|
In addition to three days of insane gameplay, MLG Columbus provided us with a wealth of information that should be of great interest to players and fans alike. From the most banelings made in a single game, right down to a detailed race-by-race and map-by-map win/loss breakdown, we've got the stats, and we're sharing them with you. So without further ado, here it is: MLG Columbus by the numbers.
Win Percentage by Matchup: Entire Tournament
PvZ:
Protoss: 46.51%
Zerg: 53.49%
PvT:
Protoss: 54%
Terran: 46%
TvZ:
Terran: 43.21%
Zerg: 56.79%
Fun Facts:
Most Banelings in a single Game: 380 – (18)SlayersSMMA vs (17)IMLosira Game 4
Most times Spawn Larva was cast in a single Game: 185 – (58)mouzMorroW vs (71)goswser Game 1
For the rest of the funf facts, as well as Average APM & Matchups by Map, head over to the Stats Breakdown.
On June 22 2011 08:22 MLGStephen wrote:
The matchup stats are for the Championship Bracket only.
|
Damn, That is a lot of Banelings. Crazy to see the percentage of wins for each race, I thought Terrans would be a bit lower than it is.
|
Awesome stats breakdown. The winning %'s were the most impressive.
(This might be fixed by time i post it but that is racial breakdown not winning percentage in the OP)
Win Percentage by Matchup: Entire Tournament
PvZ:
Z: 53.49%
P: 46.51%
PvT:
P: 54%
T: 46%
ZvT:
Z: 56.79%
T: 43.21%
|
On June 22 2011 06:49 Slasher wrote:In addition to three days of insane gameplay, MLG Columbus provided us with a wealth of information that should be of great interest to players and fans alike. From the most banelings made in a single game, right down to a detailed race-by-race and map-by-map win/loss breakdown, we've got the stats, and we're sharing them with you. So without further ado, here it is: MLG Columbus by the numbers. Win Percentage by Race: Entire Tournament Zerg: 34.38% Protoss: 25% Terran: 40.63% Fun Facts: Most Banelings in a single Game: 380 – (18)SlayersSMMA vs (17)IMLosira Game 4Most times Spawn Larva was cast in a single Game: 185 – (58)mouzMorroW vs (71)goswser Game 1For the rest of the funf facts, as well as Average APM & Matchups by Map, head over to the Stats Breakdown.
Those numbers for "win percentage by race" I believe are incorrect, since on the website it says "Percentage of Players of Each Race", not win rate. The actual win rates are much much closer.
If that's correct, you should probably edit the OP before 10 pages of people throwing around the words "Sample size" with no idea what they are talking about, etc...
nice data btw, thanks!
|
Select with 374 average apm..
6.2 actions per second.. omfg thats fast :O
|
On June 22 2011 06:49 Slasher wrote:
Win Percentage by Race: Entire Tournament
Zerg: 34.38%
Protoss: 25%
Terran: 40.63%
you mean the Percentage of Players of Each Race
|
On June 22 2011 06:49 Slasher wrote:In addition to three days of insane gameplay, MLG Columbus provided us with a wealth of information that should be of great interest to players and fans alike. From the most banelings made in a single game, right down to a detailed race-by-race and map-by-map win/loss breakdown, we've got the stats, and we're sharing them with you. So without further ado, here it is: MLG Columbus by the numbers. Win Percentage by Race: Entire Tournament Zerg: 34.38% Protoss: 25% Terran: 40.63% Fun Facts: Most Banelings in a single Game: 380 – (18)SlayersSMMA vs (17)IMLosira Game 4Most times Spawn Larva was cast in a single Game: 185 – (58)mouzMorroW vs (71)goswser Game 1For the rest of the funf facts, as well as Average APM & Matchups by Map, head over to the Stats Breakdown.
Those are "Percentage of Player of Each Race" not "Win Percentage" like you listed in OP.
|
Percentage of Players of Each Race
Zerg: 34.38%
Protoss: 25%
Terran: 40.63%
|
On June 22 2011 06:56 JoeSchmoe wrote:Show nested quote +On June 22 2011 06:49 Slasher wrote:
Win Percentage by Race: Entire Tournament
Zerg: 34.38%
Protoss: 25%
Terran: 40.63%
you mean the Percentage of Players of Each Race
lol it's ironic because terran actually had the lowest win rate by quite a margin
|
On June 22 2011 06:54 j4ckd4v13z wrote:
Select with 374 average apm..
6.2 actions per second.. omfg thats fast :O
Is that in-game apm? Cause if so, that would actually be about 516 irl apm, 8.6 actions per second if I'm not mistaken.
|
On June 22 2011 06:57 Fr33t wrote:Show nested quote +On June 22 2011 06:56 JoeSchmoe wrote:On June 22 2011 06:49 Slasher wrote:
Win Percentage by Race: Entire Tournament
Zerg: 34.38%
Protoss: 25%
Terran: 40.63%
you mean the Percentage of Players of Each Race lol it's ironic because terran actually had the lowest win rate by quite a margin
If quite a margin is 3-5% than sure...... But i think it is interesting how some races do better in different regions. By this i mean zergs are doing really well in north America (10/16) in nasl, 5/8 in mlg, while terrans are dominating korea, and lets not forget about the European protoss that had a good showing at dream hack.
|
Well, since most people are just interested in the win % here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop.
Edit:LOL I am an idiot, sorry OP I chastised people fro not reading and yet I missed some stuff in your OP
|
Vibe has a pretty damn crazy APM for a white dude.
|
|
|
These stats don't reflect all of the games played at MLG. What games do they cover? I know for a fact that a 0% winrate by protoss against terran on typhon is false because I lost a tvp there.
|
On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop.
So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know.
However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG.
|
I like it. Machine actually is a machine with exactly 200 APM
|
On June 22 2011 07:11 SuperStyle wrote:Show nested quote +On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG.
The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average.
That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance"
|
Interesting to see, for sure, but be careful not to read too much into it. It's just results from one tournament.
|
Is that apm ingame apm or 1.4(real) apm?
|
On June 22 2011 07:20 worldsnap wrote:Show nested quote +On June 22 2011 07:11 SuperStyle wrote:On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG. The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average. That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance"
The stats show the matchups win percentages, not individual races win percentages. Please read the information correctly before posting.
|
I'd like to see the stats for only the top 20 players tbh
|
On June 22 2011 07:27 andeh wrote:
Is that apm ingame apm or 1.4(real) apm?
In-game APM.
|
|
|
On June 22 2011 07:30 LagT_T wrote:Show nested quote +On June 22 2011 07:20 worldsnap wrote:On June 22 2011 07:11 SuperStyle wrote:On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG. The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average. That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance" The stats show the matchups win percentages, not individual races win percentages. Please read the information correctly before posting.
That's exactly what i'm saying.. read my post before posting.
Of all the Z and T players at MLG, T will win ~43% of the time against Z. That doesn't tell you about the balance of the game nearly as much as it tells you about the selection of players that represented Z and T at MLG.
Think of it this way: if Idra played 100 games against Sjow and won 70 of them, would you say that Z is OP? No, you would say that Idra is better than than Sjow in ZvT.
|
lol@terran some of highest AND lowest APM (MMA vs. Sjow: 360 vs. 140). Axslav wins the slow and steady race -.- too bad the hare didn't attend MLG.
Typhon peaks? imba imba imba. 100% TvP rate 
-Bleh, I don't like these statistics --> they are nice but misleading. They don't give enough background: are these from championship bracket, or overall? How do the numbers break down (like instead of %, non-simplified fractions).
|
On June 22 2011 07:11 SuperStyle wrote:Show nested quote +On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG.
Its around 5% so statistically its an acceptable outlier. I play random so I don´t even have anything invested in Zerg being OP lol. I have an extreme disgust with any kinda of QQ though
|
Could we get only from the pool play and beyond win rates? I just have a feeling a lot of silver leaguers or similar joined in the open bracket
|
On June 22 2011 07:34 worldsnap wrote:Show nested quote +On June 22 2011 07:30 LagT_T wrote:On June 22 2011 07:20 worldsnap wrote:On June 22 2011 07:11 SuperStyle wrote:On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG. The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average. That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance" The stats show the matchups win percentages, not individual races win percentages. Please read the information correctly before posting. That's exactly what i'm saying.. read my post before posting. Of all the Z and T players at MLG, T will win ~43% of the time against Z. That doesn't tell you about the balance of the game nearly as much as it tells you about the selection of players that represented Z and T at MLG. Think of it this way: if Idra played 100 games against Sjow and won 70 of them, would you say that Z is OP? No, you would say that Idra is better than than Sjow in ZvT.
Exactly , this kind of statistics are nice but don´t really say much. They are still nice and the less ammo people have to QQ about the better IMHO
|
On June 22 2011 07:30 LagT_T wrote:Show nested quote +On June 22 2011 07:20 worldsnap wrote:On June 22 2011 07:11 SuperStyle wrote:On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG. The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average. That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance" The stats show the matchups win percentages, not individual races win percentages. Please read the information correctly before posting.
He is correct actually, its easy to calculate individual race win percentage if u have all of the match up percentages, so his conclusion that zerg players won more than terran players on average is correct.
As for the guy that wanted only top 20 stats u can find that in TLPD but u have to check manually and individually each one.
|
Pretty sure that APM is adjusted APM.
I recall watching games of Losira/MC/MMA and their APM was around 270~ish from what I remember.
|
I think the really interesting thing is the match ratio per map.
It seems maps are really, really influencing the win ratio for each match ups, but we cannot say for sure since we do not have the sheer number of game played on every maps (for exemple, I'm sure Testbug, Shakuras and Tal'Darim were played quite a bit, but I don't think many played Typhon peaks for exemple).
Just giving an exemple for PvZ
Testbug : PvZ 20%
Xel Naga Cavern : PvZ: 33,30%
Those two maps seems to favor of Zergs a lot.
Then Metalopolis with a 40% PvZ, seems to favor zergs a bit.
Then, in the middle you have Shattered Temple and Shakuras Plateau both at 50% win ratio, perfect balance.
Then, at the other side of the spectrum, Tal'Darim Altar with a 70% win ratio in PvZ.
Seems, like a lot of guys said, that the solution to everything is in the map making.
|
yea losira had around 250apm from the replays i saw so tht must be real apm not in-game.
|
|
|
On June 22 2011 07:34 worldsnap wrote:Show nested quote +On June 22 2011 07:30 LagT_T wrote:On June 22 2011 07:20 worldsnap wrote:On June 22 2011 07:11 SuperStyle wrote:On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG. The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average. That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance" The stats show the matchups win percentages, not individual races win percentages. Please read the information correctly before posting. That's exactly what i'm saying.. read my post before posting. Of all the Z and T players at MLG, T will win ~43% of the time against Z. That doesn't tell you about the balance of the game nearly as much as it tells you about the selection of players that represented Z and T at MLG. Think of it this way: if Idra played 100 games against Sjow and won 70 of them, would you say that Z is OP? No, you would say that Idra is better than than Sjow in ZvT.
I'm sorry, how do YOU calculate balance from statistics?
Try using a big enough top level player pool with a big enough sample of games. If MLG was big enough or not is another debate.
|
The APM is "real" APM.
The matchup stats are for the Championship Bracket only.
Here are the stats for entire tournament (Open + Championship):
Percentage of Players of Each Race:
Zerg: 33.45%
Protoss: 37.72%
Terran: 28.83%
Matchups:
PvZ:
Zerg: 45.51%
Protoss: 54.49%
PvT:
Terran: 54.97%
Protoss: 45.03%
TvZ:
Zerg: 41.94%
Terran: 58.06%
Matchups by Map:
MLG Metalopolis:
PvZ:
Zerg: 53.57%
Protoss: 46.43%
PvT:
Protoss: 50%
Terran: 50%
TvZ:
Zerg: 42.11%
Terran: 57.89%
MLG Shakuras Plateau:
PvT:
Terran: 62.5%
Protoss: 37.5%
TvZ:
Zerg: 46.67%
Terran: 53.33%
PvZ:
Protoss: 52.17%
Zerg: 47.83%
MLG Xel'Naga Caverns:
PvZ:
Zerg: 46.88%
Protoss: 53.13%
PvT:
Terran: 57.41%
Protoss: 42.59%
TvZ:
Zerg: 38.1%
Terran: 61.9%
MLG Testbug:
PvT:
Terran: 61.11%
Protoss: 38.89%
PvZ:
Zerg: 54.84%
Protoss: 45.16%
TvZ:
Zerg: 57.89%
Terran: 42.11%
MLG Shattered Temple:
PvT:
Protoss: 55.56%
Terran: 44.44%
PvZ:
Zerg: 44.12%
Protoss: 55.88%
TvZ:
Zerg: 43.9%
Terran: 56.1%
MLG Typhon Peaks:
TvZ:
Zerg: 33.33%
Terran: 66.67%
PvZ:
Zerg: 44.12%
Protoss: 55.88%
PvT:
Terran: 65.52%
Protoss: 34.48%
MLG Tal'darim Altar:
PvZ:
Protoss: 71.74%
Zerg: 28.26%
TvZ:
Zerg: 37.84%
Terran: 62.16%
PvT:
Protoss: 47.83%
Terran: 52.17%
|
The APM stats are pretty interesting. Especially "If one player's APM was at least 50 greater than the other's, that player won 65.7% of the time."
|
On June 22 2011 08:26 Draconicfire wrote:
The APM stats are pretty interesting. Especially "If one player's APM was at least 50 greater than the other's, that player won 65.7% of the time."
But this statistic is greatly skewed by the Koreans, unless we assume Korean apm is what is causing them to win and dominate (which may be true).
|
On June 22 2011 08:33 Micket wrote:Show nested quote +On June 22 2011 08:26 Draconicfire wrote:
The APM stats are pretty interesting. Especially "If one player's APM was at least 50 greater than the other's, that player won 65.7% of the time." But this statistic is greatly skewed by the Koreans, unless we assume Korean apm is what is causing them to win and dominate (which may be true).
The APM/Win % stat included all games in both Open in Championship brackets, so the Koreans didn't skew it *too* much.
FYI: Thats approximately 1300 games.
|
Testbug is going to be very problematic. 80% Zerg overall winrate against P/T.
|
On June 22 2011 08:09 LagT_T wrote:Show nested quote +On June 22 2011 07:34 worldsnap wrote:On June 22 2011 07:30 LagT_T wrote:On June 22 2011 07:20 worldsnap wrote:On June 22 2011 07:11 SuperStyle wrote:On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG. The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average. That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance" The stats show the matchups win percentages, not individual races win percentages. Please read the information correctly before posting. That's exactly what i'm saying.. read my post before posting. Of all the Z and T players at MLG, T will win ~43% of the time against Z. That doesn't tell you about the balance of the game nearly as much as it tells you about the selection of players that represented Z and T at MLG. Think of it this way: if Idra played 100 games against Sjow and won 70 of them, would you say that Z is OP? No, you would say that Idra is better than than Sjow in ZvT. I'm sorry, how do YOU calculate balance from statistics? Try using a big enough top level player pool with a big enough sample of games. If MLG was big enough or not is another debate.
Well, tbh something as ambiguous as balance is extremely difficult to determine by any means. Its a good sample this is a case that can´t really be measured and adjusted numerically(nor by gut reaction). I am not saying this data is 100% useless and its nice data to have but at one moment we will have to accept that no matter how hard we try some people will always consider something Imbalanced while some people won´t.
We just need to discourage all forms off QQing and try to make the game better instead of crying OP at everything
|
LOL Typhon peaks TvP 100%
are you sure about the 1300 games sample size?
|
APM is so useless, for example you can see select at 373 APM and MMA at 359 APM, but from watching both players in FPVODs and replays I can tell you that MMA is __much__ faster than select, select just spams 123123123123123123 all game long and people give him credit playing fast, whereas the true magicians are almost all of the time the Koreans and a few select Europeans (e.g. Strelok) who have impressed me the most with their hand speed.
|
Every Korean has higher APM than every foreigner except for Select. I think that proves that APM is important.
|
Testbug is the only map that Zerg has a good win rate all other maps besides metalopolis PvZ favor terran and protoss vs zerg just from those stats
|
On June 22 2011 07:31 ZAiNs wrote:In-game APM.
I'm not sure about that. Unless sjow and axslav suddenly have lan jitters, they definitely dont have 140 apm. Also, i think select averages around 250-300 from the replays ive seen of him. 250*1.4 = 350 so that would be close enough.
|
Testbug is the only map that Zerg has a good win rate all other maps besides metalopolis PvZ favor terran and protoss vs zerg just from those stats
Xel'Naga?
|
On June 22 2011 08:39 ChickenLips wrote:
APM is so useless, for example you can see select at 373 APM and MMA at 359 APM, but from watching both players in FPVODs and replays I can tell you that MMA is __much__ faster than select, select just spams 123123123123123123 all game long and people give him credit playing fast, whereas the true magicians are almost all of the time the Koreans and a few select Europeans (e.g. Strelok) who have impressed me the most with their hand speed.
There are FPVODS from MMA? Link please? I want them @_@
|
On June 22 2011 08:41 EnWara wrote:
Every Korean has higher APM than every foreigner except for Select. I think that proves that APM is important.
That proves that being mechanically proficient is important. You can have a trillion APM by spamming yet never build a unit. See the difference?
|
On June 22 2011 08:39 ChickenLips wrote:
APM is so useless, for example you can see select at 373 APM and MMA at 359 APM, but from watching both players in FPVODs and replays I can tell you that MMA is __much__ faster than select, select just spams 123123123123123123 all game long and people give him credit playing fast, whereas the true magicians are almost all of the time the Koreans and a few select Europeans (e.g. Strelok) who have impressed me the most with their hand speed.
I 100% agree
watching MMA and Select play in FPVODs is like night and day
This is why I argue that APM DOES matter because the really high level players actually make use of it and they go up that high when they macro/multitask
A lot of people seem to think that if anyone has over 150 apm they're "spamming" but it's obviously not true. Players like MMA are doing important things with it
|
On June 22 2011 08:37 wxwx wrote:
Testbug is going to be very problematic. 80% Zerg overall winrate against P/T.
How many games were played on Testbug? Because its such an nice number, its either like 4 out of 5 games or 8 out of 10- neither of which are large enough for a ample sample size
|
so tl;dr, zerg > protoss and zerg > terran
i wonder how much worse the PvZ stats would be if you disregarded MC's games...
|
On June 22 2011 08:09 LagT_T wrote:
I'm sorry, how do YOU calculate balance from statistics?
Try using a big enough top level player pool with a big enough sample of games. If MLG was big enough or not is another debate.
It's worth noting that, in a tournament situation, a few players who place highly with a given race can completely skew the numbers. The reason is that winners keep playing while losers stop.
The ultimate winner of a single-elimination tournament with 16 entrants is going to win four times, while the median player is going to win once (since half the players get eliminated in the first round.)
Edit: If you really wanted statistics that showed race balances clearly, you'd have to have much more random matchups between players, and all players playing the same number of games. The ladder would be a good place for that, while a single or double-elimination tournament would not be.
Edit 2: I'll bet if you looked at the stats from the GSL Super Tournament,
+ Show Spoiler +
the numbers would heavily favor Terran, because the most number of rounds won by any non-Terran player was three, while the winner won 7.
|
On June 22 2011 08:59 lilky wrote:so tl;dr, zerg > protoss and zerg > terran i wonder how much worse the PvZ stats would be if you disregarded MC's games...
This is what should happen when 4 of the 6 best players in the tournament are zerg.
|
On June 22 2011 09:00 Lysenko wrote:Show nested quote +On June 22 2011 08:09 LagT_T wrote:
I'm sorry, how do YOU calculate balance from statistics?
Try using a big enough top level player pool with a big enough sample of games. If MLG was big enough or not is another debate. It's worth noting that, in a tournament situation, a few players who place highly with a given race can completely skew the numbers. The reason is that winners keep playing while losers stop. The ultimate winner of a single-elimination tournament with 16 entrants is going to win four times, while the median player is going to win once (since half the players get eliminated in the first round.) Edit: If you really wanted statistics that showed race balances clearly, you'd have to have much more random matchups between players, and all players playing the same number of games. The ladder would be a good place for that, while a single or double-elimination tournament would not be. Edit 2: I'll bet if you looked at the stats from the GSL Super Tournament, + Show Spoiler +
the numbers would heavily favor Terran, because the most number of rounds won by any non-Terran player was three, while the winner won 7.
That's applicable when you are looking at a race overall win percentage, but this is matchup win percentage, which is independent of any individual result. You use the total amount of games of that matchup played, regardless of the tournament round, etc, they all weight the same.
|
On June 22 2011 08:41 EnWara wrote:
Every Korean has higher APM than every foreigner except for Select. I think that proves that APM is important.
A lot of the top 16 GSL koreans average 170-220
Idra and Naniwa, as another example, average 160-220~
The ones averaging 300+ don't necessarily play better, yet.
|
On June 22 2011 08:50 Itsmedudeman wrote:Show nested quote +On June 22 2011 08:39 ChickenLips wrote:
APM is so useless, for example you can see select at 373 APM and MMA at 359 APM, but from watching both players in FPVODs and replays I can tell you that MMA is __much__ faster than select, select just spams 123123123123123123 all game long and people give him credit playing fast, whereas the true magicians are almost all of the time the Koreans and a few select Europeans (e.g. Strelok) who have impressed me the most with their hand speed. I 100% agree watching MMA and Select play in FPVODs is like night and day This is why I argue that APM DOES matter because the really high level players actually make use of it and they go up that high when they macro/multitask A lot of people seem to think that if anyone has over 150 apm they're "spamming" but it's obviously not true. Players like MMA are doing important things with it
Where did you find MMA's FPVODs? I've never seen one before.
|
wow at select and vibe's apm
|
Yeah, I didn't expect select and vibe's apm to be above/around the Koreans. I guess it's possible they like to spam click more than the normal person, however.
|
the MLG Stats are real APM, i watched the MMA replays versus losira and they all say like 220 APM but then you multiply it by i think somewhere around 1.3 and then you get real apm
|
I'm pretty sure they got all the statistics using SC2Gears, since all that stats listed are available by doing a multi-rep analysis in SC2Gears.
It's Real APM listed, not SC2 APM. 100 In-game APM = 138 Real APM. Just look at the replays or open some in SC2Gears and you'll see Idra has 170 In-Game APM (236 Real APM).
The Fun Facts listed are incorrect. If you open the Losira vs MMA Game 4 in SC2Gears and go to the Builds/Tech Stats chart, it shows 380 Baneling morphs. However, the Builds/Tech Stats chart isn't accurate, because it includes some of the attempted morphs (when you press the button but don't have enough resources). There was actually ~198 banelings (I counted), not 380.
SC2Gears does the same thing with buildings/abilities/researches/upgrades/trains. It counts the action whether or not the order went through. For example, you select a drone, tell it to build a spawning pool, then before it gets there you make it build a spawning pool somewhere else instead, it'll show 2 Spawning Pools in the Build Stats
I'm assuming there wasn't 47 Ultralisks or 52 Ghosts or 416 Infested Terrans in those other games either.
|
That is some pretty equal distrisbution of wins there, I like it. Too bad ladder isn't more like that.. not for me anyway
|
So any map on which Protoss is allowed to Forge Expand they seem to be favoured, on the others Zerg is favoured.
I would love to see which player made the most ultralisks/warp prisms/carriers throughout the tournament. I don't suppose it's possible to do stats by player and not by game?
|
On June 22 2011 09:27 LagT_T wrote:
That's applicable when you are looking at a race overall win percentage, but this is matchup win percentage, which is independent of any individual result. You use the total amount of games of that matchup played, regardless of the tournament round, etc, they all weight the same.
If there are more Zerg in the final brackets than there were in the starting brackets, that indicates that Zerg had to beat T and P disproportionately to get there. The extra ZvZ matches don't make up for that.
Another factor is that if any race is overrepresented in the starting bracket, that race is likely to lose more than the other races, because more of their players will get knocked out early. Terran players made up 40% of the participants, so more of them are likely to lose in early rounds (since the average number of games played is 2, both losses.)
Remember, in a 6-round double elimination tournament, the average player will lose twice, while the winner is guaranteed to have at most 1 loss but at least 6 wins. Anyone who makes it to the last three rounds is guaranteed to be winning more than the average.
|
On June 22 2011 11:25 Lysenko wrote:Show nested quote +On June 22 2011 09:27 LagT_T wrote:
That's applicable when you are looking at a race overall win percentage, but this is matchup win percentage, which is independent of any individual result. You use the total amount of games of that matchup played, regardless of the tournament round, etc, they all weight the same. If there are more Zerg in the final brackets than there were in the starting brackets, that indicates that Zerg had to beat T and P disproportionately to get there. The extra ZvZ matches don't make up for that. Another factor is that if any race is overrepresented in the starting bracket, that race is likely to lose more than the other races, because more of their players will get knocked out early. Terran players made up 40% of the participants, so more of them are likely to lose in early rounds (since the average number of games played is 2, both losses.) Remember, in a 6-round double elimination tournament, the average player will lose twice, while the winner is guaranteed to have at most 1 loss but at least 6 wins. Anyone who makes it to the last three rounds is guaranteed to be winning more than the average.
1) That disproportion is the win % linked in the OP.
2) The overrepresentation in the opening bracket is an issue I'm willing to discuss because I haven't found out how to solve it or if there is a need to solve it.
3) Same as point 1.
|
Lets Face it, Foreign terrans are really bad atm
Select is the only foreign terran that is decent which did poorly last MLG.
|
On June 22 2011 07:11 SuperStyle wrote:Show nested quote +On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG.
Speaking as a terran, there were a lot of superb zergs in this tournament, while the skill level of the terran players wasn't just as high. Things like that can have a big effect on the results. However people saying 7% difference is small... Well it ain't. I'm just so thankful July is playing SC2, else zergs would probably still be QQing about T like idra.
Anyway MLG XC, TvP/PvT has 105% win rate.
And jesus those koreans are fast.
|
On June 22 2011 07:01 Yamulo wrote:Show nested quote +On June 22 2011 06:57 Fr33t wrote:On June 22 2011 06:56 JoeSchmoe wrote:On June 22 2011 06:49 Slasher wrote:
Win Percentage by Race: Entire Tournament
Zerg: 34.38%
Protoss: 25%
Terran: 40.63%
you mean the Percentage of Players of Each Race lol it's ironic because terran actually had the lowest win rate by quite a margin If quite a margin is 3-5% than sure...... But i think it is interesting how some races do better in different regions. By this i mean zergs are doing really well in north America (10/16) in nasl, 5/8 in mlg, while terrans are dominating korea, and lets not forget about the European protoss that had a good showing at dream hack.
It's hard to deduce anything meaningful from a single tournament. Morrow wisely said in a dreamhack interview that you could replay a tournament and probably end up with different stats
|
On June 22 2011 12:23 LagT_T wrote:
1) That disproportion is the win % linked in the OP.
2) The overrepresentation in the opening bracket is an issue I'm willing to discuss because I haven't found out how to solve it or if there is a need to solve it.
3) Same as point 1.
The point is that disproportion will also show up in the matchup win/loss percentages. For all the Terrans and Protoss to lose before the final round, Zerg must have extra wins vs. P and T beyond their representation in the starting population (and beyond 50/50.)
Remember, in a double elimination tournament, the *median* player loses *all* their games, and I believe the average player wins two.
|
Selects average apm 373 holy shit, I didn't even think people hit that in sc2
|
On June 22 2011 12:29 DarkRise wrote:
Lets Face it, Foreign terrans are really bad atm
Select is the only foreign terran that is decent which did poorly last MLG.
I wouldn't say foreign terrans in general more NA terrans probably. I imagine the EU terrans are still super strong but I don't know to much about the EU scene just seems they always have super strong terrans kinda like Korea.
|
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Qxc: Loses to ThorZain and SeleCT. 8-2
26th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place.
|
On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place.
Because Moonan took 1 more game off someone in a bo3.
|
On June 22 2011 13:01 Suc wrote:Show nested quote +On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place. Because Moonan took 1 more game off someone in a bo3.
The point is players losing every single match and placing higher than Anyone else in the tournament.
You cannot be top 32 and not win a single bo3. That's just pathetic tournament structure.
I can lose 6 bo3s too. I'm plat league.
|
Wow. I already had a hunch from following the tourment and watching reps, but the Terran stats are pretty insane. They're almost identical to NASL. If there weren't so many Terrans, we might not have even seen one in the championship bracket at all :p Except MMA of course.
|
On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Qxc: Loses to ThorZain and SeleCT. 8-2
26th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place.
This is BY FAR the worst part of the MLG system.
|
Im pretty sure alot of those stats can't be real. I refuse to believe losira created near 400 banelings that simply isnt possible.
Edit: Ahh, post on top of page explains it.
|
On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place.
Richman: 6-2
31st place.
Wow that is quite pathetic, its ridiculous how gretorp was still even in the championship bracket, he went 1-5 in dallas =/
|
On June 22 2011 08:44 windsupernova wrote:Show nested quote +On June 22 2011 08:39 ChickenLips wrote:
APM is so useless, for example you can see select at 373 APM and MMA at 359 APM, but from watching both players in FPVODs and replays I can tell you that MMA is __much__ faster than select, select just spams 123123123123123123 all game long and people give him credit playing fast, whereas the true magicians are almost all of the time the Koreans and a few select Europeans (e.g. Strelok) who have impressed me the most with their hand speed. There are FPVODS from MMA? Link please? I want them @_@
You could download the replays and watch them from MMA's view point right (clicking on the camera icon)? Thats the same as fpv I think.
And check out sjows apm,you dont need to spam to be a top player! 
|
lol at the map stats testbug the only map worth playing as zerg, xel'naga tal'darim and typhon peaks so skewed to p and t. Yet i still like playing those maps.
|
On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Qxc: Loses to ThorZain and SeleCT. 8-2
26th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place.
Ya this is why I'm not a fan of the structure. It is too forgiving. Oh well.
|
On June 22 2011 13:13 dartoo wrote:Show nested quote +On June 22 2011 08:44 windsupernova wrote:On June 22 2011 08:39 ChickenLips wrote:
APM is so useless, for example you can see select at 373 APM and MMA at 359 APM, but from watching both players in FPVODs and replays I can tell you that MMA is __much__ faster than select, select just spams 123123123123123123 all game long and people give him credit playing fast, whereas the true magicians are almost all of the time the Koreans and a few select Europeans (e.g. Strelok) who have impressed me the most with their hand speed. There are FPVODS from MMA? Link please? I want them @_@ You could download the replays and watch them from MMA's view point right (clicking on the camera icon)? Thats the same as fpv I think. And check out sjows apm,you dont need to spam to be a top player!
That´s not the same and you know it >_<
I need to see the mouse cursor. Anyways thanks for the answer `
|
On June 22 2011 13:02 two.watup wrote:Show nested quote +On June 22 2011 13:01 Suc wrote:On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place. Because Moonan took 1 more game off someone in a bo3. The point is players losing every single match and placing higher than Anyone else in the tournament. You cannot be top 32 and not win a single bo3. That's just pathetic tournament structure. I can lose 6 bo3s too. I'm plat league.
It is called a circuit. Each stop is one event in the season, as you play in more events you will see the weaker players fall lower. Don't think of MLG events as one tournament where the past results did not matter. Give it a year or two and you will see a insane pool in the top 32.
|
I don't think we should look too much into balance with these win ratios. The Koreans basically dominated this tournament so any stats without Koreans is basically meaningless in my opinion.
|
|
|
[QUOTE]On June 22 2011 07:34 worldsnap wrote:
[QUOTE]On June 22 2011 07:30 LagT_T wrote:
[QUOTE]On June 22 2011 07:20 worldsnap wrote:
[QUOTE]On June 22 2011 07:11 SuperStyle wrote:
[QUOTE]On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop.[/QUOTE]
So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know.
However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG.[/QUOTE]
The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average.
That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance"[/QUOTE]
[/QUOTE]
QFT
The problem with stats such as these for open tournaments such as MLG is that since anyone could sign up, there could theoretically be a tournament with 1 zerg, 254 terrans, 1 protoss. Zerg now represents less than 1% of the playing population, however say its idra in the open bracket and he makes it to the quarter finals (in a single elim bracket) without losing a game and playing only terrans, then goes 0-2 he would finish with a record of 12-2 which would make that matchup read ZvT is 85% win rate for zerg.
The issue is that the general population (in my experience) doesn't understand statistics well enough to differentiate between correlation and causation. Assuming all races were perfectly balanced and had the exact same skill level at spots 1-100 of their race with the other races, in a tournament with 40% terrans, the average skill of a terran will be lower, thus pulling down their average win rate.
Now I'm not arguing that the game is balanced or these stats mean nothing, but i think outside of the oooo factor, they are mostly useless unless you cut to like top 64/32 of the tournament where the nobodies/for funsies people have all been knocked out
|
vibe has shockingly high apm and i kind of want to watch the avilo game with 52 nooks
|
I can't believe so many people are talking about apm.
Now, you made me mention it
|
On June 22 2011 14:57 windsupernova wrote:That´s not the same and you know it >_< I need to see the mouse cursor. Anyways thanks for the answer `
Dig through the history of Cella's stream and you will find one. Complete with webcam if you go back far enough.
|
On June 22 2011 07:30 LagT_T wrote:Show nested quote +On June 22 2011 07:20 worldsnap wrote:On June 22 2011 07:11 SuperStyle wrote:On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG. The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average. That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance" QFT The problem with stats such as these for open tournaments such as MLG is that since anyone could sign up, there could theoretically be a tournament with 1 zerg, 254 terrans, 1 protoss. Zerg now represents less than 1% of the playing population, however say its idra in the open bracket and he makes it to the quarter finals (in a single elim bracket) without losing a game and playing only terrans, then goes 0-2 he would finish with a record of 12-2 which would make that matchup read ZvT is 85% win rate for zerg. The issue is that the general population (in my experience) doesn't understand statistics well enough to differentiate between correlation and causation. Assuming all races were perfectly balanced and had the exact same skill level at spots 1-100 of their race with the other races, in a tournament with 40% terrans, the average skill of a terran will be lower, thus pulling down their average win rate. Now I'm not arguing that the game is balanced or these stats mean nothing, but i think outside of the oooo factor, they are mostly useless unless you cut to like top 64/32 of the tournament where the nobodies/for funsies people have all been knocked out
The only thing you can conclude concerning balance is the impact of maps on every match ups, and that's the main thing interesting here in my opinion, if you consider the difference in win ratio in each match ups per maps.
At such a high level of play (let's say MLG is high master / grand master ? or just GM ?) the win loss ratio are very, very different from one map to another. Testbug being a zerg favored map (80% win ratio) for exemple.
|
On June 22 2011 15:00 Kralic wrote:Show nested quote +On June 22 2011 13:02 two.watup wrote:On June 22 2011 13:01 Suc wrote:On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place. Because Moonan took 1 more game off someone in a bo3. The point is players losing every single match and placing higher than Anyone else in the tournament. You cannot be top 32 and not win a single bo3. That's just pathetic tournament structure. I can lose 6 bo3s too. I'm plat league. It is called a circuit. Each stop is one event in the season, as you play in more events you will see the weaker players fall lower. Don't think of MLG events as one tournament where the past results did not matter. Give it a year or two and you will see a insane pool in the top 32.
So players constantly getting top 32 placement despite losing everything and playing much worse than a lot of the people placing significantly lower for a "year or two" sounds fine to you?
Saying that it's called <insert bad format name here> as an excuse for it being a bad format is not really an excuse. It's the same as saying "it's called having a bad format", except this one has a clear meaning, while yours tries to make it seem like something else.
|
On June 22 2011 07:16 BRJ wrote:
I like it. Machine actually is a machine with exactly 200 APM
Ahaha that's a hilarious coincidence.
or is it...
Well, with Terran so dominant in Korea, maybe it shows that Terran rewards skill more than any other race (which makes sense considering the highest amount of multitasking with drops and whatnot). At the lower levels Terran is clearly not as dominant, perhaps perhaps.
|
On June 22 2011 17:47 Ruscour wrote:Show nested quote +On June 22 2011 07:16 BRJ wrote:
I like it. Machine actually is a machine with exactly 200 APM Ahaha that's a hilarious coincidence. or is it... Well, with Terran so dominant in Korea, maybe it shows that Terran rewards skill more than any other race (which makes sense considering the highest amount of multitasking with drops and whatnot). At the lower levels Terran is clearly not as dominant, perhaps perhaps.
Terran is not that dominant in Korea, GSL ST is just a frame of a period, just a tournament.
Yes, in GSL ST, Terran is dominant, but if you look at the LG Cinema 3D special league ro16 which ran the same time with GSL, all the Terran (MMA, Ryung, MarineKing, Teaja, SuperNova) got knocked out (mostly in 2:0), and the only non-Terran got knocked out is Genius (defeated by Rain), and sC got in by TvT.
|
On June 22 2011 15:12 K3Nyy wrote:
I don't think we should look too much into balance with these win ratios. The Koreans basically dominated this tournament so any stats without Koreans is basically meaningless in my opinion.
Good point.
Throw in only race X players from Korea and you'll see how imba that race is.
So pretty much GSL is the best source for statistics.
|
On June 22 2011 15:44 thebig1 wrote:Show nested quote +On June 22 2011 14:57 windsupernova wrote:That´s not the same and you know it >_< I need to see the mouse cursor. Anyways thanks for the answer ` Dig through the history of Cella's stream and you will find one. Complete with webcam if you go back far enough.
I'm interested in this, how far back do you believe it was?
Edit: nvm, found one: http://www.justin.tv/slayerscella/b/283474828 Starts around 27min in, only one game of MMA in this one though, sadly.
Edit 2: Found another better one: http://www.justin.tv/slayerscella/b/285427634 Multiple games from Ryung and MMA in this one. (first 2 parts only)
|
On June 22 2011 15:20 dennistoo wrote:
i lol-ed at PvZ.
Now while it does appear you've made a TL account just to whine about the UPness of/cheer for Protoss, one really can't help but think what an idiot you are for thinking the PvZ stats are totally unacceptable while Z's advantage in that matchup is identical to Protoss' advantage in PvT.
Furthermore these stats don't show shit, and it's worthless to engage in balance discussions - especially on TL where half the playerbase happens to be zerg.
|
On June 22 2011 12:52 darklordjac wrote:
Selects average apm 373 holy shit, I didn't even think people hit that in sc2
That isn't APM in SC2 time, it has been converted into real apm. Whether that number is actually EAPM is something I don't know.
|
|
|
On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Qxc: Loses to ThorZain and SeleCT. 8-2
26th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place.
Funny thing is, this shows nothings. Shows that Qxc met a bunch of lower level players and won against them. Would be funny for Anaheim if gretorp does the same.
Remember how it went for qxc MLG Dallas? He won 1 more game than Gretorp in the group stages, but Gretorp was in the group of death.. They both lost in their first losers round. Moonan actually got the same round as idra at MLG Dallas.
There will allways be people that get these statistics because there's a large player base, who are good enough to get into group plays, but not good enough to really do anything there, unless they have a really good weekend.
|
On June 22 2011 07:30 LagT_T wrote:Show nested quote +On June 22 2011 07:20 worldsnap wrote:On June 22 2011 07:11 SuperStyle wrote:On June 22 2011 07:02 windsupernova wrote:
Well, since the OP got it wrong here are the Win percentages
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
I wonder how many people will claim T are broken before they read the thread LOL
IMO things seem quite balanced, its a shame that the QQing never seems to stop. So thats how balance looks like, zergs having winning percentage against every other race by a high margin, good to know. However looking at every tournament around things do look balanced, but i dont see how you can say that things look balanced looking just at MLG. The only thing you can conclude from this is that given the pool of Terran and Zerg players that played at MLG, the zerg players will win more often than the terran players on average. That's the only thing you can say. The disparity in quality of players between races, especially given that there are many more terran players playing than any other race, allows you to draw no conclusion on "balance" QFT The problem with stats such as these for open tournaments such as MLG is that since anyone could sign up, there could theoretically be a tournament with 1 zerg, 254 terrans, 1 protoss. Zerg now represents less than 1% of the playing population, however say its idra in the open bracket and he makes it to the quarter finals (in a single elim bracket) without losing a game and playing only terrans, then goes 0-2 he would finish with a record of 12-2 which would make that matchup read ZvT is 85% win rate for zerg. The issue is that the general population (in my experience) doesn't understand statistics well enough to differentiate between correlation and causation. Assuming all races were perfectly balanced and had the exact same skill level at spots 1-100 of their race with the other races, in a tournament with 40% terrans, the average skill of a terran will be lower, thus pulling down their average win rate. Now I'm not arguing that the game is balanced or these stats mean nothing, but i think outside of the oooo factor, they are mostly useless unless you cut to like top 64/32 of the tournament where the nobodies/for funsies people have all been knocked out
Well, the stats for the championship bracket matches point in the same directionthe NASL racial stats.
NASL:
PvZ: 44 %
TvZ: 44%
TvP:42%
MLG (Championship bracket only)
PvZ:46.51 %
TvZ: 43.21%
TvP:46%
While you get skewed numbers in a KO system, because the better players play more games (and that way favoring the race with the most exceptionally good players), a round robin system ensures numbers that actually hint at balance, given the assumption that the skill level of the 3 races is very similar (a reasonable assumption, despite what Zergs try to tell us ).
|
I faced two gold Terrans at MLG, and a GM Protoss.
my Zvt was 100% and Zvp was 0%
I contributed to skewing the open bracket results most likely. If anyone can join the OB, it seems as if it is a silly set to look at.
I don't get this use of win % in tournaments, other than as a cute number to check out. I mean looking at championship bracket, of the top 8, 4 were korean and 2 lived/trained in Korea. IT so happened 4 of these korean trained players were Zerg. Can't really conclude much about this really. Each player has good and bad matchups, and the better players win regardless of race.
|
On June 22 2011 17:41 lololol wrote:Show nested quote +On June 22 2011 15:00 Kralic wrote:On June 22 2011 13:02 two.watup wrote:On June 22 2011 13:01 Suc wrote:On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place. Because Moonan took 1 more game off someone in a bo3. The point is players losing every single match and placing higher than Anyone else in the tournament. You cannot be top 32 and not win a single bo3. That's just pathetic tournament structure. I can lose 6 bo3s too. I'm plat league. It is called a circuit. Each stop is one event in the season, as you play in more events you will see the weaker players fall lower. Don't think of MLG events as one tournament where the past results did not matter. Give it a year or two and you will see a insane pool in the top 32. So players constantly getting top 32 placement despite losing everything and playing much worse than a lot of the people placing significantly lower for a "year or two" sounds fine to you? Saying that it's called <insert bad format name here> as an excuse for it being a bad format is not really an excuse. It's the same as saying "it's called having a bad format", except this one has a clear meaning, while yours tries to make it seem like something else.
Let me see,  2011 MLG Pro Circuit/Rankings. Looks like the top 32 is getting to be more familiar to what we are used to seeing in the top 32 of any foreigner tournament. It won't take a year or two for the terrible people to be ousted, I am saying it will be a very deep pool of skilled players in the top 32 as time progresses. It is a season, when the last event on the circuit hits the top 32 should be a pretty amazing pool of players.
New tournaments will always have weaker players ranked the highest the first few events because the amount of skilled players competing is not as high. Look at the GSL opens in the past and tell me that the top 32 were all deserving of being in the top 32(unless you think scv all ins are pro skill).
edit: anyway I am done with discussing this in this topic because it is very off topic to something that was done for fun and neat little statistics. I can always discuss it in pm's.
|
On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Qxc: Loses to ThorZain and SeleCT. 8-2
26th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place.
I honestly see nothing too crazy about this. Gret0rp is playing people like Naniwa and Slush from the very beginning, while qxc gets to roflcruise through the open bracket, getting wins vs people like Ryze, Sadistx, and TrueRedemption only to get mangled when he has to play legit players like Thorzain and SeleCt. Should Gret0rp get a pool play spot next season? Heck no. But neither should qxc, tbh. Cruising through the open bracket shouldn't be that hard for a legit "top 32" player.
|
First thing I noted was that ViBe is the only non-Korean player (or not of Korean origin, before someone starts the whole SeleCT debate again) to have 300+ average APM. =P
|
On June 22 2011 09:06 Nik0 wrote:Show nested quote +On June 22 2011 08:59 lilky wrote:so tl;dr, zerg > protoss and zerg > terran i wonder how much worse the PvZ stats would be if you disregarded MC's games... This is what should happen when 4 of the 6 best players in the tournament are zerg.
And when maps are more or less rigged to give Zerg the advantage every game...
|
On June 23 2011 01:02 aksfjh wrote:Show nested quote +On June 22 2011 09:06 Nik0 wrote:On June 22 2011 08:59 lilky wrote:so tl;dr, zerg > protoss and zerg > terran i wonder how much worse the PvZ stats would be if you disregarded MC's games... This is what should happen when 4 of the 6 best players in the tournament are zerg. And when maps are more or less rigged to give Zerg the advantage every game...
You mean "rigged to give a fair and even game amongst all players". Sorry, but no close spawns is fair, not imbalanced.
|
Are there unit stats?
I was interested to see if theres any unit in the game that was never or rarelly used. Changelings? Carriers?
|
On June 23 2011 02:14 Flight wrote:
Are there unit stats?
I was interested to see if theres any unit in the game that was never or rarelly used. Changelings? Carriers?
Changelings were used frequently. Carriers were only built in one game.
|
On June 23 2011 01:42 Stiver wrote:Show nested quote +On June 23 2011 01:02 aksfjh wrote:On June 22 2011 09:06 Nik0 wrote:On June 22 2011 08:59 lilky wrote:so tl;dr, zerg > protoss and zerg > terran i wonder how much worse the PvZ stats would be if you disregarded MC's games... This is what should happen when 4 of the 6 best players in the tournament are zerg. And when maps are more or less rigged to give Zerg the advantage every game... You mean "rigged to give a fair and even game amongst all players". Sorry, but no close spawns is fair, not imbalanced.
Seeing as how Blizzard has confirmed time and time again that the ladder maps, even with close spawns in the mix, appear relatively balanced, removing close spawns is likely to give way to imbalances. In a tournament where over half of the top 16 were Zerg, it just so happens that many of the maps used took out close spawns, something the community agrees is unfavorable to Zerg. Statistically, removing close spawns seems to give Zerg an advantage.
|
On June 23 2011 02:25 aksfjh wrote:Show nested quote +On June 23 2011 01:42 Stiver wrote:On June 23 2011 01:02 aksfjh wrote:On June 22 2011 09:06 Nik0 wrote:On June 22 2011 08:59 lilky wrote:so tl;dr, zerg > protoss and zerg > terran i wonder how much worse the PvZ stats would be if you disregarded MC's games... This is what should happen when 4 of the 6 best players in the tournament are zerg. And when maps are more or less rigged to give Zerg the advantage every game... You mean "rigged to give a fair and even game amongst all players". Sorry, but no close spawns is fair, not imbalanced. Seeing as how Blizzard has confirmed time and time again that the ladder maps, even with close spawns in the mix, appear relatively balanced, removing close spawns is likely to give way to imbalances. In a tournament where over half of the top 16 were Zerg, it just so happens that many of the maps used took out close spawns, something the community agrees is unfavorable to Zerg. Statistically, removing close spawns seems to give Zerg an advantage.
Dude, seriously, there is no logic to this at all. Removing a map, balanced or not, in no way affects the balance of the other maps.
|
the zvt stats are insane. That figure + the fact idra had won the game but it counts towards a loss :p I'm not sure on the validity of people claiming to know how many matches were used in this calculation.
|
On June 22 2011 23:54 Kralic wrote:Show nested quote +On June 22 2011 17:41 lololol wrote:On June 22 2011 15:00 Kralic wrote:On June 22 2011 13:02 two.watup wrote:On June 22 2011 13:01 Suc wrote:On June 22 2011 12:56 two.watup wrote:
I think the most impressive statistic is this:
Gret0rp: Loses every single match. 0-6 (2-12).
27th place.
Moonan: Loses every single match. 0-6 (3-12).
25th place. Because Moonan took 1 more game off someone in a bo3. The point is players losing every single match and placing higher than Anyone else in the tournament. You cannot be top 32 and not win a single bo3. That's just pathetic tournament structure. I can lose 6 bo3s too. I'm plat league. It is called a circuit. Each stop is one event in the season, as you play in more events you will see the weaker players fall lower. Don't think of MLG events as one tournament where the past results did not matter. Give it a year or two and you will see a insane pool in the top 32. So players constantly getting top 32 placement despite losing everything and playing much worse than a lot of the people placing significantly lower for a "year or two" sounds fine to you? Saying that it's called <insert bad format name here> as an excuse for it being a bad format is not really an excuse. It's the same as saying "it's called having a bad format", except this one has a clear meaning, while yours tries to make it seem like something else. Let me see, 2011 MLG Pro Circuit/Rankings. Looks like the top 32 is getting to be more familiar to what we are used to seeing in the top 32 of any foreigner tournament. It won't take a year or two for the terrible people to be ousted, I am saying it will be a very deep pool of skilled players in the top 32 as time progresses. It is a season, when the last event on the circuit hits the top 32 should be a pretty amazing pool of players. New tournaments will always have weaker players ranked the highest the first few events because the amount of skilled players competing is not as high. Look at the GSL opens in the past and tell me that the top 32 were all deserving of being in the top 32(unless you think scv all ins are pro skill). edit: anyway I am done with discussing this in this topic because it is very off topic to something that was done for fun and neat little statistics. I can always discuss it in pm's.
That had nothing to do with the tournament format of the GSL Opens, you even had to qualify all over again in all of them, there weren't any players going 0-5 and still getting top 32, while there are in MLG and the format is all that can be blamed for it.
|
On June 23 2011 02:25 aksfjh wrote:Show nested quote +On June 23 2011 01:42 Stiver wrote:On June 23 2011 01:02 aksfjh wrote:On June 22 2011 09:06 Nik0 wrote:On June 22 2011 08:59 lilky wrote:so tl;dr, zerg > protoss and zerg > terran i wonder how much worse the PvZ stats would be if you disregarded MC's games... This is what should happen when 4 of the 6 best players in the tournament are zerg. And when maps are more or less rigged to give Zerg the advantage every game... You mean "rigged to give a fair and even game amongst all players". Sorry, but no close spawns is fair, not imbalanced. Seeing as how Blizzard has confirmed time and time again that the ladder maps, even with close spawns in the mix, appear relatively balanced, removing close spawns is likely to give way to imbalances. In a tournament where over half of the top 16 were Zerg, it just so happens that many of the maps used took out close spawns, something the community agrees is unfavorable to Zerg. Statistically, removing close spawns seems to give Zerg an advantage.
Rofl no. Removing the close positions is not favoring zerg, its changing the win percentages of zerg from 20% to 50% on this map.
|
|
|
|
|
|
EPT
