|
Alright, so here are some charts I made on race winrates.
All data is from TLPD, so this is ONLY pro games from tournaments and leagues. NO ladder games. Mirror matchups are completely excluded from the data set since they would skew the results towards 50%.
MLG dallas is not included in April since they haven't released all their replays yet.
First one is from the international scene:
![[image loading]](http://imgur.com/dA8rr.png)
And here is the korean version. NOTE THAT THE KOREAN VERSION HAS A MUCH LOWER SAMPLE SIZE! The number of players and tournaments in Korea is much lower, so therefore these numbers are more volatile.
![[image loading]](http://imgur.com/Z2rbu.png)
More stuff on my twitter here and on reddit here.
|
Very Interesting stuff. Thanks for sharing!
|
Nice to know the PvZ troubles i have been having is echoed in korea. Very interesting results.
|
Wow, I can almost understand why not that many koreans pick up SC2 if their statistics look like that. :O
Foreign stats look surprisingly balanced, shouldn't be surprising that it evens out like that after this amount of time. Strange the koreans haven't though.
|
Wow, nice charts. Amazed how different things are in KR.
|
wow could have sworn some of this stuff is the opposite. Welp.
|
I think it's because in Korea, they don't have a lot of tournament, so not a lot of players ^^
|
The Korean stats are based on a much lower number of games. There just aren't that many tournaments over there. Since the sample size is smaller, the stats are more volatile.
|
Is there anyway you can add more games for each of these? The differences between Korea and International are huge. It almost seems like in Korea the game is becoming almost rock, paper, scissors based on race.
|
I wonder. Is it fair to say that korean protosses are generally terrible? That's the feeling I get from watching them but maybe I'm trying too hard to say zerg up.
Just when I watch a top 8 code s player like anypro pylon his nexus first, pull half his probes to kill it and then push out with only gateway units vs burrow roaches, I kind of have to wonder. (of course, he still won that series with 2 very skill less builds, the first of which he won even though he had 0 micro on the void ray, which was the focus of the entire build.)
It's certainly interesting to see regardless, and I'm surprised at terrans overall success.
|
I guess I should stop voting on protoss when in doubt on liquibet. Probably explains my bad stats.
Hope it all evens out. Crazy in the Foreign vs Korean difference.
|
The international one looks -VERY- promising... The Korean one not so much, haha...
I'd like to think that the Korean one is so off due to the fact that there's not enough data and it's only focusing on a small group of players at the veryveryvery top level of the game, while the international one is more widespread and generalized view of top players.
I'd like to see game counts per datapoint on these... I'm sure the massive skews in the Korean data is due to the fact that some of these plot points only include like 20-40 games of data.
|
This is extremely interesting.
Is it at all possible to put the patch release dates on the graphs? Might be asking for a bit too much I guess.
|
I posted on the Reddit post but I thought I'd post here too: even though these charts indicate balance (or lack thereof) in certain ways, they're taken only for certain time periods which basically means certain events or specific players have disproportionately larger impact on the data.
For example, Naniwa's MLG run (I know it wasn't charted but it's a good example) or recent Losira domination in GSL would both skew results and make it seem like P is OP or Z is OP when really, neither can be conclusively decided as true.
Basically, players who make deep runs play more games and skew it. So what we sometimes see are only the best of the best represented their races. For example, the GSL currently has 4 zergs in the RO16. These zergs are the absolute best when it comes to their race, as opposed to mediocre protosses and terrans who made it in. This makes it look like Zv* is insane when it may not truly be, because the fewer number of zergs and their comparative skill means they'll obviously win more games.
Similarly, I imagine if IdrA continues his streak into NASL further and in other tournaments, he will single-handedly make zerg look OP in NA even if he's the only zerg playing major tournaments and getting far.
|
Could you elaborate on what the graphs actually show, like for example how far back in time does the the win percentage in a given point include game from?
Or did you just calculate monthly win percentages and then interpolate the points?
|
Very cool, thanks for putting it together.
|
When i saw the first graph i though "good job blizzard, you may be on track" then I saw the korea graph and tried to see why it was that way and this is what I came up with
Koreans protoss generally stick to their guns in the sense that they will continue to do a BO untill it starts losing, and once it starts losing they will start doing a BO that counters the counter which explains the constant back and fourth of the protoss charts, as you can see after the drop in protoss wins they come back to have >50% win rate very soon after. This is simply development in the game.
Now in both charts Z has never had a good ratio vs T. Oddly tho the 2 charts are inverses of each other, when then international zergs do worse the Korean zergs do better. Personally I believe that the zerg generally has a low win rate vs T because T has alot of things that they can do which are not even close to all ins but can result in insta loss for the zerg. Also terran is simply much more forgiving endgame as opposed to zerg, if zerg loses 1 engagement late game they just lose the game.
my thoughts, take what you wish
|
lol right when the amulet nerf and infestor buff happened toss went to shit in korea
|
This is super interesting, I especially like the graphs where the winrates switch places oppositely and then start to even out. It really tells you about the math behind the metagame...
|
Another thing to note is that key protoss players like MC have been knocked into up/down in GSL, so they haven't really had many games in Korea lately. Since the KR TLPD is based almost all on GSL, that means there are basically no protoss games to chart.
So what we have are dozens of terran games and maybe 4 protoss matches.
|
|
|
|
|
|
EPT
