| Forum Index > SC2 General |
|
Big J
Austria16289 Posts
On January 03 2012 05:03 secretary bird wrote: If anyone doesnt take ladder seriously at all ean GM players though. just because I brought it up I would like to add here that I only wanted to use this 65% stats against some dumb comment that claimed that TvP is unplayable for nonpros... did not think that this would turn out into a discussion abou numbers that are completly offtopic. | ||
|
docvoc
United States5491 Posts
| ||
|
Let it Raine
Canada1245 Posts
to elaborate here's a quote: International winrates are pointless because they include the most ridiculously pointless games with incredible skill gaps. Let's look at International PvT versus Korean PvT in December and you tell me what's a better indicator of balance. International: Brat_OK (T) vs. Pomi (P) Polt (T) vs. MaNa (P) MVP (T) vs. BlinG (P) MVP (T) vs. Cobo (P) MVP (T) vs. Splendour (P) MarineKing (T) vs. Wern (P) MarineKing (T) vs. Babyknight (P) MarineKing (T) vs. Hui (P) Korean: Clide (T) vs. JYP (P) Bomber (T) vs. Creator (P) Bomber (T) vs. MC (P) Noblesse (T) vs. Sage (P) MarineKing (T) vs. Squirtle (P) ASD (T) vs. Genius (P) NaDa (T) vs. HuK (P) Polt (T) vs. Naniwa (P) | ||
|
Drowsy
United States4876 Posts
| ||
|
Big J
Austria16289 Posts
On January 03 2012 05:26 Let it Raine wrote: these stats =/= balance to elaborate here's a quote: well we/blizzard need some stats to discuss balance. seeing how international tournament stats at least have a useful monthly samplesize in the amount of game/players i prefer those furthermore i think (have no stats for this, but i think it has been argued by people with actual stats like blizzard) that race distribution is quite good among the international proscene, so these pro vs nonpro games should balance out overall... though this also means that everything gets closer to 50-50 and therefore 2% in those stats might actually be 10% in pro vs pro matches... fuck... the more i think about it, the less certain I am wether there are any good stats available at all from the proscene and all we can really take from them is something like "race X probably favored" kind of get now why blizzard uses their own stats... | ||
|
DarkPlasmaBall
United States45003 Posts
::shrugs:: 10% off 50/50 is supposedly concerning, right? I wonder if Terran is going to get nerfed again. | ||
ZeromuS
Canada13389 Posts
On January 03 2012 05:58 DarkPlasmaBall wrote: TvZ is getting closer, PvZ is just looping around, and TvP is just lol. ::shrugs:: 10% off 50/50 is supposedly concerning, right? I wonder if Terran is going to get nerfed again. No no no. Remember P is op since Terran had a 6% difference last time around in favour of P. Remember  | ||
|
rd
United States2586 Posts
On January 03 2012 04:05 Badfatpanda wrote: But there's a very high chance it will be considering the previous posts >.> OK, that's valid but since I look at ELO as a measure of skill in all matchups, mirror does come into account. So considering the fact that mirror matchups will dilute whatever potential imbalance there is, the resulting calculations will be a more accurate indication of performance as a whole. Regarding imbalance, any cries of imbalance relating to these graphs is relating to a fear that eventually if trends continue imbalance would exist. These numbers aren’t static. They shift almost constantly with the metagame as newly discovered strategies spread through the community, and that heavily influences how they’re interpreted. Also, due to the way the math works out we will almost never see ratios of 50:50; we expect a variance of +/- 5% in these results. So, if a win/loss ratio is approximately 55%:45%, this would indicate that the matchup is well balanced—we expect those numbers to fluctuate within that range to some degree. Ratios just outside of that range are still within acceptable boundaries. It is only after win/loss ratios exceed 60%:40% that there is an indication that a potential imbalance might exist. We keep a sharp eye on these variations from day to day and week to week, staying constantly alert for where the numbers are changing and what the possible causes could be. It’s fairly common, for example, for a new strategy or build order to skew the numbers in favor of a particular race for a brief period, until the metagame catches up and the counter strategies spread through the community. From the balance team, source: http://us.battle.net/sc2/en/blog/3551858/StarCraft_II_Balance_Snapshot_-9_22_2011#blog So disregarding April, the game has been in a state of overall balance. And even the April turn only reinforces the point of the article I provided. Directly after the amulet removal toss were in a depression, 1 month later with no balance patch the MU magically was buffed up 14%. So I feel ELO up to this point is determined more from player skill than it exists from a state of imbalance in the game overall. Oh I'm not arguing whether or not the stats indicate balance or player skill. I'm just pointing out the correlation between match-up statistics and Terrans in the high ELO are essentially the same thing. | ||
|
Sadistx
Zimbabwe5568 Posts
On January 02 2012 23:45 Ermac wrote: I wonder what TvP statistics would look like if you took out the all ins. Imho the matchup is balanced terribly at the moment. Protoss dominates most straight up games while Terran keeps crushing them with all ins. Unfortunately, that's unlikely to change until HOTS, because they are unlikely to do a big balance patch until then (I could be mistaken obviously, but unless they do something with pylon warp-in, nothing will change). | ||
|
Scila
Canada1849 Posts
You could see a similar situation in pretty much every match up. In PvZ, Protoss tends to win with either a big 2 base all-in, or a super late game situation where the Zerg can't handle the deathball. In TvZ, Terran also tends to lose much more in the late game, and win much more in the early and mid game. It's no big secret that Terran overall is weaker in the late game than the other races, and instead tends to rely on gimmicky aggression-based builds and timings to win a lot if not most of their games. I like the fact that most match ups are going towards 50% balance, but that doesn't look at the issue of different races struggling at different points in a game, like Terran in Late game TvP, or Protoss fast expanding versus Zerg. Every race should win and lose at ANY point of the game resulting DIRECTLY from the skill and decisions of the player. THAT is good design, and what I think was almost achieved in Brood War. | ||
|
K3Nyy
United States1961 Posts
On January 03 2012 05:26 Let it Raine wrote: these stats =/= balance to elaborate here's a quote: What's wrong with the Korean PvT? It looks perfectly fine to me. | ||
|
Belisarius
Australia6232 Posts
I do wish Ctuchik would standardise his axes. The international looks a lot worse than it is because it's a 20% range and KR is 40. I would also love it if TLPD stored game lengths. I think you'd see a huge correlation in PvT for that. | ||
|
skatbone
United States1005 Posts
On January 03 2012 04:37 SeaSwift wrote: Interesting. It looks like on ladder, at least, Zerg is underrepresented in almost every league. Is the race just not appealing to play as or something? I wouldn't read too much into these numbers. Just 3 or 4 weeks ago, Terran was so underrepresented, according to SC2Ranks, that we had that painful thread about Terrans dying out. Once Christmas break came around, the Terrans seemed to re-emerge in diamond. I play against more T now that I do P or Z. tl;dr These things are in flux. | ||
|
Big J
Austria16289 Posts
On January 03 2012 08:26 Scila wrote: What I think most people in this thread don't realize is that balance does not = good design and a good game. Sure, by looking at these statistics of roughly 50% win ratios at the top level (which is how it should be measured) you can say that the game is "balanced". However, this doesn't look at the real issues of game design. For example, in TvP, Terran tends to win much more before the "late" game. Also, Terran tends to use all-ins like the 1 base or 2 base 1/1/1 a lot to achieve these wins. Protoss, on the other hand, tends to have much higher win ratios in the late game. You could see a similar situation in pretty much every match up. In PvZ, Protoss tends to win with either a big 2 base all-in, or a super late game situation where the Zerg can't handle the deathball. In TvZ, Terran also tends to lose much more in the late game, and win much more in the early and mid game. It's no big secret that Terran overall is weaker in the late game than the other races, and instead tends to rely on gimmicky aggression-based builds and timings to win a lot if not most of their games. I like the fact that most match ups are going towards 50% balance, but that doesn't look at the issue of different races struggling at different points in a game, like Terran in Late game TvP, or Protoss fast expanding versus Zerg. Every race should win and lose at ANY point of the game resulting DIRECTLY from the skill and decisions of the player. THAT is good design, and what I think was almost achieved in Brood War. until there are any statistics for this (which there wont be because it's simply not true... the game has been designed for each race to have achievable composition in the lategame that can at least be even with your opponents) this is nothing but the usual "races i dont play are imba"-crap paired with a little bit of "bw is better"-crap. | ||
|
Thrombozyt
Germany1269 Posts
Just calculate it on a 95% confidence level (p<0.05) based on the number of matches going into each graph. That would clear up a lot of doubts, because I personally have a feeling that the data base for the December graph - especially the Korean - is rather shaky. | ||
|
Mowr
Sweden791 Posts
On January 03 2012 05:58 DarkPlasmaBall wrote: TvZ is getting closer, PvZ is just looping around, and TvP is just lol. ::shrugs:: 10% off 50/50 is supposedly concerning, right? I wonder if Terran is going to get nerfed again. I guess they will continue the pattern and to nothing about the early game and keep buffing protoss late game. Hey, balance! | ||
Catatonic
United States699 Posts
On January 02 2012 21:10 HowardRoark wrote: Korean stats are the only that ought to matter, and fairly balanced. Zerg having the worst winrates, but with the highest skill ceiling for Z I do not see any problem with this. Zerg will in some month's come up on top without patches. Elitism for the win man? No not really though, they should definatly take in winrates from more then Koreans seeing as how the majority of the people playing the game are non Koreans. Yea the Koreans are better but when the majority of the populace isn't Korean why only focus on winrates that don't apply to them as players. These winrates are here to get a general consensus of which race is winning an discuss the reasons for that. Only adding Koreans negates more then 50% of the community cause most don't have the skill of a Korean so why have us compare ourselves against them? | ||
|
Wyk
314 Posts
| ||
|
Shield
Bulgaria4824 Posts
On January 03 2012 09:36 Big J wrote: until there are any statistics for this (which there wont be because it's simply not true... the game has been designed for each race to have achievable composition in the lategame that can at least be even with your opponents) this is nothing but the usual "races i dont play are imba"-crap paired with a little bit of "bw is better"-crap. Not really. He is right about design. | ||
|
Meega
Germany35 Posts
On January 03 2012 08:26 Scila wrote: What I think most people in this thread don't realize is that balance does not = good design and a good game. Sure, by looking at these statistics of roughly 50% win ratios at the top level (which is how it should be measured) you can say that the game is "balanced". However, this doesn't look at the real issues of game design. For example, in TvP, Terran tends to win much more before the "late" game. Also, Terran tends to use all-ins like the 1 base or 2 base 1/1/1 a lot to achieve these wins. Protoss, on the other hand, tends to have much higher win ratios in the late game. You could see a similar situation in pretty much every match up. In PvZ, Protoss tends to win with either a big 2 base all-in, or a super late game situation where the Zerg can't handle the deathball. In TvZ, Terran also tends to lose much more in the late game, and win much more in the early and mid game. It's no big secret that Terran overall is weaker in the late game than the other races, and instead tends to rely on gimmicky aggression-based builds and timings to win a lot if not most of their games. I like the fact that most match ups are going towards 50% balance, but that doesn't look at the issue of different races struggling at different points in a game, like Terran in Late game TvP, or Protoss fast expanding versus Zerg. Every race should win and lose at ANY point of the game resulting DIRECTLY from the skill and decisions of the player. THAT is good design, and what I think was almost achieved in Brood War. I really like this post! I think you are 100% right - these are things blizzard should recognize and deal with! | ||
| ||
StarCraft 2 StarCraft: Brood War Dota 2 League of Legends Other Games |
|
The PondCast
LAN Event
OSC
Replay Cast
OSC
LAN Event
Korean StarCraft League
CranKy Ducklings
WardiTV Korean Royale
LAN Event
[ Show More ] |
|
|
EPT
" to "mech is so bad, why can't I go mech even though Protoss can't go stargate either :'(".
