Some interesting ladder numbers from sc2ranks. - Page 2

On July 02 2012 07:15 EnE wrote:

@Elldar:

Yeah......you seem extremely ignorant, just from how you act like GM is like another league or even like masters. Here is the difference: GM is the top 200 mmr. You can't compare it to another league because there can ONLY EVER BE 200 people with top 200 MMR. The more skiled the people are, the higher the top 200 MMR ceiling will be, but it will always be an OBJECTIVE STATISTIC.

On July 02 2012 07:19 Cokefreak wrote:

What about all the pro's that don't sink countless hours into laddering but instead practice in custom games and therefore are not in GM? GM doesn't really mean anything, it's just a way to show off for some not that popular players.

This matters because the fact is, even if the GM players arent all doing good tournament results, they ARE all top 200 players on ladder. Therefore, they are the best by the most and only objective standard we have. Ladder is the only thing everyone plays. It has enough players for the top 200 mmr in the ladder by race to be very relevant.

On July 02 2012 07:21 monkybone wrote:


How can you say that when their skill (MMR) is determined by who they win against? Imagine for the sake of the argument the ladder had only two races:
A: Terran
B: Terran with 90% health

People playing A will stabilize with an MMR which gives them a 50% winrate against both A and B. The reason is that any mirror matchup necessarily will be 50% for each player on average, so the other matchup (A v B) will 50% for each player on average as well in order to maintain a total of 50% winrate for each player. So we end up with the situation where some proportion of players are playing Terran, and some are playing Terran with a severe disadvantage, while the race winrates are 50-50. So people playing B will be more skilled at the game than people playing A even though they have similar MMR.

The same principle applies (in a more complicated manner) where we have three races A,B and C and some races are more powerful than others. However, what is symptomatic for the situation described above with only A and B is that the higher leagues would have an underrepresentation of B players because of the racial imbalance.

On July 02 2012 07:12 darkness wrote:
I don't think Blizzard care about "interesting" numbers too. Back when MC was complaining, there were, I think, 6-8 terrans in top10 gm in KR. Nothing happened.

On July 02 2012 07:21 EnE wrote:
they are the best by the most and only objective standard we have.

On July 02 2012 07:28 buckKeefe wrote:


ladder ranking is not a measure of skill, which is the relevant variable for considering game balance. any data you draw from the ladder is confounded by in-game race anyway. i'd go so far as to claim that you can't use this tiny subset of the ladder to measure balance at all. imbalance would presumably be defined as something like "differences in win % by race controlling for skill." the problem is that the ladder's measure of skill would be mediated by race to begin with if the game were imbalanced, so you'd never get an accurate answer. BTW i play terran so i should be relatively unbiased here.

On July 02 2012 07:15 EnE wrote:
This is very relevant . Having 6 terrans in top 10 is COMPLETELY different from having 38 in top 200. Just simply worlds apart.

The GM means whoever had the HIGHEST MMR. It's literally the top 200 highest MMR who are put in the league. Therefore if only 38 of the top 200 HIGHEST MMR are terran it says infinitely more than some of the top 200 winning a few more games than other people in the top 200.

@Elldar:

Yeah......you seem extremely ignorant, just from how you act like GM is like another league or even like masters. Here is the difference: GM is the top 200 mmr. You can't compare it to another league because there can ONLY EVER BE 200 people with top 200 MMR. The more skiled the people are, the higher the top 200 MMR ceiling will be, but it will always be an OBJECTIVE STATISTIC.


This matters because the fact is, even if the GM players arent all doing good tournament results, they ARE all top 200 players on ladder. Therefore, they are the best by the most and only objective standard we have. Also, European ladder is the most relevant ladder... why? It's the most competitive ladder apart from Korea and Korea is a far more isolated system with a mere 50 million people living in it... 19% being terrans in GM is bad.

On July 02 2012 07:30 DarkInfinity wrote:
This has no statistical meaning, there are countless reasons why this doesn't represent the balance of the game.