|
On October 14 2014 05:50 Tuczniak wrote:Show nested quote +On October 14 2014 05:44 Ghanburighan wrote:On October 14 2014 05:41 Thieving Magpie wrote:On October 14 2014 05:25 Tuczniak wrote:On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you! Any system with 3 races with an imbalance would look like this with this method. This is not a proof of anything. Not that I disagree with your claim though. I guess I'm too stupid to see what you mean. How can a race be too powerful if only the best players can use it? I think the argument is that there will be more top players for that race. I meant that Top100 will always have winrates closer to 50% than Top50. And that's because in Top100 there is a lot more games with big skill difference between players for example TLO vs Innovation. No matter the balance state, this game will most likely end up with Innovation winning. Those games are more common in top100 and naturally will push winrates towards 50%. PS: sorry TLO
Please elaborate on this mechanism.
So, let's assume that the top 50 is actually top 3 - one from each race. And the top 100 is top 6. Also, the bottom 3 always lose to the top 3. Now, wouldn't the winrates for top 3 be the same as for top 6?
|
On October 14 2014 05:50 Tuczniak wrote:Show nested quote +On October 14 2014 05:44 Ghanburighan wrote:On October 14 2014 05:41 Thieving Magpie wrote:On October 14 2014 05:25 Tuczniak wrote:On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you! Any system with 3 races with an imbalance would look like this with this method. This is not a proof of anything. Not that I disagree with your claim though. I guess I'm too stupid to see what you mean. How can a race be too powerful if only the best players can use it? I think the argument is that there will be more top players for that race. I meant that Top100 will always have winrates closer to 50% than Top50. And that's because in Top100 there is a lot more games with big skill difference between players for example TLO vs Innovation. No matter the balance state, this game will most likely end up with Innovation winning. Those games are more common in top100 and naturally will push winrates towards 50%. PS: sorry TLO
If over infinite games the matchup is 50/50, then doesn't that mean it's balanced?
|
Czech Republic12134 Posts
On October 14 2014 03:52 DinoMight wrote:Show nested quote +On October 14 2014 03:44 antiRW wrote:On October 14 2014 03:14 Ghanburighan wrote:On October 14 2014 03:11 antiRW wrote:On October 14 2014 02:49 Ghanburighan wrote:
This is a nice new data point. Although it's clear why the top 50 gives a weird result when you compare it to top100. It has people like Targa, Jaedong, Leenock, etc over people like Hydra, Dark, Curious.
I do dislike that both lists bring in Mvp, Kas and some guy called EnDerr. Perhaps it's possible to refine the search a bit more. Easy to refine in principle. For starters I just took the aligulac top X. Could be something like current GSL players or some mixed group instead. Agreed, players like Kas and EnDerr are why I personally prefer the top 50 over the top 100. It cannot be reduced much below 50 if we want to have a decent number of games (and thus valid statistics). Only GSL players sounds quite viable on paper. Does that yield anything different from Top50 or GSL games only? Players from last GSL Code S: 'Stats', 'Flash', 'Hurricane', 'Reality', 'Stork', 'Maru', 'Rain', 'Avenge', 'EffOrt', 'PartinG', 'DongRaeGu', 'ParalyzE', 'Soulkey', 'Zest', 'TY', 'TRUE', 'soO', 'Dark', 'Rogue', 'Solar', 'Shine', 'Trap', 'INnoVation', 'Bbyong', 'MyuNgSiK', 'Classic', 'Hush', 'sOs', 'Cure', 'Terminator', 'Dear', 'Trust' 69 meetings between these players (not just GSL, but all aligulac games since Sep 01, 2014) Total games: 204 PvT - 43.48% - 46 PvZ - 52.54% - 59 TvZ - 54.55% - 44 31 PvP, 12 TvT, 12 ZvZ => Similar directions in all three match-ups as with the Top 50/100 (obviously there is some player overlap as well). If you look only at Bo5 series and up, Terran becomes hugely favored both in GSL and outside vs Z.
In current map pool? Who would have guessed  I think the current map pool heavily favors Terran, since we have 7 maps in map pool, in BO5 you can veto just one, in BO7 none. So we are back to the blinking seasons, where the map pool was heavily favoring Protoss blinking race.
|
On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you!
Thanks.
It does not quite support that but something very similar: In TvZ, Terran has a huge amount of micro-ability at the top allowing the absolute best players to utilise their units way better than those in the top 100-1000. With build refinement and army positioning there is only so much you can do. Innovation's builds are not insanely betters than those of a top 100 Terran, but his execution and micro is.
The problem is, how do you balance now? Balance for the top 10 or the top 100 or the top 1000? Right now TvZ appears imbalanced at all these.
p.s.: Can we just leave ironic tone out of it? Let just have a civilised discussion about the state of the game. If it helps, I don't have a stake in the current debate - I play toss.
|
On October 14 2014 06:06 Ghanburighan wrote:Show nested quote +On October 14 2014 05:50 Tuczniak wrote:On October 14 2014 05:44 Ghanburighan wrote:On October 14 2014 05:41 Thieving Magpie wrote:On October 14 2014 05:25 Tuczniak wrote:On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you! Any system with 3 races with an imbalance would look like this with this method. This is not a proof of anything. Not that I disagree with your claim though. I guess I'm too stupid to see what you mean. How can a race be too powerful if only the best players can use it? I think the argument is that there will be more top players for that race. I meant that Top100 will always have winrates closer to 50% than Top50. And that's because in Top100 there is a lot more games with big skill difference between players for example TLO vs Innovation. No matter the balance state, this game will most likely end up with Innovation winning. Those games are more common in top100 and naturally will push winrates towards 50%. PS: sorry TLO Please elaborate on this mechanism. So, let's assume that the top 50 is actually top 3 - one from each race. And the top 100 is top 6. Also, the bottom 3 always lose to the top 3. Now, wouldn't the winrates for top 3 be the same as for top 6? If there is no imbalance the winrates will be the same. But assume imbalance, same skilled players have winrate zvt 40%, and as you said top4-6 will always lose to top1-3.
1.So for top3 we have winrates 40% (same skill, assumed imbalance).
2.For top top6 we would have 4 situations.
a) zerg1 (from top3) meets terran1 (from top3) -> 40%
b) zerg1 (from top3) meets terran2 (from top4-6) -> 100%
c) zerg2 (from top4-6) meets terran1(from top3) -> 0%
d) zerg2 (from top4-6) meets terran2(from top4-6) -> 40%
Assuming they play each other the same amount of times we get 45% winrate ((100+40+40+0)/4)
For what Thieving Magpie wanted, if terrans have better winrates at higher levels it would be better to compare (top50 vs top51-100) not (top50 vs top100).
|
On October 14 2014 06:25 Tuczniak wrote:Show nested quote +On October 14 2014 06:06 Ghanburighan wrote:On October 14 2014 05:50 Tuczniak wrote:On October 14 2014 05:44 Ghanburighan wrote:On October 14 2014 05:41 Thieving Magpie wrote:On October 14 2014 05:25 Tuczniak wrote:On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you! Any system with 3 races with an imbalance would look like this with this method. This is not a proof of anything. Not that I disagree with your claim though. I guess I'm too stupid to see what you mean. How can a race be too powerful if only the best players can use it? I think the argument is that there will be more top players for that race. I meant that Top100 will always have winrates closer to 50% than Top50. And that's because in Top100 there is a lot more games with big skill difference between players for example TLO vs Innovation. No matter the balance state, this game will most likely end up with Innovation winning. Those games are more common in top100 and naturally will push winrates towards 50%. PS: sorry TLO Please elaborate on this mechanism. So, let's assume that the top 50 is actually top 3 - one from each race. And the top 100 is top 6. Also, the bottom 3 always lose to the top 3. Now, wouldn't the winrates for top 3 be the same as for top 6? If there is no imbalance the winrates will be the same. But assume imbalance, same skilled players have winrate zvt 40%, and as you said top4-6 will always lose to top1-3. 1.So for top3 we have winrates 40% (same skill, assumed imbalance). 2.For top top6 we would have 4 situations. a) zerg1 (from top3) meets terran1 (from top3) -> 40% b) zerg1 (from top3) meets terran2 (from top4-6) -> 100% c) zerg2 (from top4-6) meets terran1(from top3) -> 0% d) zerg2 (from top4-6) meets terran2(from top4-6) -> 40% Assuming they play each other the same amount of times we get 45% winrate ((100+40+40+0)/4) For what Thieving Magpie wanted, if terrans have better winrates at higher levels it would be better to compare (top50 vs top51-100) not (top50 vs top100).
Top 51-100 active players: 'MajOr', 'Trap', 'Jim', 'Rogue', 'Lilbow', 'Dark', 'Hydra', 'MC', 'RagnaroK', 'TRUE', 'Impact', 'Center', 'Nerchio', 'Reality', 'KangHo', 'Golden', 'Petraeus', 'ShoWTimE', 'HeRoMaRinE', 'Pigbaby', 'Journey', 'Sen', 'EffOrt', 'MaNa', 'SuperNova', 'MaSa', 'Symbol', 'Avenge', 'Dayshi', 'Kas', 'Trust', 'Ryung', 'Hack', 'Check', 'HerO', 'EnDerr', 'Curious', 'Stork', 'FanTaSy', 'Starbuck', 'uThermal', 'Revival', 'Squirtle', 'ParalyzE', 'TLO', 'Spear', 'MorroW', 'HuK', 'Sorry', 'Bly', 'KeeN', 'PiG', 'YongHwa', 'sKyHigh', 'Armani', 'MarineLorD', 'Alicia', 'Creator', 'MyuNgSiK', 'Welmu', 'Kane', 'Daisy', 'iaguz', 'Action', 'MarineKing', 'Strelok', 'JonnyREcco', 'Dream', 'Shine', 'TooDming', 'CranK', 'Apocalypse', 'HasuObs', 'BBoongBBoong', 'Serral', 'Bunny', 'Seed', 'eMotion', 'XiGua', 'aLive', 'Sleep', 'Has', 'Suppy', 'Ret', 'Billowy', 'Elazer', 'Harstem', 'NAKSEO', 'elfi', 'Grubby', 'Oz', 'puCK', 'Blysk', 'Zanster', 'LiveZerg', 'SortOf', 'MacSed', 'fraer', 'Pezz', 'JYP'
370 matches between these players since 2014-09-01
Total series: 370
PvT - 44.58% - 83
PvZ - 46.79% - 109
TvZ - 49.38% - 81
Total games: 932
PvT - 47.39% - 211
PvZ - 49.26% - 270
TvZ - 50.24% - 211
|
On October 14 2014 06:25 Tuczniak wrote:Show nested quote +On October 14 2014 06:06 Ghanburighan wrote:On October 14 2014 05:50 Tuczniak wrote:On October 14 2014 05:44 Ghanburighan wrote:On October 14 2014 05:41 Thieving Magpie wrote:On October 14 2014 05:25 Tuczniak wrote:On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you! Any system with 3 races with an imbalance would look like this with this method. This is not a proof of anything. Not that I disagree with your claim though. I guess I'm too stupid to see what you mean. How can a race be too powerful if only the best players can use it? I think the argument is that there will be more top players for that race. I meant that Top100 will always have winrates closer to 50% than Top50. And that's because in Top100 there is a lot more games with big skill difference between players for example TLO vs Innovation. No matter the balance state, this game will most likely end up with Innovation winning. Those games are more common in top100 and naturally will push winrates towards 50%. PS: sorry TLO Please elaborate on this mechanism. So, let's assume that the top 50 is actually top 3 - one from each race. And the top 100 is top 6. Also, the bottom 3 always lose to the top 3. Now, wouldn't the winrates for top 3 be the same as for top 6? If there is no imbalance the winrates will be the same. But assume imbalance, same skilled players have winrate zvt 40%, and as you said top4-6 will always lose to top1-3. 1.So for top3 we have winrates 40% (same skill, assumed imbalance). 2.For top top6 we would have 4 situations. a) zerg1 (from top3) meets terran1 (from top3) -> 40% b) zerg1 (from top3) meets terran2 (from top4-6) -> 100% c) zerg2 (from top4-6) meets terran1(from top3) -> 0% d) zerg2 (from top4-6) meets terran2(from top4-6) -> 40% Assuming they play each other the same amount of times we get 45% winrate ((100+40+40+0)/4) For what Thieving Magpie wanted, if terrans have better winrates at higher levels it would be better to compare (top50 vs top51-100) not (top50 vs top100).
Yeah, you're right.
The easiest way to show this is to just assume that a top 3 T has a 100% winrate against Z. If we add a top 6 T who always loses to a top 3 player, the winrate goes down.
|
On October 14 2014 06:36 antiRW wrote:Show nested quote +On October 14 2014 06:25 Tuczniak wrote:On October 14 2014 06:06 Ghanburighan wrote:On October 14 2014 05:50 Tuczniak wrote:On October 14 2014 05:44 Ghanburighan wrote:On October 14 2014 05:41 Thieving Magpie wrote:On October 14 2014 05:25 Tuczniak wrote:On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you! Any system with 3 races with an imbalance would look like this with this method. This is not a proof of anything. Not that I disagree with your claim though. I guess I'm too stupid to see what you mean. How can a race be too powerful if only the best players can use it? I think the argument is that there will be more top players for that race. I meant that Top100 will always have winrates closer to 50% than Top50. And that's because in Top100 there is a lot more games with big skill difference between players for example TLO vs Innovation. No matter the balance state, this game will most likely end up with Innovation winning. Those games are more common in top100 and naturally will push winrates towards 50%. PS: sorry TLO Please elaborate on this mechanism. So, let's assume that the top 50 is actually top 3 - one from each race. And the top 100 is top 6. Also, the bottom 3 always lose to the top 3. Now, wouldn't the winrates for top 3 be the same as for top 6? If there is no imbalance the winrates will be the same. But assume imbalance, same skilled players have winrate zvt 40%, and as you said top4-6 will always lose to top1-3. 1.So for top3 we have winrates 40% (same skill, assumed imbalance). 2.For top top6 we would have 4 situations. a) zerg1 (from top3) meets terran1 (from top3) -> 40% b) zerg1 (from top3) meets terran2 (from top4-6) -> 100% c) zerg2 (from top4-6) meets terran1(from top3) -> 0% d) zerg2 (from top4-6) meets terran2(from top4-6) -> 40% Assuming they play each other the same amount of times we get 45% winrate ((100+40+40+0)/4) For what Thieving Magpie wanted, if terrans have better winrates at higher levels it would be better to compare (top50 vs top51-100) not (top50 vs top100). Top 51-100 active players: 'MajOr', 'Trap', 'Jim', 'Rogue', 'Lilbow', 'Dark', 'Hydra', 'MC', 'RagnaroK', 'TRUE', 'Impact', 'Center', 'Nerchio', 'Reality', 'KangHo', 'Golden', 'Petraeus', 'ShoWTimE', 'HeRoMaRinE', 'Pigbaby', 'Journey', 'Sen', 'EffOrt', 'MaNa', 'SuperNova', 'MaSa', 'Symbol', 'Avenge', 'Dayshi', 'Kas', 'Trust', 'Ryung', 'Hack', 'Check', 'HerO', 'EnDerr', 'Curious', 'Stork', 'FanTaSy', 'Starbuck', 'uThermal', 'Revival', 'Squirtle', 'ParalyzE', 'TLO', 'Spear', 'MorroW', 'HuK', 'Sorry', 'Bly', 'KeeN', 'PiG', 'YongHwa', 'sKyHigh', 'Armani', 'MarineLorD', 'Alicia', 'Creator', 'MyuNgSiK', 'Welmu', 'Kane', 'Daisy', 'iaguz', 'Action', 'MarineKing', 'Strelok', 'JonnyREcco', 'Dream', 'Shine', 'TooDming', 'CranK', 'Apocalypse', 'HasuObs', 'BBoongBBoong', 'Serral', 'Bunny', 'Seed', 'eMotion', 'XiGua', 'aLive', 'Sleep', 'Has', 'Suppy', 'Ret', 'Billowy', 'Elazer', 'Harstem', 'NAKSEO', 'elfi', 'Grubby', 'Oz', 'puCK', 'Blysk', 'Zanster', 'LiveZerg', 'SortOf', 'MacSed', 'fraer', 'Pezz', 'JYP' 370 matches between these players since 2014-09-01 Total series: 370 PvT - 44.58% - 83 PvZ - 46.79% - 109 TvZ - 49.38% - 81 Total games: 932 PvT - 47.39% - 211 PvZ - 49.26% - 270 TvZ - 50.24% - 211
That's a remarkable difference between top 50 and top 100. But also notice how 51-100 includes great players from every race. The GSL restriction might be more viable to weed out the EnDerr's of this world.
|
On October 14 2014 06:06 Ghanburighan wrote:Show nested quote +On October 14 2014 05:50 Tuczniak wrote:On October 14 2014 05:44 Ghanburighan wrote:On October 14 2014 05:41 Thieving Magpie wrote:On October 14 2014 05:25 Tuczniak wrote:On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you! Any system with 3 races with an imbalance would look like this with this method. This is not a proof of anything. Not that I disagree with your claim though. I guess I'm too stupid to see what you mean. How can a race be too powerful if only the best players can use it? I think the argument is that there will be more top players for that race. I meant that Top100 will always have winrates closer to 50% than Top50. And that's because in Top100 there is a lot more games with big skill difference between players for example TLO vs Innovation. No matter the balance state, this game will most likely end up with Innovation winning. Those games are more common in top100 and naturally will push winrates towards 50%. PS: sorry TLO Please elaborate on this mechanism. So, let's assume that the top 50 is actually top 3 - one from each race. And the top 100 is top 6. Also, the bottom 3 always lose to the top 3. Now, wouldn't the winrates for top 3 be the same as for top 6?
First of, you have greater variance because of less games. If your expectany is 50% it is much more likely that top50 average is over 60% when you take a sample, than top100 (bigger samplesize of top100 in the same periode).
To your exercise (though only one matchup): say we have top2 and the winrate is 60:40.
Then we have top4 and the winrate is 60:40, however, as you suggest X1 always beats Y2 and Y1 always beats X2. Letting everybody play everbody would now lead to an expectancy of (60+60+100+0)/4=55. So actually the winrate dropped by 5% under the true value. (The imbalance halved by doubling the samplesize).
The reason why this works like that is that we modelled the winchance of better players vs worse players as independend from the true balance. Therefore the introduction of 1 sure win per side (so 50% on average) on top of 2 normal games (60% on average) distorts the winrate.
It remains to be discussed whether Innovation really always beats TLO though. If occasionally TLO takes just one game, that model does not hold.
|
On October 14 2014 06:20 antiRW wrote:Show nested quote +On October 14 2014 05:22 Thieving Magpie wrote:
@antiRW
That is fascinating work and greatly supports a lot of previous theories of Terran being the hardest race as it seems that the higher the percieve skill levels of the players, the higher the overall winrates is for Terran. Conversely, the lower the overall skill rate of the population, the better it is for non-terrans. I've never seen such definitive proof that it's a player skill problem and not a race problem. Thank you! Thanks. It does not quite support that but something very similar: In TvZ, Terran has a huge amount of micro-ability at the top allowing the absolute best players to utilise their units way better than those in the top 100-1000. With build refinement and army positioning there is only so much you can do. Innovation's builds are not insanely betters than those of a top 100 Terran, but his execution and micro is. The problem is, how do you balance now? Balance for the top 10 or the top 100 or the top 1000? Right now TvZ appears imbalanced at all these. p.s.: Can we just leave ironic tone out of it? Let just have a civilised discussion about the state of the game. If it helps, I don't have a stake in the current debate - I play toss.
Terran microability hasn't changed since Innovation vs. DRG. Attack priority Thors, Tanks +10% attack speed, and no upgrade Hellbats aren't exponentially better in the hands of top 10 Terrans as opposed to top 50 Terrans.
What's clearly changed is that Terrans have a lot more compositions available to them in TvZ, and Zerg either do not have a proper response to some of these compositions, or have not yet found those proper responses. I think that 2 months of playtime following a dearth of TvZs is inevitably going to lead to Zergs losing some games that they just don't know how to win. There's no way around it. This kind of meta shift takes time. The only way to keep things level is to give Zerg some weird compositional buff of their own, so that Zerg can catch Terrans off guard in the midgame with a huge Corruptor transition (???) or something. I have nothing against that. But I do think that Zerg's reactive play in TvZ will make this goldilocks solution pretty tough to find.
I admit I haven't seen all the big premier TvZs in the last two months, but of the ones I have seen, there's just nothing to be frustrated about as a Zerg. Let's start with Snute almost beating Flash. Shouldn't Flash playing a foreign Zerg during Flash's reign of supremacy AND Terran's imbalance mean a 3-0 sweep? Apparently not, considering it was a nail-biting 3-2. I can only assume that everyone here thinks it's only natural for Snute to be able to bring Flash, a regular Ro16 Code S player, to the brink of elimination. I don't happen to share that sentiment, but there you go.
soO vs Innovation was just WOL beta cheeses over and over until g6, which I wouldn't dream of basing a balance complaint around no matter how that game had gone. These guys had both gone so far out of their comfort zones in that series, it was sheer insanity.
Scarlett played Snute's style of ZvT against Flash and lost. Well, what do you expect? Scarlett hasn't even been the best foreign Zerg this year when relying on builds she knows inside and out, before Terran got buffed. But after the buff her ZvT should suddenly improve? Well that's not very likely.
Innovation vs DRG is a straight up stomp. Game 2 has zero patch-influenced shenanigans, it's bio vs. ling/bling/Muta all day and DRG looks helpless. This game could have been a part of their series last year for all that balance has to do with it. Game 3 has one Thor over 23 minutes of playtime, no Hellbats, no Tanks, nothing else. How do you make a balance complaint out of that? What's more likely is that DRG (with all the love in the world to the guy) was playing bad Starcraft. Look at the engagement he chooses to take 19 minutes in, sending his Banelings off creep. It's suicidal. The only game that balance complaints could potentially be made from is game 1, in which Inno brought out a mech timing build and crushed through DRG's SHs. I don't know enough about mech vs. SH to know what was going on, but I know that Soulkey crushed Bbyong's mech in Ro32 this season and made it look pretty easy. So I'm not inclined to jump to balance as the culprit, but "DRG doesn't know how to deal with this build that didn't use to exist until 4-5 months ago" does seem possible.
With the exception of DRG vs. Inno g1, there isn't even the potential for the results to be based off a patch in these series. It just can't be done. Eliminate these series from the statistics and DinoMight's 64.5% TvZ winrate drops down to 59% in an instant. That's how small our sample sizes are. I haven't seen the other games, but I have to wonder how many of them had more than a token Thor, or a Hellbat timing that hit before Transformation Servos would have been researched.
edit: fixed my numbers. I need to go back to school.
|
What you're missing is that the threat of a unit being produced can have an effect on the game even if it isn't produced.
Terran now has more options than ever, which makes it harder for Zerg to respond. Yeah you could give excuses as to why every single Zerg lost but in the end, in the elimination stages of all major tournaments since the patch, Terran is 13-4 in series vs. Zerg.
That says something.
|
Gamewise it says nothing. Instead of providing those numbers maybe you can say something about the games instead?
That says 500% more to me than "13-4"
|
On October 14 2014 08:55 Foxxan wrote:
Statwise it says nothing. Instead of providing those numbers maybe you can say something about the games instead?
That says 500% more to me than "13-4"
What is there to say beyond top Zergs cannot beat top Terrans and here are the last 17 series of such games? TvZ win rate for the "top T" and "top Z" was posted earlier in this thread and its 60%+ for every T and above 50% for only 2 players.
The fact that in Bo5+ series the stats get worse for Zerg indicate to me that there is a greater balance problem than people think. Zerg can win once or twice with one off strategies that catch their opponents off guard, but they can't win consistently to take down top Terrans in tournaments. Those stats illustrate the lack of options Z has and all the openers that T has.
|
Quote Made by TheDwf
Just look at the quality of some of the material among all of this!
soO loses an unscouted 2b Marines/Tanks all-in after faking a Roach/Baneling bust
YoDa defends a Roach/Nydus all-in
Golden is caught with only 5 Queens against a Hellbat timing
TRUE loses Marines/Tanks (!) after throwing a gigantic lead
Snute loses MMA's 2b all-in after droning to 80
viOLet loses one map against TaeJa going Roach/Infests after losing 15 drones to Hellion raid as usual
Snute fails 3 bane busts in a row against Flash
Scarlett overdrones against a 2b all-in and loses
Scarlett goes for a Roach/Baneling bust and gets a build order loss
Scarlett's opening completely fails, she ends up behind and thus can't stop a 2b all-in
HyuN gets 79 drones and infest pit against a 2b all-in and loses
HyuN stays on lings/banes for 15 minutes, regularly crashing banes against Marauders
HyuN goes for a Roach/Baneling bust and gets a build order loss
If the fact you can't drone to 80 against a 2b all-in doesn't prove there's a problem in TvZ, I don't know what will
This says 5000% more to me than some statistics on a paper. Especially with such a low sample.
The only game iam curious about is game 2 that hyun played.
|
On October 14 2014 08:46 DinoMight wrote:
What you're missing is that the threat of a unit being produced can have an effect on the game even if it isn't produced.
The threat of Innovation making Thors that shoot Mutalisks first made soO drop two games to two rax? Innovation played a straight up bio-mine style in games 2 and 3, what's DRG scared of, Innovation teching out of bio-mine into mech at the 25 minute mark?
Terran was screwed against Protoss because so many of their openings looked alike, and it was so easy to hide tech. Blink Stalker poke and Blink Stalker all-in and DT rush look nearly indistinguishable. These are concrete complaints. What's the concrete TvZ equivalent? DRG couldn't tell Innovation was going mech in game 1 and had to defend as though it might turn into a bio build...?
Terran now has more options than ever, which makes it harder for Zerg to respond. Yeah you could give excuses as to why every single Zerg lost but in the end, in the elimination stages of all major tournaments since the patch, Terran is 13-4 in series vs. Zerg.
That says something.
It says "in the elimination stages of all major tournaments since the patch, Terran is 13-4 in series vs. Zerg." If you want it to mean something on top of that, like "the game isn't balanced," you've gotta make that argument yourself. A temporarily unstable meta explains those results just as well.
|
Northern Ireland27469 Posts
Do we have a list of those 13-4?
|
On October 14 2014 09:21 Wombat_NI wrote:
Do we have a list of those 13-4?
DinoMight's post.
|
Add Scarlett vs Heart and Hyun vs Heart from WCS AM to that list and its 13-4.
I guess there's no sense in arguing this. No number of games will be a "large enough sample size" and even when there are enough games I'm sure TheDwf can think of a thousand reasons why each Zerg lost.
Snute drones to 80... maybe because he feels its the only way he has a chance of winning later in the game?
Hyun goes for a Roach Bane allin because... the game is imbalanced and he needs to try and find an edge?
Etc.
You can go on forever justifying why XYZ happened. But at the end of the day top Zergs cannot beat top Terrans in Bo5+.
But you're right, Zerg players just need to adapt and "l2p brah."
|
Northern Ireland27469 Posts
If I was to have Liquibetted those I think I might have predicted maybe two or three wrongly (Solar v Cure I thought the former was going to win). Really doesn't mean a massive amount, but in terms of my intuition over who I would have favoured for their ability in the matchup those were pretty much what I would have expected.
|
On October 14 2014 09:26 DinoMight wrote:Add Scarlett vs Heart and Hyun vs Heart from WCS AM to that list and its 13-4. I guess there's no sense in arguing this. No number of games will be a "large enough sample size" and even when there are enough games I'm sure TheDwf can think of a thousand reasons why each Zerg lost. Snute drones to 80... maybe because he feels its the only way he has a chance of winning later in the game? Hyun goes for a Roach Bane allin because... the game is imbalanced and he needs to try and find an edge? Etc.
You do realize that you're complaining that Snute didn't beat Flash, right?
Snute.
Flash.
Last month.
After patch.
3-2.
It's like being upset that Thorzain loses 2-3 to herO one season back. You gotta pick your battles, man.
You can go on forever justifying why XYZ happened. But at the end of the day top Zergs cannot beat top Terrans in Bo5+.
But you're right, Zerg players just need to adapt and "l2p brah."
When Terrans were losing to Protoss, Terrans had CONCRETE COMPLAINTS. Offer a concrete complaint and we'll have something to talk about. If you can't think of one, entertain the possibility that it's because there is no concrete complaint to be had.
|
|
|
|
|
|
EPT
