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.
EPT
![[image loading]](https://i.imgur.com/phGlO68.png)
KCM/Race Survival
![[image loading]](https://i.imgur.com/SdTJBmR.png)
![[image loading]](https://i.imgur.com/uGAOWNP.png)
![[image loading]](https://i.imgur.com/wewkyas.png)
