On June 21 2021 00:25 att wrote: We cant ignore that there have been quite many terran bonjwas, and some zerg bonjwas, but protoss bonjwa has been weaker and shorter reign compared to the others so seems like when terrans is on top of the world they really must be an exceptional player
That is why Bisu is the best lol Seriously tho, there is no protoss bonjwa. And as great as the bonjwas are (and we should all be fans), for purposes of discussing these topics, imo their reign should be disregarded/excluded. These guys basically changed the game and everyone’s understanding of it. Thats what made them such important figures (apart from winning a bunch of tournaments). But as such, if we want to discuss the game as it exists today, how it was before the bonjwas changed it is irrelevant.
Im not sure how I feel about creating asterisks' for certain players. No reputable sports analyst would remove Michael Jordan or Lebron James from a certain stat to prove a point. It just seems very arbitrary to me and a bit vunerable to bias. Those players would have to have identical metrics between them to justify removing them from a pool of statistics. They exist, they play by the same rules and have the same opportunities as the other players, therefore you absolutely have to include them.
I think you completely misuderstood. I am not arguing to exclude them from the list of the very best to play this game. I am just saying we should exclude old games that were played with obsolete styles/strategies. For obvious reasons.
I quoted the wrong message but was talking about the post giving OSL MSL KSL and ASL gold medals but excluding Flash Jaedong and Bisu. This is just compiling stats arbitrarily to mold a narrative. It assumes that those 3 players are equals when they are not. Any statastic that creates an asterisks for Flash is meaningless if you think about it. Why do we create an exception for him? Does he not play with the same rules and opprotunities as other players? His accomplishments should absolutely be included in any compilation of stats. Now if we had a large sample size of him play R or off race, THEN maybe we could leverage his games played differently to create some sort of meaningful race balance discussion. Until then, Flash wins = terran wins, period.
Ya I got it a little bit after I replied that you quoted the wrong post. Anyway, this started when I said that I firmly believe Terrans cant compete without Flash. Take from that statement what you will. I just posted some stuff to back that up.
Let me spin it to you a different way. If you were compiling stats for average IQ of a population, would you remove the top X samples from the data? I understand people want to asterisks Flash but the reasons seem so arbitrary. If Broodwar is played for another 100 years and 2 or 3 players obtain more golds than flash, will the asterisks all of a sudden be removed?
Sorry but I really have no idea what you are trying to get at. Against my better judgment, I’ll respond to the IQ question: outliers can be excluded. That is why median/mode are also used.
Maybe the IQ example was bad. Ill return to the original sports related analogy for this discussion. If a sports (NBA) analyst were to compile data for top 10 players in each of the respected positions, do you think they would purposely not include certain players because they were too good? I get removing outliers from certain data sets but the methodology should be well defined enough to apply to any player and should contain concrete benchmarks.
It also leads to a slippery slope of creating asterisks for anything that might be deemed "outlier" for example bad map pools during certain seasons, weaker player brackets etc. You see this a lot when people try to create asterisk seasons in the NBA when players were injured during certain seasons. The consensus is its best to just leave the data intact and let the numbers speak for themselves once they reach a certain sample size. Although I agree post-savior era kespa should be a starting point for statistics gathering to focus on what we call the "modern era".
I would say we always need to apply critical thinking when analyzing data. There is danger of bias for sure, but not looking at the details and taking numbers at face value can be just as dangerous. Let me ask you, if the numbers were something like this: 99 tournaments, 33 wins for each race. All T wins were from Flash. No P/Z player won more than 1 title. Do we still say “all Flash wins are T wins period”? Should we really be so closed minded as to not entertain the possibility that maybe its actually really hard to win with T, that its so hard to play that only one guy has managed to win with it after 99 tries? If Flash doesnt play on the 100th tournament, what unbiased rational person will pick T to win it? I hate exaggerating it this way but just doing it to make a point.
On June 21 2021 10:05 SlayerS_BunkiE wrote: I would say we always need to apply critical thinking when analyzing data. There is danger of bias for sure, but not looking at the details and taking numbers at face value can be just as dangerous. Let me ask you, if the numbers were something like this: 99 tournaments, 33 wins for each race. All T wins were from Flash. No P/Z player won more than 1 title. Do we still say “all Flash wins are T wins period”? Should we really be so closed minded as to not entertain the possibility that maybe its actually really hard to win with T, that its so hard to play that only one guy has managed to win with it after 99 tries? If Flash doesnt play on the 100th tournament, what unbiased rational person will pick T to win it? I hate exaggerating it this way but just doing it to make a point.
In this hypothetical example, I'd be more inclined to suspect an error in the dataset rather than draw the conclusion that terran is too difficult to play except for Flash. In my job I frequently do data analysis and I've never seen anything come even remotely close to such an extreme example. Whenever I discover something like that, then - assuming that the sample size isn't too small - I find that I've made some error while preparing or going through the data, i.e. it's the result of a bad input or a selection bias. So even though your example is hypothetically possible, in reality it's best to suspect an input/selection error rather than go ahead and continue using the data without further inspection.
A good example would be Dream's case of cheating in a Minecraft speedrun. Even while definitive proof was lacking (or faulty) at first, people were right to remain suspicious and to reject his time until a mathematician definitively solved the case, proving beyond all reasonable doubt that Dream's time was too far off from the rest of the dataset. And that proof is not perfectly logical by the way. Hypothetically it's possible that every other speedrunner before and after Dream was playing extremely illogical strats that would explain why he was able to set himself apart from all other runners. But that assumption is so unreasonable that it borders on insanity, because it would require that our perception of all speedrunners including Dream was completely off (i.e. everyone sucks at Minecraft except for Dream, and every observer misperceives every speedrunner including Dream. A completely absurd - yet statistically possible - explanation).
I love Arty and the terran comrades, but their story about the poor Terran race is getting out of hand by more than lots. At this point it isnt even a parody anymore. Reminds me about 1948. I wonder if the terran players actually would meet their goals, if they actually used their time to reach it instead of trashing the forums.
On June 20 2021 11:53 Dante08 wrote: Have you actually played TvP before and if yes at what level? Here is a video between Flash and Nada where Nada who is one of the best Terrans ever asking for TvP tips and discussing how hard the matchup is.
Do you base your opinions off just pro games or something?
What? You debate a problem by counter-arguing the others points, not by reiterating your original opinion time and time again.
Here's Flash and Nada discussing how hard TvP is. Err, so what? Of course it's hard. Any match-up is hard. Are they supposed to say "TvP is a piece of cake for me"? The question is, is it harder than PvT? If so, is it harder because Terran can die from DT/Reaver (lol)?
I can have Bisu and Stork sit there discussing how hard PvT is as well. Does that prove PvT is harder than TvP? No.
God dam I think it's pointless to argue at this point, obviously you don't play the game.
On June 21 2021 10:05 SlayerS_BunkiE wrote: I would say we always need to apply critical thinking when analyzing data. There is danger of bias for sure, but not looking at the details and taking numbers at face value can be just as dangerous. Let me ask you, if the numbers were something like this: 99 tournaments, 33 wins for each race. All T wins were from Flash. No P/Z player won more than 1 title. Do we still say “all Flash wins are T wins period”? Should we really be so closed minded as to not entertain the possibility that maybe its actually really hard to win with T, that its so hard to play that only one guy has managed to win with it after 99 tries? If Flash doesnt play on the 100th tournament, what unbiased rational person will pick T to win it? I hate exaggerating it this way but just doing it to make a point.
In this hypothetical example, I'd be more inclined to suspect an error in the dataset rather than draw the conclusion that terran is too difficult to play except for Flash. In my job I frequently do data analysis and I've never seen anything come even remotely close to such an extreme example. Whenever I discover something like that, then - assuming that the sample size isn't too small - I find that I've made some error while preparing or going through the data, i.e. it's the result of a bad input or a selection bias. So even though your example is hypothetically possible, in reality it's best to suspect an input/selection error rather than go ahead and continue using the data without further inspection.
The hypothetical is mostly a shit test. You can tell a lot about the validity of someone's position/thought process by how they answer or don't answer.
Everyone who have played the game with all 3 races know that Protoss is easiest and Terran his hardest.
One of the things that made Bisu so good was his ability to multitask, and he did it with the “easiest” race. That speaks to the insanely high skill ceiling in Starcraft, so one can only imagine what is possible to do with Z or T
For most beginners, the time it takes to start reaching X rank and win some games is the lowest for Protoss, and highest for Terran. Some individuals may have a special affinity or talent for a particular race, but this is generally the case. I've played every race from beginner to around B+ on the old ICCup, and my experience is entirely consistent with conventional wisdom. However, once you get past certain levels, Zerg starts being strong, and then Terran starts being strong.
This is a copied comment from a Youtube video replying to the claim that Terran is a weak race. The statistics are about two weeks dated at the time of this posting.
My conclusion from the statistics is that most maps that people like to make are slightly or significantly Terran favoured in one or both match ups. They are usually slightly or significantly Zerg favoured in ZvP. However, truly balanced maps can exist, and for Protoss and Zerg, maps are possible that give them one even match up and one favoured match up, although this post is only about Terran, so it doesn't show ZvP. Maps on which either Zerg or Protoss is favoured in both match ups exist but are rare.
It's also possible that the more a map is played, the better Terran becomes on it.
Also, an interesting fact is that from one year to the next, Circuit Breaker went from being significantly Terran favoured against Zerg but not Protoss, to being significantly favoured against Protoss but not Zerg. _____________________________________________________________________________________________
KSL Second place finishes (no Flash in KSL finals):
1 Terran 1 Zerg 2 Protoss
KSL first place finishes:
2 Terran 1 Zerg 1 Protoss
ASL second place finishes:
4 Terran (3 if you remove Flash) 3 Zerg 4 Protoss
ASL first place finishes:
5 Terran (1 if you remove Flash, look at that Protoss dominance without Flash, whew lads) 2 Protoss 4 Zerg
Terran Sponbbang win rates on sponsored matches by map in 2021 so far:
Match Point: TvZ: 53% (247 games) TvP: 46% (150 games)
Most current (and recently passed) maps are slightly Terran favoured at the highest level of human skill in one or both match ups. A few are significantly Terran favoured in one match up. There are a few maps which favour Z or P against T, but overall, most maps are good for Terran.
If you look at statistics for previous years, you get similar results. You can do your own research to stop being ignorant at sponbbang.com. Use ipvoid.com or google translate to translate. If Flash is the reason for these statistics, then he must have played hundreds of games more than other Terran players in sponsored matches. I watch sponsored matches, and that hasn't been my experience, and now that he's not playing at all, unless the norms for maps change, map statistics will continue to be the same without him, because most maps that people make are slightly Terran favoured in one or both match ups. There was one point in history at which the Terran winrate was only good because of Flash, and was significantly worse without him, but that was during the KeSPA era. It was true once, it's not true now. It's silly to use that excuse for every thing, and from now on, because he's gone to to the military, you can't use it for any future results.
On June 22 2021 19:21 vOdToasT wrote: There was one point in history at which the Terran winrate was only good because of Flash, and was significantly worse without him, but that was during the KeSPA era. It was true once, it's not true now. It's silly to use that excuse for every thing, and from now on, because he's gone to to the military, you can't use it for any future results.
Well, Terran kinda sucks for the first 5 weeks of this current KCM season, now that Flash's gone. But then again, he barely played in the previous KCM seasons either.
Maybe it's not Flash contributing to the stats by playing the game himself. It's his aura. When he's around, other Terran players just automatically become better.
the units can be fragile, the mechanics for building structures and defending certain rushes can cause stupid losses or disadvantages.
people love to exploit this. everyone wants to just cheese out a terran player.
terran armies once they’re built up can be easily mismanaged by bad decisions. the other races have ways of either exploiting the compositions or backstabbing or negating it’s effectiveness.
that all being said, a good terran player should be very difficult to kill, and very scary once they’re on the map with a high supply. scanner sweep, upgrades, spider mines, vessels and siege tanks are strong.
the best terran players look invincible: their builds account for all possible scenarios and feature sharp kill timings. and they have a knack for recovering in bad scenarios mounting strong comebacks
On June 08 2021 12:49 Virus(shield)08 wrote: So im not sure anyone even is on this forum anymore,but i was watching Artosis stream and he said terran is the weakest, and hardest race,but if thats true why does flash play terran,and why has he had more success with terran than the other races. the only thing i can think of is tvt is easier to be consistant than the other mirrors. I honestly have no clue. can someone who knows what they are talking about answer this question plz.
For the same reason God took human form as Jesus Christ.
On June 21 2021 00:25 att wrote: We cant ignore that there have been quite many terran bonjwas, and some zerg bonjwas, but protoss bonjwa has been weaker and shorter reign compared to the others so seems like when terrans is on top of the world they really must be an exceptional player
That is why Bisu is the best lol Seriously tho, there is no protoss bonjwa. And as great as the bonjwas are (and we should all be fans), for purposes of discussing these topics, imo their reign should be disregarded/excluded. These guys basically changed the game and everyone’s understanding of it. Thats what made them such important figures (apart from winning a bunch of tournaments). But as such, if we want to discuss the game as it exists today, how it was before the bonjwas changed it is irrelevant.
Im not sure how I feel about creating asterisks' for certain players. No reputable sports analyst would remove Michael Jordan or Lebron James from a certain stat to prove a point. It just seems very arbitrary to me and a bit vunerable to bias. Those players would have to have identical metrics between them to justify removing them from a pool of statistics. They exist, they play by the same rules and have the same opportunities as the other players, therefore you absolutely have to include them.
I think you completely misuderstood. I am not arguing to exclude them from the list of the very best to play this game. I am just saying we should exclude old games that were played with obsolete styles/strategies. For obvious reasons.
I quoted the wrong message but was talking about the post giving OSL MSL KSL and ASL gold medals but excluding Flash Jaedong and Bisu. This is just compiling stats arbitrarily to mold a narrative. It assumes that those 3 players are equals when they are not. Any statastic that creates an asterisks for Flash is meaningless if you think about it. Why do we create an exception for him? Does he not play with the same rules and opprotunities as other players? His accomplishments should absolutely be included in any compilation of stats. Now if we had a large sample size of him play R or off race, THEN maybe we could leverage his games played differently to create some sort of meaningful race balance discussion. Until then, Flash wins = terran wins, period.
Ya I got it a little bit after I replied that you quoted the wrong post. Anyway, this started when I said that I firmly believe Terrans cant compete without Flash. Take from that statement what you will. I just posted some stuff to back that up.
Let me spin it to you a different way. If you were compiling stats for average IQ of a population, would you remove the top X samples from the data? I understand people want to asterisks Flash but the reasons seem so arbitrary. If Broodwar is played for another 100 years and 2 or 3 players obtain more golds than flash, will the asterisks all of a sudden be removed?
Sorry but I really have no idea what you are trying to get at. Against my better judgment, I’ll respond to the IQ question: outliers can be excluded. That is why median/mode are also used.
Maybe the IQ example was bad. Ill return to the original sports related analogy for this discussion. If a sports (NBA) analyst were to compile data for top 10 players in each of the respected positions, do you think they would purposely not include certain players because they were too good? I get removing outliers from certain data sets but the methodology should be well defined enough to apply to any player and should contain concrete benchmarks.
It also leads to a slippery slope of creating asterisks for anything that might be deemed "outlier" for example bad map pools during certain seasons, weaker player brackets etc. You see this a lot when people try to create asterisk seasons in the NBA when players were injured during certain seasons. The consensus is its best to just leave the data intact and let the numbers speak for themselves once they reach a certain sample size. Although I agree post-savior era kespa should be a starting point for statistics gathering to focus on what we call the "modern era".
I would say we always need to apply critical thinking when analyzing data. There is danger of bias for sure, but not looking at the details and taking numbers at face value can be just as dangerous. Let me ask you, if the numbers were something like this: 99 tournaments, 33 wins for each race. All T wins were from Flash. No P/Z player won more than 1 title. Do we still say “all Flash wins are T wins period”? Should we really be so closed minded as to not entertain the possibility that maybe its actually really hard to win with T, that its so hard to play that only one guy has managed to win with it after 99 tries? If Flash doesnt play on the 100th tournament, what unbiased rational person will pick T to win it? I hate exaggerating it this way but just doing it to make a point.
Reductio ad absurdum is a great too to get your point across, provided the people reading/listening are aware and capable of understanding it. Personally I liked your point very much.
On June 21 2021 00:38 SlayerS_BunkiE wrote: [quote] That is why Bisu is the best lol Seriously tho, there is no protoss bonjwa. And as great as the bonjwas are (and we should all be fans), for purposes of discussing these topics, imo their reign should be disregarded/excluded. These guys basically changed the game and everyone’s understanding of it. Thats what made them such important figures (apart from winning a bunch of tournaments). But as such, if we want to discuss the game as it exists today, how it was before the bonjwas changed it is irrelevant.
Im not sure how I feel about creating asterisks' for certain players. No reputable sports analyst would remove Michael Jordan or Lebron James from a certain stat to prove a point. It just seems very arbitrary to me and a bit vunerable to bias. Those players would have to have identical metrics between them to justify removing them from a pool of statistics. They exist, they play by the same rules and have the same opportunities as the other players, therefore you absolutely have to include them.
I think you completely misuderstood. I am not arguing to exclude them from the list of the very best to play this game. I am just saying we should exclude old games that were played with obsolete styles/strategies. For obvious reasons.
I quoted the wrong message but was talking about the post giving OSL MSL KSL and ASL gold medals but excluding Flash Jaedong and Bisu. This is just compiling stats arbitrarily to mold a narrative. It assumes that those 3 players are equals when they are not. Any statastic that creates an asterisks for Flash is meaningless if you think about it. Why do we create an exception for him? Does he not play with the same rules and opprotunities as other players? His accomplishments should absolutely be included in any compilation of stats. Now if we had a large sample size of him play R or off race, THEN maybe we could leverage his games played differently to create some sort of meaningful race balance discussion. Until then, Flash wins = terran wins, period.
Ya I got it a little bit after I replied that you quoted the wrong post. Anyway, this started when I said that I firmly believe Terrans cant compete without Flash. Take from that statement what you will. I just posted some stuff to back that up.
Let me spin it to you a different way. If you were compiling stats for average IQ of a population, would you remove the top X samples from the data? I understand people want to asterisks Flash but the reasons seem so arbitrary. If Broodwar is played for another 100 years and 2 or 3 players obtain more golds than flash, will the asterisks all of a sudden be removed?
Sorry but I really have no idea what you are trying to get at. Against my better judgment, I’ll respond to the IQ question: outliers can be excluded. That is why median/mode are also used.
Maybe the IQ example was bad. Ill return to the original sports related analogy for this discussion. If a sports (NBA) analyst were to compile data for top 10 players in each of the respected positions, do you think they would purposely not include certain players because they were too good? I get removing outliers from certain data sets but the methodology should be well defined enough to apply to any player and should contain concrete benchmarks.
It also leads to a slippery slope of creating asterisks for anything that might be deemed "outlier" for example bad map pools during certain seasons, weaker player brackets etc. You see this a lot when people try to create asterisk seasons in the NBA when players were injured during certain seasons. The consensus is its best to just leave the data intact and let the numbers speak for themselves once they reach a certain sample size. Although I agree post-savior era kespa should be a starting point for statistics gathering to focus on what we call the "modern era".
I would say we always need to apply critical thinking when analyzing data. There is danger of bias for sure, but not looking at the details and taking numbers at face value can be just as dangerous. Let me ask you, if the numbers were something like this: 99 tournaments, 33 wins for each race. All T wins were from Flash. No P/Z player won more than 1 title. Do we still say “all Flash wins are T wins period”? Should we really be so closed minded as to not entertain the possibility that maybe its actually really hard to win with T, that its so hard to play that only one guy has managed to win with it after 99 tries? If Flash doesnt play on the 100th tournament, what unbiased rational person will pick T to win it? I hate exaggerating it this way but just doing it to make a point.
Reductio ad absurdum is a great too to get your point across, provided the people reading/listening are aware and capable of understanding it. Personally I liked your point very much.
Ty sir. I couldnt believe the response I got was “most likely an error in data set”.
Reductio ad absurdum can also be a fallacy. your example u use is too absurd to draw any conclusions from, when real life data have already shown that statements like " All T wins were from Flash." is clearly untrue