|
On September 13 2014 11:55 Loccstana wrote:
Siege tanks are very underpowered at the moment. They need a range increase and some sort of damage bonus against Protoss units.
Tanks shoot plenty far and a range increase would be dumb, the trouble is and always has been that their damage against non-armored units is terrible and basically everything Protoss has on the ground counters them. Chargelots, archons, and immortals slaughter them and stalkers can just blink around a tank-based army all day.
|
Zerg seems to be doing fine at the general level but rather poorly at the top level. Only about 12% of the premier tournaments in 2014 were won by zerg, and that's a long time sample. Given that the win percentage against Terran has dropped following both hellbat and mine patch I doubt they're going to win any particular amount of future tournaments without some sort of patch. Just not sure what exactly they would patch.
|
I think they should add siege tank bonus damage to shields.. i dont know but i still think terran needs additional firepower in TvP.. maybe this will help in the late game scenario in the positioning of deathballs.. I just think Protoss have the upper hand right now since they have the collosus which has range and AOE, and terran has nothing that has AOE except mines (which is not used in late games) and EMP which only has moderate range..
Also and i dont know if I am allowed to post this here but i kind of like the inclusion of mech in TvP just like in Broodwar..
Any thoughts? Maybe i am wrong but i still think this is a good angle to look into.. 
|
On September 13 2014 12:47 bo1b wrote:Show nested quote +On September 13 2014 12:22 sibs wrote:
Right now at the very top T&P > Z for sure, like I mentioned a few pages ago, there's a reason every Zerg got picked so fast for ro16 groups, aligulac paints the same picture.
Terran is winning TvZ by the same margins they won at the beginning of hots, and ZvP is very map dependant, but in general favors Protoss. And yet before the widowmine nerf things were 50/50 with zerg having adapted. Now is not the time for a kneejerk reaction
Just saying winrates were 50% for 1 month doesn't say that much, I don't want the patch reversed either, I say let it play it out, but to think the game is anywhere near balanced is naive.
http://aligulac.com/periods/101/
Zergs still weren't winning much, if anything at all, Hots went from Terran dominance to Protoss dominance, to either Terran or Protoss dominance now.
|
Err, Z is still the most represented race. Looking at the previous and the ongoing Aligulac list. This suggests that it can't be struggling that much. And the favourite to win is Soo who just took out Flash, the hottest terran. I see no evidence for your claims.
|
On September 13 2014 15:26 Ghanburighan wrote:
Err, Z is still the most represented race. Looking at previous and the ongoing Aligulac list. This suggests that it can't be struggling that much. And the favourite to win is Soo who just took out Flash, the hottest terram. I see no evidence for your claims.
To be fair, soo played what looked like the best he ever had that series, and flash didn't split, and played an inferior build.
|
On September 13 2014 15:27 bo1b wrote:Show nested quote +On September 13 2014 15:26 Ghanburighan wrote:
Err, Z is still the most represented race. Looking at previous and the ongoing Aligulac list. This suggests that it can't be struggling that much. And the favourite to win is Soo who just took out Flash, the hottest terram. I see no evidence for your claims. To be fair, soo played what looked like the best he ever had that series, and flash didn't split, and played an inferior build.
Yeah, sure, but Flash just used that inferior build to win a tournament... And beat tons of great players in an insane run. Yet Soo had the tools to beat him handily. In the end, winning doesn't suggest a balance problem no matter what could be improved.
|
On September 13 2014 13:01 ZombieFrog wrote:
Zerg seems to be doing fine at the general level but rather poorly at the top level. Only about 12% of the premier tournaments in 2014 were won by zerg, and that's a long time sample. Given that the win percentage against Terran has dropped following both hellbat and mine patch I doubt they're going to win any particular amount of future tournaments without some sort of patch. Just not sure what exactly they would patch.
Thats true, but theres been a lot of 2nd places, so to me it seems like zergs are unable to win a Bo7 finals only.
|
On September 13 2014 16:00 Ghanburighan wrote:Show nested quote +On September 13 2014 15:27 bo1b wrote:On September 13 2014 15:26 Ghanburighan wrote:
Err, Z is still the most represented race. Looking at previous and the ongoing Aligulac list. This suggests that it can't be struggling that much. And the favourite to win is Soo who just took out Flash, the hottest terram. I see no evidence for your claims. To be fair, soo played what looked like the best he ever had that series, and flash didn't split, and played an inferior build. Yeah, sure, but Flash just used that inferior build to win a tournament... And beat tons of great players in an insane run. Yet Soo had the tools to beat him handily. In the end, winning doesn't suggest a balance problem no matter what could be improved.
Eh, that tournaments best zerg he played against had a huge mental breakdown and still almost lost to him. Not that I'm saying flash is bad, I just think his tvz style is pretty weak comparatively.
|
On September 13 2014 15:26 Ghanburighan wrote:
Err, Z is still the most represented race. Looking at the previous and the ongoing Aligulac list. This suggests that it can't be struggling that much. And the favourite to win is Soo who just took out Flash, the hottest terran. I see no evidence for your claims.
Proving or disproving balance based on a single series is not a good way to go about it. If one race wins tournament after tournament after tournament, which was the case with Terran in 2010-2011, Zerg near the end of WoL and Protoss recently, then there is something wrong. Right now it's really too early to tell.
|
On September 13 2014 12:47 bo1b wrote:Show nested quote +On September 13 2014 12:22 sibs wrote:
Right now at the very top T&P > Z for sure, like I mentioned a few pages ago, there's a reason every Zerg got picked so fast for ro16 groups, aligulac paints the same picture.
Terran is winning TvZ by the same margins they won at the beginning of hots, and ZvP is very map dependant, but in general favors Protoss. And yet before the widowmine nerf things were 50/50 with zerg having adapted. Now is not the time for a kneejerk reaction
It's funny how this 50/50 is constantly being brought up. Zerg was struggling against Terran for months and when DRG manages to play absolutely flawless and wins a single match vs. Innovation, that's a clear indicator that Zergs were 'beginning to adapt'.
|
On September 13 2014 21:53 ReMinD_ wrote:Show nested quote +On September 13 2014 12:47 bo1b wrote:On September 13 2014 12:22 sibs wrote:
Right now at the very top T&P > Z for sure, like I mentioned a few pages ago, there's a reason every Zerg got picked so fast for ro16 groups, aligulac paints the same picture.
Terran is winning TvZ by the same margins they won at the beginning of hots, and ZvP is very map dependant, but in general favors Protoss. And yet before the widowmine nerf things were 50/50 with zerg having adapted. Now is not the time for a kneejerk reaction It's funny how this 50/50 is constantly being brought up. Zerg was struggling against Terran for months and when DRG manages to play absolutely flawless and wins a single match vs. Innovation, that's a clear indicator that Zergs were 'beginning to adapt'.
He then went on to not drop a map when it got to the muta/ling/bling stage for like 6 months lol. The guys a beast.
Having said that, curious beat innovation, and hyvaa almost did too.
|
On September 13 2014 21:53 ReMinD_ wrote:Show nested quote +On September 13 2014 12:47 bo1b wrote:On September 13 2014 12:22 sibs wrote:
Right now at the very top T&P > Z for sure, like I mentioned a few pages ago, there's a reason every Zerg got picked so fast for ro16 groups, aligulac paints the same picture.
Terran is winning TvZ by the same margins they won at the beginning of hots, and ZvP is very map dependant, but in general favors Protoss. And yet before the widowmine nerf things were 50/50 with zerg having adapted. Now is not the time for a kneejerk reaction It's funny how this 50/50 is constantly being brought up. Zerg was struggling against Terran for months and when DRG manages to play absolutely flawless and wins a single match vs. Innovation, that's a clear indicator that Zergs were 'beginning to adapt'.
Eh you find no support from statistics that zerg was struggling vs terran after summer 2013. According to Aliguac statistics, zerg win/rates were doing just fine, despite being overpresented relatiev to terran. In WCS Korea, win/rates didn't favor terran either.
However, post-Widow mine nerf, adjusted win/rates began favoring zerg quite a bit. Check out the below graph I posted 2 weeks ago.
![[image loading]](http://i57.tinypic.com/2m336up.png)
The unadjusted win/rates for August are exactly what we would predict given that the TvZ and TvP matchup are becoming more balanced.
|
Where are you pulling those graphs from?
They're at odds with other data.
|
On September 13 2014 23:41 sibs wrote:
Where are you pulling those graphs from?
They're at odds with other data.
They are not at adds with other data. They are based on Aliguac win/rates, but adjused to take into account race distribution (also based on Aliguac data).
Read more about the appoach at page 1078 and 1079. Note that, it's by no means a perfect approach, but far better than naked win/rates which are very deceiving for multiple different reasons (which is explained further in page 1078 and 1079-1080).
|
On September 13 2014 23:50 Hider wrote:Show nested quote +On September 13 2014 23:41 sibs wrote:
Where are you pulling those graphs from?
They're at odds with other data. They are not at adds with other data. They are based on Aliguac win/rates, but adjused to take into account race distribution (also based on Aliguac data). Read more about the appoach at page 1078 and 1079. Note that, it's by no means a perfect approach, but far better than naked win/rates which are very deceiving for multiple different reasons (which is explained further in page 1078 and 1079-1080).
Adjusting the data is sketchy. You will notice that people using any sort of "adjustment" in a balance whine thread typically only do so when their argument is buttressed by the assumptions underlying the "adjustment." I'm fine with temporarily giving the benefit of the doubt that high Terran win rates are the result of meta disruption rather than imbalance. But that chart just seems sketchy because it hides the fact that Terran is currently doing very well in both non-mirror matchups.
|
On September 13 2014 23:50 Hider wrote:Show nested quote +On September 13 2014 23:41 sibs wrote:
Where are you pulling those graphs from?
They're at odds with other data. They are not at adds with other data. They are based on Aliguac win/rates, but adjused to take into account race distribution (also based on Aliguac data). Read more about the appoach at page 1078 and 1079. Note that, it's by no means a perfect approach, but far better than naked win/rates which are very deceiving for multiple different reasons (which is explained further in page 1078 and 1079-1080).
Could you link to the comment where you explain your method?
|
On September 14 2014 00:27 cheekymonkey wrote:Show nested quote +On September 13 2014 23:50 Hider wrote:On September 13 2014 23:41 sibs wrote:
Where are you pulling those graphs from?
They're at odds with other data. They are not at adds with other data. They are based on Aliguac win/rates, but adjused to take into account race distribution (also based on Aliguac data). Read more about the appoach at page 1078 and 1079. Note that, it's by no means a perfect approach, but far better than naked win/rates which are very deceiving for multiple different reasons (which is explained further in page 1078 and 1079-1080). Could you link to the comment where you explain your method?
It would be good to call it something other than a win rate, say 'balance indicator' or 'strength indicator'. Otherwise you will always get flack for it not actually being a true win rate.
I would also be curious to see the exact formula. Judging balance in a skewed distribution is an interesting mathematical problem.
|
On September 14 2014 00:45 antiRW wrote:Show nested quote +On September 14 2014 00:27 cheekymonkey wrote:On September 13 2014 23:50 Hider wrote:On September 13 2014 23:41 sibs wrote:
Where are you pulling those graphs from?
They're at odds with other data. They are not at adds with other data. They are based on Aliguac win/rates, but adjused to take into account race distribution (also based on Aliguac data). Read more about the appoach at page 1078 and 1079. Note that, it's by no means a perfect approach, but far better than naked win/rates which are very deceiving for multiple different reasons (which is explained further in page 1078 and 1079-1080). Could you link to the comment where you explain your method? It would be good to call it something other than a win rate, say 'balance indicator' or 'strength indicator'. Otherwise you will always get flack for it not actually being a true win rate. I would also be curious to see the exact formula. Judging balance in a skewed distribution is an interesting mathematical problem.
It's taking a data set from a total population group (pro games) then weighing the wins of different races more heavily depending on the bias of the adjustor. Politicians do it all the time, totally "legit."
|
On September 14 2014 00:10 Salient wrote:Show nested quote +On September 13 2014 23:50 Hider wrote:On September 13 2014 23:41 sibs wrote:
Where are you pulling those graphs from?
They're at odds with other data. They are not at adds with other data. They are based on Aliguac win/rates, but adjused to take into account race distribution (also based on Aliguac data). Read more about the appoach at page 1078 and 1079. Note that, it's by no means a perfect approach, but far better than naked win/rates which are very deceiving for multiple different reasons (which is explained further in page 1078 and 1079-1080). Adjusting the data is sketchy. You will notice that people using any sort of "adjustment" in a balance whine thread typically only do so when their argument is buttressed by the assumptions underlying the "adjustment." I'm fine with temporarily giving the benefit of the doubt that high Terran win rates are the result of meta disruption rather than imbalance. But that chart just seems sketchy because it hides the fact that Terran is currently doing very well in both non-mirror matchups.
On one hand I agree with you in the sense that people will tend to manipulate data to support their beliefs, on the other hand, I don't see a lot of poepole actually coming up with adjusted metrics that takes into account more than one factor. Since we know that the distribution of players can explain some part of balnace that win/rates cannot, it makes sense to look at both win/rates and distribution of players in a balance assesment.
I have been recommending people to do that for a couple of years now, and I am glad to see that more and more people are looking at the amount of mirror-matchups along with win/rates to get a more comprehensive view of balance.
But I still see people who make statements like "race X has 55% win/rates and therefore it is OP". These statements are comparable to saying "China has a higher level of GDP than Denmark, and is therefore richer". That's obviously nonsense, and we here need to take into account population and divide it by the GDP to get GDP per citicizen, which is a much better indicator of how rich people are.
The issue with adjusting win/rates is that we do not not have the "population data" that we just can divide by "GDP". So instead we have to use metrics which aren't 100% perfect, but still gives a good indication of the "population". That means that the exact adjusted win/rates should be taken by grain of salt. Instead, it's more important to look at how the adjusted win/rates have changed over time and see whether the numbers generally seem to make sense.
Assumptions
Allow me to elaborate a bit further on the assumptions behind the adjusted win/rates.
Assumption 1: There are no structural reasons to expect that competitive terrans should have fewer games played than competitive zerg/toss players. That may not always be true in the case where there is a local weekly tournament with lots of games played, and the players participating there are mostly zergs, while more competitive terrans besides in regions with less tournament activity.
Assumption 2: The importance of win/rates relative to distribution of players in the estimation of adjusted win/rates is somewhat arbitrarly decided upon (I discussed why this approach had to be used at page 1079 compared to a more statistical "correct" one).
Track-record of adjusted/win-rates
In quantifying the "importance"-factor, I studed the historical data compared to general consensus of balance and decided upon the values based on that. I would never have posted the graph if the adjusted win/rates had not had a much better track-record in the assesment of balance than naked win/rates. Unlike naked win/rates, the adjusted win/rates could identify the below balance issues much quicker:
(1) Terran imba early WOL
(2) TvZ continued to be T-favored after the Reaper/Siege Tank-nerf.
(2) Toss needing a buff vs terran in summer 2011
(3) Queen buff being unneeded from a balance-perspective. If Blizzard wanted to buff zerg early game, they should have copensated terran late game.
(4) It took Blizzard untill December 2012 to nerf Infestors. According to the adjusted win/rates there were already very strong signs of a significant imbalance in summer 2012.
(5) Mine-nerf was never needed balance-wise.
(6) Terrans significant struggles in the first half of 2014.
Are the assumptions wrong?
If you believe the adjustments underestimate the importance of win/rates (relative to distribution), and make a general adjustment to the mode, then the adjusted win/rates for ZvT during the patch-zerg era will at the same time be much closer to 50%. Does that really make sense (?)
The only way to maintain the the hypothesis that (1) TvZ was T favored prior to the Widow Mine nerf and (2) Zerg was too strong during the patch-zerg era at the same time, would be to assume that a structural change from 2012 to 2013 had occured.
I don't see the evidence of this structural change, and I prefer to use this methdology consistently rather than make adjustments on case-by-case basis unless I see very convincing arguments for why I should do that.
If you start getting into the habit of small adjustments, then I believe race-bias is gonna have a much more meaningful impact on the win/rates.
|
|
|
|
|
|
EPT
