MSLs combined, dated back from 2007. Why 2007? Partly because I don't have the time to go further. But also because it's when the Bisu's revolution happened which balanced up the PvZ matchup which was considered hugely Z favored before. So, P is indeed under-represented in tournaments, but Zerg is no longer over-represented in tournaments. It's Terran. So it kinda answer your question here:
If 1) then why is zerg overrepresented, but not terran? The terran winrate (at the highest level) is higher, therefore the winrate doesn't explain the zerg overrepresentation, since it's not leading to a terran overrepresentation.
But then, even the quote above has flawed in its logic as well. You see, the win rate we have is the win rate in officical tournaments. It's an event that occured after the event of race distribution in tournaments (hope you get what I mean). The win rate that can possibly affect racial distribution in tournaments is the win rate at sub-pro levels, in teamhouse practices, in qualifiers. Of that probably no one has the data as well.
So you see, with the mess of data and flawed logic you're stuck in, there's no way your claim is a valid explanation like you want to say.
On July 16 2022 13:31 Avi-Love wrote: In this thread: People looking to balance whine to make up for their own short comings, people incapable of analyzing data, someone that thinks the kespa era is still the pinnacle of bw (and still relevant) and lastly only a couple of genuinely unbiased and reasonable people.
All around terrible thread, the original post should have been locked instantly, the quality and 'discussion' it hoped to spark were doomed from the get go.
Most people regard the years leading up to and including 2012 the most competitive era, because that's when the trio of Flash/JD/Bisu was dominating the scene. That era is almost always mentioned first. Today's standard of games is quite significantly lower, albeit still at a very high level.
Today's gameplay is on another level, competitiveness doesn't matter as much when the game isn't fully discovered yet. Execution and game knowledge has a huge impact on balance.
All sports evolve with time, BW is no different. With more time people become better at decision making, macro, micro, multitask, builds get optimized and the game completely changes. It's kind of like you're unlocking new levels, players are able to do things they weren't able to do before, that in turn gives them more options b.o-wise.
It's the reason why 2h muta basically killed 3h muta for example. When high lvl pros started to mineral boost every game ~3 years ago, all the pros were forced to learn mining optimizations so they wouldn't fall behind. It's the same reason why as more time goes on builds become more and more greedy (players game knowledge evolves, their micro becomes good enough to defend with the bare minimum while maximizing their eco). It's the same reason why PvT is moving away from arbiter to mass speed shuttle storm style. Basically I'm giving you examples of why past games are meaningless when you're talking about todays balance, because the game wasn't played at the highest level execution-wise and b.o-wise.
You can think of it like this, if we take our modern day ladder ranks (1500/C being the default), how would you rank all the different eras, skill-wise? The earliest era would obviously be the lowest rank and eras would rank up as more time went on (due to mechanics/execution and game knowledge improvements).
It's no different than game knowledge/execution among players today, a D/C/B/A/S rank player doesn't have enough game knowledge or execution to understand the intricacies of higher level play, which ends up clouding their perspective. This applies to everyone up until the highest lvl of play.
TLDR: you over-think/over-analyze things when the issue is much more simple, just look at the stats of the highest lvl players in any era (compare them against their own era, not against all eras)
On July 17 2022 15:58 Lachrymose wrote: At this point I'm just going to accept you can't read and stop trying to talk to you.
Ad hominem is not an argument. I can read perfectly well, and your argument doesn't make sense and relies on too much speculation.
To confirm your hypothesis you could simply provide the racial distribution on the fish server. It would then be possible to figure out the survival rate of each race into the pro scene. I don't have access to that data, but you're very convinced of your argument so you should be able to get it somewhere.
On July 17 2022 06:14 Magic Powers wrote: I think you guys are not getting the logical contradiction.
Dude, your reasoning is wrong on so many levels it's absurd. You used the wrong data, a baseless assumption, and a flawed logic to come to your conclusion. It's all over the place. Lachrymose already spent a page explaining the flaws in your logic (the positive feedback loop) and other things already. But I'll point out a few more here:
In the Kespa era progaming scene, the racial distribution of zerg is severely above expectation, terran is slightly below expectation, and protoss is severely below expectation. And expectation would be 1/3 (since virtually no one in the pro scene played random).
See, this is a baseless assumption. You just automatically assumed expectation would be 1/3. Why? In fact it depends on a number of factors, including win rate at sub-pro levels (before the players were drafted into teamhouses), or historical factor (like, players tend to pick a race more because of their idols or something). To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data.
Observation: Zerg is overrepresented (in tournaments and overall in the pro scene, at the highest level). Terran is a lot more appropriately represented (in tournaments and overall yada yada highest level). Protoss is underrepresented (in tournaments yada yada highest level).
I told you before that the piece of data you used is just a list of players who have a liquipedia entry. It's not accurate. Here, I took the time to collect a more accurate one on for you: P: 235 (28.5%) T: 314 (38.1%) Z: 275 (33.4%) This is the number of participants of each race in the last 13 OSLs + last 13 MSLs combined, dated back from 2007. Why 2007? Partly because I don't have the time to go further. But also because it's when the Bisu's revolution happened which balanced up the PvZ matchup which was considered hugely Z favored before. So, P is indeed under-represented in tournaments, but Zerg is no longer over-represented in tournaments. It's Terran. So it kinda answer your question here:
If 1) then why is zerg overrepresented, but not terran? The terran winrate (at the highest level) is higher, therefore the winrate doesn't explain the zerg overrepresentation, since it's not leading to a terran overrepresentation.
But then, even the quote above has flawed in its logic as well. You see, the win rate we have is the win rate in officical tournaments. It's an event that occured after the event of race distribution in tournaments (hope you get what I mean). The win rate that can possibly affect racial distribution in tournaments is the win rate at sub-pro levels, in teamhouse practices, in qualifiers. Of that probably no one has the data as well.
So you see, with the mess of data and flawed logic you're stuck in, there's no way your claim is a valid explanation like you want to say.
1/3 is the expectation given all else is equal. The expectation can be adjusted according to data that supports a deviation. I didn't assume that 1/3 has to be true, I only used it as the initial benchmark.
"To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data." If no one has access to that data, then no one - not you or I or anyone else - can argue that they can tell the real reason for the racial distribution in the individual tournaments, because such information is essential to determine the cause. But you're claiming that things like learning difficulty and winrates determine racial distribution, so you have to be able to fully support that. If you don't provide the necessary data to prove your claim, then it can be easily dismissed. So far I haven't seen you provide any data at all, only speculation. (Edit: s.b.)
Edit: Ok I see you have actually provided data this time. You say this is from OSL and MSL since 2007.
P: 235 (28.5%) T: 314 (38.1%) Z: 275 (33.4%)
The first question that I would ask is how does a distribution go from being heavily zerg favored in the TLPD entries (most being legit progamers from the Kespa era) to being heavily terran favored, while there is no change to the field of protoss players? The winrate alone wouldn't explain that, because zerg has a very balanced overall winrate and should therefore not lose too many players, and protoss has a slightly negative overall winrate and should therefore lose players.
On July 17 2022 06:14 Magic Powers wrote: I think you guys are not getting the logical contradiction.
Dude, your reasoning is wrong on so many levels it's absurd. You used the wrong data, a baseless assumption, and a flawed logic to come to your conclusion. It's all over the place. Lachrymose already spent a page explaining the flaws in your logic (the positive feedback loop) and other things already. But I'll point out a few more here:
In the Kespa era progaming scene, the racial distribution of zerg is severely above expectation, terran is slightly below expectation, and protoss is severely below expectation. And expectation would be 1/3 (since virtually no one in the pro scene played random).
See, this is a baseless assumption. You just automatically assumed expectation would be 1/3. Why? In fact it depends on a number of factors, including win rate at sub-pro levels (before the players were drafted into teamhouses), or historical factor (like, players tend to pick a race more because of their idols or something). To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data.
Observation: Zerg is overrepresented (in tournaments and overall in the pro scene, at the highest level). Terran is a lot more appropriately represented (in tournaments and overall yada yada highest level). Protoss is underrepresented (in tournaments yada yada highest level).
I told you before that the piece of data you used is just a list of players who have a liquipedia entry. It's not accurate. Here, I took the time to collect a more accurate one on for you: P: 235 (28.5%) T: 314 (38.1%) Z: 275 (33.4%) This is the number of participants of each race in the last 13 OSLs + last 13 MSLs combined, dated back from 2007. Why 2007? Partly because I don't have the time to go further. But also because it's when the Bisu's revolution happened which balanced up the PvZ matchup which was considered hugely Z favored before. So, P is indeed under-represented in tournaments, but Zerg is no longer over-represented in tournaments. It's Terran. So it kinda answer your question here:
If 1) then why is zerg overrepresented, but not terran? The terran winrate (at the highest level) is higher, therefore the winrate doesn't explain the zerg overrepresentation, since it's not leading to a terran overrepresentation.
But then, even the quote above has flawed in its logic as well. You see, the win rate we have is the win rate in officical tournaments. It's an event that occured after the event of race distribution in tournaments (hope you get what I mean). The win rate that can possibly affect racial distribution in tournaments is the win rate at sub-pro levels, in teamhouse practices, in qualifiers. Of that probably no one has the data as well.
So you see, with the mess of data and flawed logic you're stuck in, there's no way your claim is a valid explanation like you want to say.
1/3 is the expectation given all else is equal. The expectation can be adjusted according to data that supports a deviation. I didn't assume that 1/3 has to be true, I only used it as the initial benchmark.
"To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data." If no one has access to that data, then no one - not you or I or anyone else - can argue that they can tell the real reason for the racial distribution in the individual tournaments, because such information is essential to determine the cause. But you're claiming that things like learning difficulty and winrates determine racial distribution, so you have to be able to fully support that. If you don't provide the necessary data to prove your claim, then it can be easily dismissed. So far I haven't seen you provide any data at all, only speculation.
It's you who brought racial distribution to the table and make a claim about that. No one else. We're only dismissing your claim. Roll back the pages and see.
The first question that I would ask is how does a distribution go from being heavily zerg favored in the TLPD entries (most being legit progamers from the Kespa era) to being heavily terran favored, while there is no change to the field of protoss players? The winrate alone wouldn't explain that, because zerg has a very balanced overall winrate and should therefore not lose too many players, and protoss has a slightly negative overall winrate and should therefore lose players
Did you, again, not read my last point? The win rate you have is the win rate in tournaments. It's the event occured after the racial distribution. It only explains results in tournaments, not how players get into tournaments.
On July 17 2022 06:14 Magic Powers wrote: I think you guys are not getting the logical contradiction.
Dude, your reasoning is wrong on so many levels it's absurd. You used the wrong data, a baseless assumption, and a flawed logic to come to your conclusion. It's all over the place. Lachrymose already spent a page explaining the flaws in your logic (the positive feedback loop) and other things already. But I'll point out a few more here:
In the Kespa era progaming scene, the racial distribution of zerg is severely above expectation, terran is slightly below expectation, and protoss is severely below expectation. And expectation would be 1/3 (since virtually no one in the pro scene played random).
See, this is a baseless assumption. You just automatically assumed expectation would be 1/3. Why? In fact it depends on a number of factors, including win rate at sub-pro levels (before the players were drafted into teamhouses), or historical factor (like, players tend to pick a race more because of their idols or something). To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data.
Observation: Zerg is overrepresented (in tournaments and overall in the pro scene, at the highest level). Terran is a lot more appropriately represented (in tournaments and overall yada yada highest level). Protoss is underrepresented (in tournaments yada yada highest level).
I told you before that the piece of data you used is just a list of players who have a liquipedia entry. It's not accurate. Here, I took the time to collect a more accurate one on for you: P: 235 (28.5%) T: 314 (38.1%) Z: 275 (33.4%) This is the number of participants of each race in the last 13 OSLs + last 13 MSLs combined, dated back from 2007. Why 2007? Partly because I don't have the time to go further. But also because it's when the Bisu's revolution happened which balanced up the PvZ matchup which was considered hugely Z favored before. So, P is indeed under-represented in tournaments, but Zerg is no longer over-represented in tournaments. It's Terran. So it kinda answer your question here:
If 1) then why is zerg overrepresented, but not terran? The terran winrate (at the highest level) is higher, therefore the winrate doesn't explain the zerg overrepresentation, since it's not leading to a terran overrepresentation.
But then, even the quote above has flawed in its logic as well. You see, the win rate we have is the win rate in officical tournaments. It's an event that occured after the event of race distribution in tournaments (hope you get what I mean). The win rate that can possibly affect racial distribution in tournaments is the win rate at sub-pro levels, in teamhouse practices, in qualifiers. Of that probably no one has the data as well.
So you see, with the mess of data and flawed logic you're stuck in, there's no way your claim is a valid explanation like you want to say.
1/3 is the expectation given all else is equal. The expectation can be adjusted according to data that supports a deviation. I didn't assume that 1/3 has to be true, I only used it as the initial benchmark.
"To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data." If no one has access to that data, then no one - not you or I or anyone else - can argue that they can tell the real reason for the racial distribution in the individual tournaments, because such information is essential to determine the cause. But you're claiming that things like learning difficulty and winrates determine racial distribution, so you have to be able to fully support that. If you don't provide the necessary data to prove your claim, then it can be easily dismissed. So far I haven't seen you provide any data at all, only speculation.
It's you who brought racial distribution to the table and make a claim about that. No one else. We're only dismissing your claim. Roll back the pages and see.
That's because if, using your hypothesis, you can't explain the racial distribution without contradictions, then your hypothesis is not valid. Racial distribution is an essential piece to the puzzle, because without it, you don't even have an argument. You're arguing that the racial distribution for the titles proves imbalance, therefore you can't just ignore racial distribution in other places as you please.
On July 17 2022 06:14 Magic Powers wrote: I think you guys are not getting the logical contradiction.
Dude, your reasoning is wrong on so many levels it's absurd. You used the wrong data, a baseless assumption, and a flawed logic to come to your conclusion. It's all over the place. Lachrymose already spent a page explaining the flaws in your logic (the positive feedback loop) and other things already. But I'll point out a few more here:
In the Kespa era progaming scene, the racial distribution of zerg is severely above expectation, terran is slightly below expectation, and protoss is severely below expectation. And expectation would be 1/3 (since virtually no one in the pro scene played random).
See, this is a baseless assumption. You just automatically assumed expectation would be 1/3. Why? In fact it depends on a number of factors, including win rate at sub-pro levels (before the players were drafted into teamhouses), or historical factor (like, players tend to pick a race more because of their idols or something). To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data.
Observation: Zerg is overrepresented (in tournaments and overall in the pro scene, at the highest level). Terran is a lot more appropriately represented (in tournaments and overall yada yada highest level). Protoss is underrepresented (in tournaments yada yada highest level).
I told you before that the piece of data you used is just a list of players who have a liquipedia entry. It's not accurate. Here, I took the time to collect a more accurate one on for you: P: 235 (28.5%) T: 314 (38.1%) Z: 275 (33.4%) This is the number of participants of each race in the last 13 OSLs + last 13 MSLs combined, dated back from 2007. Why 2007? Partly because I don't have the time to go further. But also because it's when the Bisu's revolution happened which balanced up the PvZ matchup which was considered hugely Z favored before. So, P is indeed under-represented in tournaments, but Zerg is no longer over-represented in tournaments. It's Terran. So it kinda answer your question here:
If 1) then why is zerg overrepresented, but not terran? The terran winrate (at the highest level) is higher, therefore the winrate doesn't explain the zerg overrepresentation, since it's not leading to a terran overrepresentation.
But then, even the quote above has flawed in its logic as well. You see, the win rate we have is the win rate in officical tournaments. It's an event that occured after the event of race distribution in tournaments (hope you get what I mean). The win rate that can possibly affect racial distribution in tournaments is the win rate at sub-pro levels, in teamhouse practices, in qualifiers. Of that probably no one has the data as well.
So you see, with the mess of data and flawed logic you're stuck in, there's no way your claim is a valid explanation like you want to say.
1/3 is the expectation given all else is equal. The expectation can be adjusted according to data that supports a deviation. I didn't assume that 1/3 has to be true, I only used it as the initial benchmark.
"To know the real racial distribution among the pro scene, you need to go deep into each teamhouse and count the number of players of each race. Of that probably no one has the data." If no one has access to that data, then no one - not you or I or anyone else - can argue that they can tell the real reason for the racial distribution in the individual tournaments, because such information is essential to determine the cause. But you're claiming that things like learning difficulty and winrates determine racial distribution, so you have to be able to fully support that. If you don't provide the necessary data to prove your claim, then it can be easily dismissed. So far I haven't seen you provide any data at all, only speculation.
It's you who brought racial distribution to the table and make a claim about that. No one else. We're only dismissing your claim. Roll back the pages and see.
The first question that I would ask is how does a distribution go from being heavily zerg favored in the TLPD entries (most being legit progamers from the Kespa era) to being heavily terran favored, while there is no change to the field of protoss players? The winrate alone wouldn't explain that, because zerg has a very balanced overall winrate and should therefore not lose too many players, and protoss has a slightly negative overall winrate and should therefore lose players
Did you, again, not read my last point? The win rate you have is the win rate in tournaments. It's the event occured after the racial distribution. It only explains results in tournaments, not how players get into tournaments.
You said this: "This is the number of participants of each race in the last 13 OSLs + last 13 MSLs combined, dated back from 2007. Why 2007? Partly because I don't have the time to go further. But also because it's when the Bisu's revolution happened which balanced up the PvZ matchup which was considered hugely Z favored before."
I do not think I missed your point, no. The racial distribution for the participants that you posted contradicts the racial distribution of the progaming landscape, since every progamer would equally try to qualify for the individual tournaments. Since there are far more zerg progamers, it's expected that, due to their overall very fair winrate, they should qualify at a rate that they're not significantly less represented among the tournament participants than among the progaming landscape. Protoss on the other hand would be expected to qualify at a lower rate. We do not observe either of these expectations. Zerg went from clear overrepresentation to a fairly normal representation (reduction of around 7%), and protoss stayed at roughly 28%. This doesn't make sense.
OMG why do you keep missing key things and mixing things up again and again.
The racial distribution for the participants that you posted contradicts the racial distribution of the progaming landscape
You don't have this one (bold part). The one piece of data you presented is inaccurate. It's just the number of players having a liquipedia entry. For the racial distribution of the progaming landscape, go to the history of each team and count. We probably can't.
it's expected that, due to their overall very fair winrate, they should qualify at a rate that they're not significantly less represented among the tournament participants than among the progaming landscape.
You don't have this one either. The "fair win rate" you have is the win rate from tournaments. The one from qualifiers and below that level might be different.
In fact, all of your "expectations" are falsed expectations, because your methodology is wrong, starting from the data you used.
The racial distribution for the participants that you posted contradicts the racial distribution of the progaming landscape
You don't have this one (bold part). The one piece of data you presented is inaccurate. It's just the number of players having a liquipedia entry. For the racial distribution of the progaming landscape, go to the history of each team and count. We probably can't.
it's expected that, due to their overall very fair winrate, they should qualify at a rate that they're not significantly less represented among the tournament participants than among the progaming landscape.
You don't have this one either. The "fair win rate" you have is the win rate from tournaments. The one from qualifiers and below that level might be different.
In fact, all of your "expectations" are falsed expectations, because your methodology is wrong, starting from the data you used.
I do have that data, but if you reject the data, ok fine, then you don't have an argument anymore, since you would then not know the racial distribution among the progaming landscape either.
On July 17 2022 18:02 TMNT wrote: OMG why do you keep missing key things and mixing things up again and again.
The racial distribution for the participants that you posted contradicts the racial distribution of the progaming landscape
You don't have this one (bold part). The one piece of data you presented is inaccurate. It's just the number of players having a liquipedia entry. For the racial distribution of the progaming landscape, go to the history of each team and count. We probably can't.
it's expected that, due to their overall very fair winrate, they should qualify at a rate that they're not significantly less represented among the tournament participants than among the progaming landscape.
You don't have this one either. The "fair win rate" you have is the win rate from tournaments. The one from qualifiers and below that level might be different.
In fact, all of your "expectations" are falsed expectations, because your methodology is wrong, starting from the data you used.
I do have that data, but if you reject the data, ok fine, then you don't have an argument anymore, since you would then not know the racial distribution among the progaming landscape either.
Lmao what is this level of discussion. I pointed out the data you used is wrong, explaining the reason why it's wrong. You proceed to say yours is right, without explaining why? Where's your complete list of progamers? Where's your qualifier winrate?
Fyi, I just quickly counted the race distribution of an OSL qualifier, as an example for you: 2010 Korean Air OSL Season 1/Results and Standings (Offlines) 57 P 60 T 50 Z So you see, Zerg is the most under-represented in this qualifier (surprise!).
You also have 4P, 4T and 8Z already seeded to the main tournament, so in total it's 61P (33%) 64T (35%) 58Z. (32%) registering for the tournaments. Fairly equal eh? In fact, Zerg is slightly under-represented. What is your data again?
Now let's see who made it out of the qualifier: 4 P (lmao) 10 T 10 Z So, P indeed qualied at an abysmal rate. suppoting the theory that they are the worst race.
Finally, this is the distribution of the main tournament: 8 P (20 %) 14 T (35%) 18 Z (45%)
Did you say before that if we flood the field with Protoss players, they'll get more chance of winning the tournament? Well, in fact, the majority of them couldn't even make it out of the qualifiers .
Ok, I can follow that reasoning. Didn't you say the winrates (or learning difficulty, I'm not sure what your argument is anymore) differ drastically between different ranks, explaining why more protoss players would be weeded out at or before the highest ranks?
On July 17 2022 19:01 Magic Powers wrote: Ok, I can follow that reasoning. Didn't you say the winrates (or learning difficulty, I'm not sure what your argument is anymore) differ drastically between different ranks, explaining why more protoss players would be weeded out at or before the highest ranks?
My main argument is "your argument is wrong". It works like this:
- You argued against it, citing Bisu PvZ win rate, racial distribution and etc., saying it's contradictory.
- I, along with some others, proceeded to point out that your way of reasoning is flawed. I particularly pointed out the data you used is not accurate.
In fact I have looked at another OSL qualifiers to see race distribution of the participants. It looks like there was always roughly 170-180 participants in qualifiers, and the distribution among them is roughly equal (2010 OSL S2 is 55/60/54). Obviously there were seasons when P qualified at a much better rate, and vice versa for T and Z. But given the distribution of participants in qualifiers is roughly equal, the inferior distribution of P in the main tournaments (after many tournaments) does suggest that they performed worst.
Edit: seems correct. 2009 Bacchus OSL has 57/63/59 participants in qualifiers (main tourney is 10/12/18). 2012 Tving OSL has 26/24/24 participants in qualifiers (main tourney is 5/8/15). And I'm not gonna count more. The qualifiants were always divided in groups of 8 with 2/3/3 distribution. Protoss was never under-represented at qualifiers. They just got wiped out more.
I'm not arguing that protoss should perform equally to the other races, I've even posted the ZvP winrate of roughly 54%, confirming that the observed winrate favors zerg. I'm only disputing the evidence for the claim of imbalance, in particular titles won, which I consider unusable as evidence, because there are many factors that can mess with the data.
A much better piece of evidence is the matchup winrate, and I have no interest in disputing that as valid evidence (although I wouldn't consider it completely perfect, but it's also not unusable).
On July 17 2022 19:37 Magic Powers wrote: I'm not arguing that protoss should perform equally to the other races, I've even posted the ZvP winrate of roughly 54%, confirming that the observed winrate favors zerg. I'm only disputing the evidence for the claim of imbalance, in particular titles won, which I consider unusable as evidence, because there are many factors that can mess with the data.
A much better piece of evidence is the matchup winrate, and I have no interest in disputing that as valid evidence (although I wouldn't consider it completely perfect, but it's also not unusable).
Yeah but you need to come up with a new methodology for that. Your old one (where you argued that lack of P leads to lack of titles), I just proved wrong. P players wasn't lacking in qualifiers, they just got wiped out more by T and Z.
For what do I need a new methodology? I'm not making a specific claim, I'm disputing evidence. My claim isn't that protoss should perform like the other races.
Some of the comments here very good... this is so cool
Anyway, I think this
On July 14 2022 00:52 Magic Powers wrote: PvZ is really the only matchup that was ever considered disadvantageous for protoss. But there are examples contradicting this view. Bisu achieved a PvZ winrate of 71.51% which is greater than that of Flash's 69.7% in TvP (the historically worst matchup for terran), and only slightly behind Flash's TvZ winrate of 72%, with both players having played a similar number of games. Bisu also had a 9-5 record against Jaedong (which is of course a small sample), while Flash achieved 20-20. This even though Jaedong had a better record against protoss with 67.38% compared to 63% vs terran. Furthermore, Bisu's best matchup has always clearly been PvZ, which puts a big question mark on the claim of racial imbalance.
These three players are typically considered the best of their respective race, and it's quite clear that their winrates contradict the idea that protoss players have it the worst, even in the allegedly worst matchup of PvZ.
I think that protoss players have never figured out what sets Bisu apart so much. He clearly has a far superior understanding of the matchup than anyone else. Why can't other protoss players study his game and apply their findings? I think that's the real question that needs to be asked. It's like the answer is right there in front of people, but they're not picking it up. We have hundreds of vods of Bisu playing PvZ that can be studied.
Instead people resort to complaining about balance, which is the lazy option.
In YGOSU, it is mostly agreed upon that when three or more players are playing on team melee, protoss becomes the strongest race due to the likelihood of strong play-making potential from units such as early game harassment from probes having a single player dedicate all his focus on it, corsairs that never die throughout the entire game versus the zergs, and game changing late-game spellcasters such as arbiters.
I was thinking the same about Protoss before, and then I saw a few team melee games between Bisu+Best vs Light+Rush, and man oh man did the Protoss team totally destroy Terran.
The thing is, the Terran army during a fight always gets some sort of value even if you set your units up badly. If Protoss units are in range, they fire. Same thing can't be said for Protoss: + Dragoons hitting Depots or a floating Barrack. + Zealots getting on top of each other, and worse, eating a mine together. + High Templars evaporating before casting any storms + Shuttles dying mid-air because frankly, by the time you have the time to grab them they're already dead. All of the above examples can be mitigated by a great player but only to some extent. Chances are, if you are able to finish sorting out the targeting of your ground army, your Shuttles are likely on auto pilot and you won't be able to cast all the storms you'd like. But if you have two or three players controlling the same battle. The extra values Protoss can get is huge.
Same thing can be said for PvZ. No more scouting Probe dying early. No more High Templar full of energy dying before storm can get off. No more Corsair wasting.
Obviously Terran and Zerg can benefit a lot from team melee as well. But I feel like the extra values are nowhere near Protoss'.
Bisu's greatest strenghts are his perfect execution and superior multitasking, those two go along pretty well and if you watch his FPVod you can tell thats how he gets the edge in PvZ.
Everybody knows that but that is not some style you can just copy . His PvZ is not a secret, it's just unreachable. With the keyboard and the mouse, he is just better than the rest.
That is also the reason why Mini is doing so well. I just hope that anyone who go and watch These two games and then read TMNT comment would see it as clearly as I do now ^^
EDIT: not only TMNT but also XenoSky and Letmelose, some very interesting ideas
On July 14 2022 00:52 Magic Powers wrote: PvZ is really the only matchup that was ever considered disadvantageous for protoss. But there are examples contradicting this view. Bisu achieved a PvZ winrate of 71.51% which is greater than that of Flash's 69.7% in TvP (the historically worst matchup for terran), and only slightly behind Flash's TvZ winrate of 72%, with both players having played a similar number of games. Bisu also had a 9-5 record against Jaedong (which is of course a small sample), while Flash achieved 20-20. This even though Jaedong had a better record against protoss with 67.38% compared to 63% vs terran. Furthermore, Bisu's best matchup has always clearly been PvZ, which puts a big question mark on the claim of racial imbalance.
These three players are typically considered the best of their respective race, and it's quite clear that their winrates contradict the idea that protoss players have it the worst, even in the allegedly worst matchup of PvZ.
I think that protoss players have never figured out what sets Bisu apart so much. He clearly has a far superior understanding of the matchup than anyone else. Why can't other protoss players study his game and apply their findings? I think that's the real question that needs to be asked. It's like the answer is right there in front of people, but they're not picking it up. We have hundreds of vods of Bisu playing PvZ that can be studied.
Instead people resort to complaining about balance, which is the lazy option.
This blog is such a gem. I am so in agreement with the following statement :
In YGOSU, it is mostly agreed upon that when three or more players are playing on team melee, protoss becomes the strongest race due to the likelihood of strong play-making potential from units such as early game harassment from probes having a single player dedicate all his focus on it, corsairs that never die throughout the entire game versus the zergs, and game changing late-game spellcasters such as arbiters.
I was thinking the same about Protoss before, and then I saw a few team melee games between Bisu+Best vs Light+Rush, and man oh man did the Protoss team totally destroy Terran.
The thing is, the Terran army during a fight always gets some sort of value even if you set your units up badly. If Protoss units are in range, they fire. Same thing can't be said for Protoss: + Dragoons hitting Depots or a floating Barrack. + Zealots getting on top of each other, and worse, eating a mine together. + High Templars evaporating before casting any storms + Shuttles dying mid-air because frankly, by the time you have the time to grab them they're already dead. All of the above examples can be mitigated by a great player but only to some extent. Chances are, if you are able to finish sorting out the targeting of your ground army, your Shuttles are likely on auto pilot and you won't be able to cast all the storms you'd like. But if you have two or three players controlling the same battle. The extra values Protoss can get is huge.
Same thing can be said for PvZ. No more scouting Probe dying early. No more High Templar full of energy dying before storm can get off. No more Corsair wasting.
Obviously Terran and Zerg can benefit a lot from team melee as well. But I feel like the extra values are nowhere near Protoss'.
Bisu's greatest strenghts are his perfect execution and superior multitasking, those two go along pretty well and if you watch his FPVod you can tell thats how he gets the edge in PvZ.
Everybody knows that but that is not some style you can just copy . His PvZ is not a secret, it's just unreachable. With the keyboard and the mouse, he is just better than the rest.
That is also the reason why Mini is doing so well. I just hope that anyone who go and watch These two games and then read TMNT comment would see it as clearly as I do now ^^
EDIT: not only TMNT but also XenoSky and Letmelose, some very interesting ideas
What’s overlooked, with regards to Bisu, is his mentors and teachers at different stages of his career. His success in Proleague was hugely influenced by Kingdom the same way Fantasy was influenced by Oov. Bisu was at his best when he was coached by brilliant minds on how to approach the map he was preparing for. Then his mechanical skills ensured what he prepared would demolish his opponents. I think a really interesting discussion that we haven’t really had is who the most influential coaches were in Brood War. Would X Player have a successful career without coach Y?
Late game, the ground Protoss army gets decimated by both T and Z. I'd say this on its own is good reason to consider Protoss the weakest race. It has to compensate strategically in early and mid game which the other races don't need to do.
Found it amusing how someone was fiercely arguing earlier in the thread that Bisu's monstrous PvZ win rate is proof that Protoss is balanced.
My immediate thought was: "Yeah, but that's only because Flash plays Terran lulz."
Followed by: "Zerg players so weak in that era lulz."
Point is that data spread over a longer period of time shows that the matchup is Zerg-favoured and Bisu had a short stint of dominance that hasn't really been replicated since.
Of course, there's the whole chicken-and-egg question of whether Zerg/Terran pros just happen to be more skilful than Protoss pros or that Zerg/Terran's edge is a key factor that draws top pros to play Zerg/Terran over Protoss.
I'm just a casual player whose main race is Terran and don't have first hand experience in high level PvZ. But my neutral view from watching games over the years is that Protoss just have a more torrid time dealing with hydra busts than Zergs dealing with zealots running around (just simplying the matchup, let's not even get into defilers...)
Terran requires best decision making, strict BO and perfect execution. Zerg require high APM, perfect micro. Protoss requires multitasking (?), patience and natural talent.
EPT
AfreecaTV StarCraft League Remastered![[image loading]](https://i.imgur.com/fv4SwyC.png)
![[image loading]](https://i.imgur.com/NadKlxT.png)
![[image loading]](https://i.imgur.com/s3q9ftO.png)
![[image loading]](https://i.imgur.com/phGlO68.png)
![[image loading]](https://i.imgur.com/SdTJBmR.png)
![[image loading]](https://i.imgur.com/uGAOWNP.png)
![[image loading]](https://i.imgur.com/wewkyas.png)
. 
