|
I expect Terrans to stop doing MMMM, and go into mech to balance out TvZ.
TvP seems balanced, but the oracle is still broken, because any unit with high damage that can obliterate mineral lines gives the game another luck factor, (like DTs, pre-nerfed hellbats, banshees during 1/1/1) and the game is already heavily luck based as is.
I only expect Protoss to continue to win more against T and Z in the future
|
I flipped a coin a couple of times and heads won with 115 out of 216 that's a 53.2% winrate for HvT. Than I flipped some again and now tails won 102 out of 195, that's a 52.3% winrate for TvH.
OMG T got better after patch!!! blah, blah, imba, imba, redesign the game, remove forciefieldstimcolossusswarmhost give us back the reaver.
|
terran just needs warhound back and all will be well
|
On December 09 2013 23:50 BakedButters wrote: I expect Terrans to stop doing MMMM, and go into mech to balance out TvZ.
One can still hope.
Sorry for being disillusional but I don't think it will happen :-p
|
Canada16217 Posts
Thanks for posting the winrates!
|
Sure thing is the game was more balanced before the patch. As expected, this patch is a failure and illustrates the fact that most of the changes were not necessary.
|
so 2400 total games.. how many of those are PVT?
I think the sample size is way too small considering the amount of metagaming that goes on in these tournaments. Also the korean vs foreigner skill gap.
Is there any tool available to look at ALL korean GM games?
|
Smaller sample size, then some people complain about it. Bigger sample size, then others complain about the quality of the games.
No matter how big or small sample sizes are, people find a way to discredit it. Maybe we should kidnap all those top Korean players and have them play millions of games in a basement so that we can finally have a huge quality sample size.
|
On December 10 2013 03:22 Orek wrote: Smaller sample size, then some people complain about it. Bigger sample size, then others complain about the quality of the games.
No matter how big or small sample sizes are, people find a way to discredit it. Maybe we should kidnap all those top Korean players and have them play millions of games in a basement so that we can finally have a huge quality sample size. Without Top5 player each race is the best. A top5 player is winning against everyone until he mets a top5 player from other race.
|
On December 10 2013 01:39 deepfield1 wrote: so 2400 total games.. how many of those are PVT?
I think the sample size is way too small considering the amount of metagaming that goes on in these tournaments. Also the korean vs foreigner skill gap.
Is there any tool available to look at ALL korean GM games?
205 TvP games after the patch. 428 games before the patch.
Btw. If I am doing my math correctly, there is around 350 game discrepancy between the sample size stated on the image and the actual games listed in the google doc.
Not sure where those games disappeared, I asked the OP about it, but so far no response from him.
|
On December 10 2013 03:22 Orek wrote: Smaller sample size, then some people complain about it. Bigger sample size, then others complain about the quality of the games.
No matter how big or small sample sizes are, people find a way to discredit it. Maybe we should kidnap all those top Korean players and have them play millions of games in a basement so that we can finally have a huge quality sample size.
all of korean GM would be a good place to start.
Who is complaining about a large sample size? and quality of games? lol
|
Still pretty good, but I fear a long fall toward an End-of-2012-like abyss for Terran.
|
On December 10 2013 03:32 Qwerty85 wrote:Show nested quote +On December 10 2013 01:39 deepfield1 wrote: so 2400 total games.. how many of those are PVT?
I think the sample size is way too small considering the amount of metagaming that goes on in these tournaments. Also the korean vs foreigner skill gap.
Is there any tool available to look at ALL korean GM games? 205 TvP games after the patch. 428 games before the patch. Btw. If I am doing my math correctly, there is around 350 game discrepancy between the sample size stated on the image and the actual games listed in the google doc. Not sure where those games disappeared, I asked the OP about it, but so far no response from him.
Sorry the data set was not updated since the mistake. It is now updated.
Pre patch TvP has a sample size of 509
post patch TvP has a sample size of 205
|
On December 10 2013 04:05 deepfield1 wrote:Show nested quote +On December 10 2013 03:22 Orek wrote: Smaller sample size, then some people complain about it. Bigger sample size, then others complain about the quality of the games.
No matter how big or small sample sizes are, people find a way to discredit it. Maybe we should kidnap all those top Korean players and have them play millions of games in a basement so that we can finally have a huge quality sample size. all of korean GM would be a good place to start. Who is complaining about a large sample size? and quality of games? lol
Sure let's use blind, random bo1s online to determine our balance. I never really figured this out, no one has ever before pointed to ladder as legitimate balance evidence until now and anyone that tried was basically mocked. A race that relies more on tricks and gimmicks(Protoss) will naturally be better on ladder.
|
That Oracle/Mine patch made things worse. It was more balanced before that it is now.
|
yes , David Kim destroyed the game
|
Does Fragbite Masters count in this tally?
|
Thanks for the interesting statistics. Balance is looking great.
|
Seems pretty close to perfect balance. The differences between pre-patch and post-patch winrates is probably statistical fluctuation.
OP can you give us p-values for the changes? I'd like to see how statistically significant the changes in winrate before and after the patch are.
|
Looks like the balance in PvT could be tuned just a tiny bit, but other than that it looks surprisingly close.
|
|
|
|