|
On May 04 2012 03:07 zmansman17 wrote:
A few explanations of the win rates: Many of the weaker Terran players have left the race entirely and switched to Protoss or Zerg. The evidence of this is in the proportion of Terran players to their Zerg and Protoss counterparts in the GM and Master leagues within NA and EU. Terran players have dropped to 1/4 of the race selections. (In KR, the Korean Terrans are on another level, and yet their representation is only equal).
Note that it is a possibility that the win rate of the remaining Terrans never changed, but the overall win rate of Terran simply increased from the reduction of the weaker Terran players.
Thus, this is one explanation of relatively even win rates TvP.
An alternative explanation of these win rates is that Terrans are opting for more All-in or Semi all-in mid game timings. Most good Terran players recognize that our Win % peters out as the game goes on, and so we go for timings when we have the greatest chance to win.
I agree with your explanation in bold, or put another way, it is just easier to win right now with p and z and therefore there are more of them in the higher leagues.
On May 04 2012 09:26 Sakagami wrote:
Lol quite honestly the korean PvT winrate is completely skewed. The korean sample is like of 150 games which is nothing, and the entire winrate in general is probably skewed by STartale_Parting who has a 80% win rate in the match up.
Parting alone can eaisly skew the winrate 5-6%
As could one of the better terrans potentially skew it in their direction. What is that number contained a bunch of MVP and MMA games and no Parting games? Then in reality it would be even more skewed to toss if Parting was included. The bottom line is we don't know.
I would just like to point out that the record in the GSL Ro8 PvT was 6 - 0. Parting was not involved (he was playing PvP). Parting is not the reason PvT is strong.
|
On May 04 2012 09:37 NHL Fever wrote:
I would just like to point out that the record in the GSL Ro8 PvT was 6 - 0. Parting was not involved (he was playing PvP). Parting is not the reason PvT is strong.
There were 3 PvTs
So it was 7-3...
|
On May 04 2012 07:32 SkimGuy wrote:Show nested quote +On May 04 2012 06:37 Seam wrote:On May 04 2012 06:29 blade55555 wrote:On May 04 2012 06:26 Seam wrote:On May 04 2012 06:11 Sajaki wrote:On May 04 2012 05:59 ACrow wrote:On May 04 2012 05:53 Seam wrote:On May 04 2012 05:45 Sajaki wrote:
YAY! Terran winrates are absolute garbage. Now maybe blizzard will actually pay attention to our race when it comes to any game past 13-14 minutes! :D Are we looking at the same graph? let me translate: "Terran winrates are absolute garbage" = Terran winrates are below 50% for the first time evar Terran winrates are garbage = terran dropping from 62-46 and 58-44. That is a BIG change. Of course the Korean rates matter the most because metagame changes usually happen first in korea, and blizzard is balancing the games for the top level of play i.e. korean server. I would assume the korean section of the graph would actually matter the least.Mainly due to having a very, very small sample size. I disagree. You have to realize for the internation graph that it counts games such as stephano vs no named joe. Korean graph is almost purely GSL with code A/S players. International counts tournaments even if the top foreigners don't play in the smaller tournaments. So hard to take the international seriously at all since it goes from super good player thorzain vs no named xxx player who is mid masters. While it's true that it does show only GSL level players, you also have to consider not all players are at the same level, even in GSL. With a sample size that's really small, there's a lot of viables. Or, for example, take the MVP vs Naniwa games. MVP beat Naniwa due to playing better, and out controlling Naniwa.(And 2 cheeses) Does that show anything about balance? 2 of the 4 games were cheese.Overall, I'm saying we shouldn't use this in any way, shape, or form as a means to base aruguments of balance on. It's good to bring up, yes, but with small sample sizes it can't be trusted as an accurate scale of balance. (I'm tired and having trouble wording things) And the better player won. What's your point?
My point is it doesn't show anything about balance.
|
parting and squirtle rippin stuff UP
my biggest issue with korean zvp is the reaction to a 3rd base. they take a 4th and go for a bust, when you can max out on roaches more efficiently without a 4th hatch
also lol at the terran outcry; only .5% over 50% in international plus korea! what an outrage!
|
On May 04 2012 10:38 BenAD wrote:Show nested quote +On May 04 2012 09:37 NHL Fever wrote:
I would just like to point out that the record in the GSL Ro8 PvT was 6 - 0. Parting was not involved (he was playing PvP). Parting is not the reason PvT is strong. There were 3 PvTs So it was 7-3...
True enough, I stand corrected. Of Korean PvT's, 6 - 0.
On May 04 2012 11:10 Alejandrisha wrote:
parting and squirtle rippin stuff UP
my biggest issue with korean zvp is the reaction to a 3rd base. they take a 4th and go for a bust, when you can max out on roaches more efficiently without a 4th hatch
also lol at the terran outcry; only .5% over 50% in international plus korea! what an outrage!
International + Korea basically = International because of the numbers involved.
|
Judging from other threads, I'd like to present the following hypothesis:
A more balanced game does not lead to less balance whine. In fact, the amount of balance whine has little correlation to the current balance.
|
Parting alone can eaisly skew the winrate 5-6%
Same can be said for MVP/MMA when Terrans had higher win rates. Are players skewing the data? Or is it the affects of both player skill and metagame/balance?
|
On May 04 2012 13:14 SarcasmMonster wrote:
Judging from other threads, I'd like to present the following hypothesis:
A more balanced game does not lead to less balance whine. In fact, the amount of balance whine has little correlation to the current balance.
I second this. People doesn't really care about balance as much as they care about themselves winning
|
4713 Posts
Another reason to look forward to the KeSPA and OGN SC2 team and individual leagues. It will be harder for people to dismiss the korean stats, despite them being more relevant skill wise, when the sample size doubles or triples.
I can't trust international tournament results when they contain games from people playing way bellow the biggest skill level and especially when the results get skewed by playhem tournaments where low league nobodies are destroyed by top level players.
If you assume the korean and international sample sizes contain nearly equal skill level from players, then yes I'd trust the international stats more, but as it stands, the international stats are just as polluted and misleading as the small sample size from Korea.
|
I'll stick with Terran even if the winrate goes to 20% as long as Flash plays Terran in SC2.
I believe <3
|
On May 04 2012 17:58 Destructicon wrote:
Another reason to look forward to the KeSPA and OGN SC2 team and individual leagues. It will be harder for people to dismiss the korean stats, despite them being more relevant skill wise, when the sample size doubles or triples.
I can't trust international tournament results when they contain games from people playing way bellow the biggest skill level and especially when the results get skewed by playhem tournaments where low league nobodies are destroyed by top level players.
If you assume the korean and international sample sizes contain nearly equal skill level from players, then yes I'd trust the international stats more, but as it stands, the international stats are just as polluted and misleading as the small sample size from Korea. With respect to MarineKing and other professional gamers Korean player, even a "decent" tournament player is a low league nobody.
If the top player is, say, a terran, then Terran will look stronger, just because that top player can have 10-2 against Zerg or Protoss due to his skill. And if the same player happens to make a lot of mirrors, then the TvZ or TvP ratio will drop.
This is not due to metagame changes, much less balance issues, it's just that the best of the best play a specific race.
If you have more players, then things will be more even: each race will have its best players destroying those below them. So you need more players to have relevant information. You will always have better players destroying lower players, even if you only take the world top 30.
|
On May 04 2012 09:26 Sakagami wrote:
Lol quite honestly the korean PvT winrate is completely skewed. The korean sample is like of 150 games which is nothing, and the entire winrate in general is probably skewed by STartale_Parting who has a 80% win rate in the match up.
Parting alone can eaisly skew the winrate 5-6%
Oh yeah, parting alone skewed it by 5-6%
Parting has 9 recorded PvTs in the korean TLPD for the month and he lost 5 of them. If there's anyone skewing Blizzard's view on balance it's the zombie russian/UA terran ladder heroes.
|
On May 04 2012 18:04 xrapture wrote:
I'll stick with Terran even if the winrate goes to 20% as long as Flash plays Terran in SC2.
I believe <3
Note that this month, Terran is back on top with about 55% win ratio and slightly ahead in both non-mirror match-ups  I don't think you'll ever have to worry about Terran- of all races- dropping down that low lol. It's always been much more stable than the other two.
But quite frankly, all 3 non-mirror match-ups have insanely close winrates. Happy for this 
|
On May 04 2012 19:16 DarkPlasmaBall wrote:Show nested quote +On May 04 2012 18:04 xrapture wrote:
I'll stick with Terran even if the winrate goes to 20% as long as Flash plays Terran in SC2.
I believe <3 Note that this month, Terran is back on top with about 55% win ratio and slightly ahead in both non-mirror match-ups I don't think you'll ever have to worry about Terran- of all races- dropping down that low lol. It's always been much more stable than the other two. But quite frankly, all 3 non-mirror match-ups have insanely close winrates. Happy for this
Close win rates don't imply a good game, though.
People are upset with T being completely dependent on bio and having all ins like 2 rax bunker, proxy rax, etc. but fail to understand that is because the game has DESIGNED to REWARD terrans that "all in" and win EARLY to MID game and PUNISHES any terran that turtles and plays macro without pressure.
It's like designing Protoss to only be able to win with cannon rush, and so long as P wins half of the time against everybody else, the game is "balanced".
But it's not balanced in any non-shallow meaning of the word.
|
On May 04 2012 19:33 adrenaLinG wrote:Show nested quote +On May 04 2012 19:16 DarkPlasmaBall wrote:On May 04 2012 18:04 xrapture wrote:
I'll stick with Terran even if the winrate goes to 20% as long as Flash plays Terran in SC2.
I believe <3 Note that this month, Terran is back on top with about 55% win ratio and slightly ahead in both non-mirror match-ups I don't think you'll ever have to worry about Terran- of all races- dropping down that low lol. It's always been much more stable than the other two. But quite frankly, all 3 non-mirror match-ups have insanely close winrates. Happy for this Close win rates don't imply a good game, though. People are upset with T being completely dependent on bio and having all ins like 2 rax bunker, proxy rax, etc. but fail to understand that is because the game has DESIGNED to REWARD terrans that "all in" and win EARLY to MID game and PUNISHES any terran that turtles and plays macro without pressure. It's like designing Protoss to only be able to win with cannon rush, and so long as P wins half of the time against everybody else, the game is "balanced". But it's not balanced in any non-shallow meaning of the word.
I agree with you that there exists a difference between merely having balanced win rates, and having a good inherent game design that allows for both success and variation in all of the early, mid, and late games and innovation of multiple unit compositions across all three races. I've also agreed with the existence of the shortcomings in these areas.
That being said, the TLPD winrates don't really mention them, so I didn't talk about them in this thread (I talked about them in other, relevant threads). But yes
|
On May 04 2012 11:44 NHL Fever wrote:Show nested quote +On May 04 2012 10:38 BenAD wrote:On May 04 2012 09:37 NHL Fever wrote:
I would just like to point out that the record in the GSL Ro8 PvT was 6 - 0. Parting was not involved (he was playing PvP). Parting is not the reason PvT is strong. There were 3 PvTs So it was 7-3... True enough, I stand corrected. Of Korean PvT's, 6 - 0.
In which the terrans played absolutely awful and completely deserved the result.
|
Korean terrans being overtaken by zerg and toss? Can anyone Explain why? And why On international T takes the lead, Way back I've read that Foreign Zerg/Protoss are better than their Terran? Im lost hehe
|
On May 04 2012 19:55 Ace1123 wrote:
Korean terrans being overtaken by zerg and toss? Can anyone Explain why? And why On international T takes the lead, Way back I've read that Foreign Zerg/Protoss are better than their Terran? Im lost hehe Error bars overlap, so it's just the graph that is not precise enough. There are only a few % difference, so we don't have enough games to reliably tell which race has the lead.
|
yeah TvP and TvZ kinda hell right now lategame...
|
International stats need to remove the first few rounds where pros don't meet pros because it adds a lot of noise. A pro from any race will beat a no-name from any race thus averaging the balance towards 50%-50%.
For every round in a single elimination you'll have twice as much games than the next round. If you remove half the rounds of tournaments where pros meet no-names you remove half the noise. How many rounds do pros play in playhem, go4sc2, zotac before they meet another pro ? The International stats are filled with such tournaments full of useless data. Korean stats have a smaller sample size but they have very little noise too. Without noise, the International stats could be further away from 50%-50% too.
|
|
|
|
|
|
EPT
