|
On January 19 2011 13:11 IdrA wrote:Show nested quote +On January 19 2011 13:02 iEchoic wrote:
^^^
This is the essence of your argument:
- When zerg does disproportionately well (Zerg players reach master at a higher ratio than other races, Zerg winning 2/3 GSLs), it is because the people playing Zerg have more skill
- When zerg does disproportionately badly, it is because of imbalance, not because the T/P players have more skill
It's a really annoying and common double-standard. its not like this is some theoretical situation watch the games from the 2 gsl's zergs won. fd was head and shoulders above all his opponents, still should have lost to top and still felt the need to allin inca. he was way better than everyone, he got a relatively easy draw, and he got lucky a number of times on top of it all. nestea got 2 zvz's in ro16 and 8, not a comment on balance, and then he held off a series of retarded cheeses from boxer and mk. and has since proven hes a really good player who has a special knack for holding off cheesy stuff. not only was he better than his opponents, he won games that werent very meaningful at all in terms of balance, poorly executed undeveloped cheeses. zerg has never done proportionately well on the whole, besides qualifications or season 3 where every z dropped out in the first 2 rounds anyway. its always been a very very few z's having success and you can watch the games, they quite clearly were more skilled than their competition.
I agree, Fruit was way ahead of his opponents, no argument there. I guess I'd have to argue about Nestea. MKP only did two scv allins in the series against NesTea and won with one of them (one on Shakuras which he won, one game 7 on Metal which he lost) if I remember right.
I don't think you can say Nestea was ridiculously ahead of Foxer, whereas Fruit was way ahead of Rainbow.
|
I put together an excel sheet of all the race stats from the International SC2 TLPD.
All-time: TvZ 57% PvZ 55%
Post 1.1: TvZ 54% PvZ 56%
|
Aha, Sc2 ranks yields all information.
I think that the ~30% proportion of zergs in the upper leagues, compared to their ~20% proportion in the lower leagues, is mostly due to a large number of Terran and Protoss players who play 'casually' at the lower leagues, simply not playing enough games to get into diamond.
We can test this. If we dont see the average number of games played by Zergs being higher than the average number of Terran and Protoss games, then I'm wrong. (But if we do see this, it doesn't mean I'm right. It might just be that low tier zergs have shorter games, six pooling every time!)
Edit: Some mistake...hold on!
Edit: Corrected numbers...
Bronze:
40M games for 350k Terrans, 114 games per T.
17M games for 145k Zergs, 117 games per Z.
Silver
22M games for 115k Ts, 191 games per T.
13.2M games for 70k Zs, 188 games per Z.
Gold
17.6M games 93k Ts, 183 games per T.
12.8M games 61k Zs 210 games per Z.
Platinum
16M games for 65k Ts, 246 games per T.
13.8M games for 48k Zs, 287 games per Z.
The number are roughly the same at higher levels, and the number of players is a lot smaller, so I didn't bother adding them.
I'd appreciate someone checking the numbers independently, I'm kind of tired.
But if I got the numbers right, it's pretty clear the average zerg in the mid leages play more games than the average Terran, with roughly the same at the very bottom levels. Are they, zerg, on average, more serious? Do they like to six pool? Something else? You decide!
|
this forum is hilarious, everyone already knows that most people on TL play zerg from past polls(not going to look them up for you, i'm not lying about this, look it up yourself) and terran and toss QQ is not allowed, while shadowed zerg QQ seems to be unabashed.
well played friends, well played.
|
On January 19 2011 13:46 Dragar wrote:
Aha, Sc2 ranks yields all information.
I think that the ~30% proportion of zergs in the upper leagues, compared to their ~20% proportion in the lower leagues, is mostly due to a large number of Terran and Protoss players who play 'casually' at the lower leagues, simply not playing enough games to get into diamond.
We can test this. If we dont see the average number of games played by Zergs being higher than the average number of Terran and Protoss games, then I'm wrong. (But if we do see this, it doesn't mean I'm right. It might just be that low tier zergs have shorter games, six pooling every time!)
Here's the data:
Bronze:
40M games for 350k Terrans. Average 114 games per Terran.
17M games for 145k Zergs. Average 117 games per Zerg.
Silver
18M games for 186k Ts, 97 games per T
12.8M games for 61k Zs, 210 games per Z.
Platinum
16M games for 130k Ts, 123 games per T.
14M games for 94k Zs, 149 games per Z.
The number are roughly the same at higher levels, and the number of players is a lot smaller, so I didn't bother adding them.
I'd appreciate someone checking the numbers independently, I'm kind of tired.
But if I got the numbers right, it's pretty clear the average zerg in the lower leages play a lot more (double in Silver!) games than the average Terran. Are they, on average, more serious? Do they like to six pool? Something else? You decide!
The reason for this is pretty obvious... People who are playing starcraft for the first time will start playing the terran campaign. This will naturally cause them to gravitate to playing terran on multiplayer due to their familiarity with it. Therefore terran in particular will have a higher representation in the lower leagues and among the less frequent players.
|
On January 19 2011 13:50 jdseemoreglass wrote:
The reason for this is pretty obvious... People who are playing starcraft for the first time will start playing the terran campaign. This will naturally cause them to gravitate to playing terran on multiplayer due to their familiarity with it. Therefore terran in particular will have a higher representation in the lower leagues and among the less frequent players.
I agree that's the right explanation. I didn't think anyone would be surprised. But it's nice to see the numbers actually confirm it.
I am a little surprised at the Silver numbers. I think that may be a mistake...
|
On January 19 2011 13:49 Silidons wrote:
this forum is hilarious, everyone already knows that most people on TL play zerg from past polls(not going to look them up for you, i'm not lying about this, look it up yourself) and terran and toss QQ is not allowed, while shadowed zerg QQ seems to be unabashed.
well played friends, well played.
Does anyone know why this is (z players being way overrpresented here)?
|
Yes, playing zerg is like punching yourself in the nuts.
User was warned for this post
|
On January 19 2011 13:52 iEchoic wrote:Show nested quote +On January 19 2011 13:49 Silidons wrote:
this forum is hilarious, everyone already knows that most people on TL play zerg from past polls(not going to look them up for you, i'm not lying about this, look it up yourself) and terran and toss QQ is not allowed, while shadowed zerg QQ seems to be unabashed.
well played friends, well played. Does anyone know why this is (z players being way overrpresented here)?
probably because of IdrA/Artosis/Ret/Haypro.
IdrA probably being the main reason, the most well-known foreigner, it kinda goes hand in hand, people want to do what he does, etc.
|
On January 19 2011 13:34 iEchoic wrote:Show nested quote +On January 19 2011 13:11 IdrA wrote:On January 19 2011 13:02 iEchoic wrote:
^^^
This is the essence of your argument:
- When zerg does disproportionately well (Zerg players reach master at a higher ratio than other races, Zerg winning 2/3 GSLs), it is because the people playing Zerg have more skill
- When zerg does disproportionately badly, it is because of imbalance, not because the T/P players have more skill
It's a really annoying and common double-standard. its not like this is some theoretical situation watch the games from the 2 gsl's zergs won. fd was head and shoulders above all his opponents, still should have lost to top and still felt the need to allin inca. he was way better than everyone, he got a relatively easy draw, and he got lucky a number of times on top of it all. nestea got 2 zvz's in ro16 and 8, not a comment on balance, and then he held off a series of retarded cheeses from boxer and mk. and has since proven hes a really good player who has a special knack for holding off cheesy stuff. not only was he better than his opponents, he won games that werent very meaningful at all in terms of balance, poorly executed undeveloped cheeses. zerg has never done proportionately well on the whole, besides qualifications or season 3 where every z dropped out in the first 2 rounds anyway. its always been a very very few z's having success and you can watch the games, they quite clearly were more skilled than their competition. I agree, Fruit was way ahead of his opponents, no argument there. I guess I'd have to argue about Nestea. MKP only did two scv allins in the series against NesTea and won with one of them (one on Shakuras which he won, one game 7 on Metal which he lost) if I remember right. I don't think you can say Nestea was ridiculously ahead of Foxer, whereas Fruit was way ahead of Rainbow.
at the time mk was pure gimmick, hes gotten a bit more solid since then but every tvz win that season was 2 rax scv/rine, the weird pure mass marine stuff, or a strange 1 time build. on scrap he did an absolutely retarded fe->marine drop build and lost, then there were at least 3 scv marine rushes. think he did a tank drop on lt.
either way it was a 4-3 win in odd games by a player who has continued to perform very well. when that's your point for zerg performing disproportionately well for the first couple of seasons your argument has serious issues.
|
On January 19 2011 13:52 iEchoic wrote:Show nested quote +On January 19 2011 13:49 Silidons wrote:
this forum is hilarious, everyone already knows that most people on TL play zerg from past polls(not going to look them up for you, i'm not lying about this, look it up yourself) and terran and toss QQ is not allowed, while shadowed zerg QQ seems to be unabashed.
well played friends, well played. Does anyone know why this is (z players being way overrpresented here)?
I can only offer tentative theories...
Maybe people who know the history of Starcraft recognize the power of macro-oriented play, and therefore favor the macro-oriented race.
Maybe people who take the game more seriously tend to favor the strategy and skills necessary in playing the reactive/adaptive race as opposed to playing the aggressive races.
Maybe playing reactively/defensively requires more external understanding of counters/metagame/etc. and therefore people who search the internet for the latest strategies and trends end up joining TL in greater numbers.
It is likely a combination of many such factors.
|
Are Z players really more common on TL forums? Or is it roughly the 30% I'd expect from the masters/diamond league distribution?
|
On January 19 2011 08:57 Silidons wrote:
but yet zerg has won 2 GSL's, which if you wanted to win any tournament, you would want to win that one.
sorry but zerg UP isn't going to work, GSL has the best players in the world.
I'll quote IdrA to respond to your comment
...
ive really never understood this
how can someone post something thats flat out wrong with such absolute certainty? what is broken in your brain to make that possible?
Zerg won the 2 GSL's a LONG time ago. The metagame HAS changed (if you even know what that means) , as well as patches that hurt zerg more such as phoenix buffs. Those zerg players are extremely good, and are some of the few zerg in Code S at the moment.. they may as well have won with terran or toss at that time because they are so skilled.
I absolutely hate it when people use this point in balance arguments.. the GSLs where Fruitdealer and Nestea won were when people played differently, terran and toss now know how to deal with zerg and as well know how to abuse zerg more.
|
On January 19 2011 13:58 IdrA wrote:Show nested quote +On January 19 2011 13:34 iEchoic wrote:On January 19 2011 13:11 IdrA wrote:On January 19 2011 13:02 iEchoic wrote:
^^^
This is the essence of your argument:
- When zerg does disproportionately well (Zerg players reach master at a higher ratio than other races, Zerg winning 2/3 GSLs), it is because the people playing Zerg have more skill
- When zerg does disproportionately badly, it is because of imbalance, not because the T/P players have more skill
It's a really annoying and common double-standard. its not like this is some theoretical situation watch the games from the 2 gsl's zergs won. fd was head and shoulders above all his opponents, still should have lost to top and still felt the need to allin inca. he was way better than everyone, he got a relatively easy draw, and he got lucky a number of times on top of it all. nestea got 2 zvz's in ro16 and 8, not a comment on balance, and then he held off a series of retarded cheeses from boxer and mk. and has since proven hes a really good player who has a special knack for holding off cheesy stuff. not only was he better than his opponents, he won games that werent very meaningful at all in terms of balance, poorly executed undeveloped cheeses. zerg has never done proportionately well on the whole, besides qualifications or season 3 where every z dropped out in the first 2 rounds anyway. its always been a very very few z's having success and you can watch the games, they quite clearly were more skilled than their competition. I agree, Fruit was way ahead of his opponents, no argument there. I guess I'd have to argue about Nestea. MKP only did two scv allins in the series against NesTea and won with one of them (one on Shakuras which he won, one game 7 on Metal which he lost) if I remember right. I don't think you can say Nestea was ridiculously ahead of Foxer, whereas Fruit was way ahead of Rainbow. at the time mk was pure gimmick, hes gotten a bit more solid since then but every tvz win that season was 2 rax scv/rine, the weird pure mass marine stuff, or a strange 1 time build. on scrap he did an absolutely retarded fe->marine drop build and lost, then there were at least 3 scv marine rushes. think he did a tank drop on lt. either way it was a 4-3 win in odd games by a player who has continued to perform very well. when that's your point for zerg performing disproportionately well for the first couple of seasons your argument has serious issues.
omg did IdrA just post in my thread...
|
Australia8532 Posts
On January 19 2011 14:04 EternaL_9 wrote:Show nested quote +On January 19 2011 08:57 Silidons wrote:
but yet zerg has won 2 GSL's, which if you wanted to win any tournament, you would want to win that one.
sorry but zerg UP isn't going to work, GSL has the best players in the world. I'll quote IdrA to respond to your comment Show nested quote +
...
ive really never understood this
how can someone post something thats flat out wrong with such absolute certainty? what is broken in your brain to make that possible?
Zerg won the 2 GSL's a LONG time ago. The metagame HAS changed (if you even know what that means) , as well as patches that hurt zerg more such as phoenix buffs. Those zerg players are extremely good, and are some of the few zerg in Code S at the moment.. they may as well have won with terran or toss at that time because they are so skilled. I absolutely hate it when people use this point in balance arguments.. the GSLs where Fruitdealer and Nestea won were when people played differently, terran and toss now know how to deal with zerg and as well know how to abuse zerg more.
Wow talk about selective bias - what about that patch fixes that buffed zerg? But no you conveniently leave those out yeah?
they may as well have won with terran or toss at that time because they are so skille
A ridiculous statement in itself; Nestea vs Foxer was a very close series
terran and toss now know how to deal with zerg and as well know how to abuse zerg more.
A zerg won a GSL before and after the reaper nerf.. but zerg abuse is at an all time high? Terran and toss had this epiphany on how to bully zerg? No. Zergs decided to hatch first, they got punished. Now zergs pool first and then early hatch and are holding it off more convincingly.. it is all a part of the game developing .. please - no more zerg tears
|
With what I can remember of stats, and wikipedia, here is the likelihood of these results:
All-time:
TvZ 57% n=1129 likelihood=0.001%
PvZ 55% n=634 likelihood=0.692%
PvT 48% n=1522 likelihood=6.496%
Post 1.1:
TvZ 54% n=779 likelihood=1.987%
PvZ 56% n=446 likelihood=1.381%
PvT 48% n=937 likelihood=20.264%
Updated spreadsheet with these calculations found here
|
On January 19 2011 14:17 Wakamex wrote:With what I can remember of stats, and wikipedia, here is the likelihood of these results: All-time: TvZ 57% n=1129 likelihood=0.001% PvZ 55% n=634 likelihood=0.692% PvT 48% n=1522 likelihood=6.496% Post 1.1: TvZ 54% n=779 likelihood=1.987% PvZ 56% n=446 likelihood=1.381% PvT 48% n=937 likelihood=20.264% Updated spreadsheet with these calculations found here
That wikipedia article is a little complicated and seems to focus on 50/50 probability rather than 33/33/33... could you give a simple explanation of how you came up with those "likelihood" numbers?
|
On January 19 2011 09:06 haffy wrote:Show nested quote +On January 19 2011 08:57 Silidons wrote:
but yet zerg has won 2 GSL's, which if you wanted to win any tournament, you would want to win that one.
sorry but zerg UP isn't going to work, GSL has the best players in the world. Yeah well done. When presented with a larger number of statistics you chose to use a smaller sample to discredit them. When using numbers you always use the larger sample size because it is less likely to be effected by variance. Also I play Zerg and I don't think they're UP. But I think your arguement is pretty badly flawed.
well the largest number of statistics includes all ladder playing, even bronze
do you want to use that?
|
On January 19 2011 14:24 travis wrote:Show nested quote +On January 19 2011 09:06 haffy wrote:On January 19 2011 08:57 Silidons wrote:
but yet zerg has won 2 GSL's, which if you wanted to win any tournament, you would want to win that one.
sorry but zerg UP isn't going to work, GSL has the best players in the world. Yeah well done. When presented with a larger number of statistics you chose to use a smaller sample to discredit them. When using numbers you always use the larger sample size because it is less likely to be effected by variance. Also I play Zerg and I don't think they're UP. But I think your arguement is pretty badly flawed. well the largest number of statistics includes all ladder playing, even bronze do you want to use that?
But bronze is irrelevant for obvious reasons for an indication of winrates at the tip level, whereas top level international tournaments aren't. You must have knew that before you posted >_>;; Why post this?
|
On January 19 2011 14:23 jdseemoreglass wrote:
That wikipedia article is a little complicated and seems to focus on 50/50 probability rather than 33/33/33... could you give a simple explanation of how you came up with those "likelihood" numbers?
Each matchup can be expected to be a 50/50 probability, no? So I calculate standard error (se) as 1/2sqrt(n), observed error (e) as W%-0.5, your experiment's z-value (x) as e/se, and your confidence level P(z < x < z) with excel's 2*(normsdist(x)-0.5).
All the steps are in the excel sheet if you want to plug in numbers.
set x=1, and you get a confidence level of 68.269%, as per wikipedia. set w% to 50% and you get likelihood of 100%. you can also tweak n. for example, the 57% TvZ result becomes "likely" with n=26.
|
|
|
|
|
|
EPT
