Think that through. Take the two high and low cases. Player 1 goes 10/0/10 against pro teams, but loses every game. Player 2 goes 10/0/10 against B5 players, but loses every game. Stats alone do not suggest a difference in the ability of each player, but the competition they face does suggest a variance in their ability. Now, let's say that player 2 (the one in B5) wins 66% of the time, but is so unlucky that the losses always occur in the promos. Would you say that player 2 is better than player 1, since player 2 is the only one that has wins?
NOTE: we can reduce the data set enough that the plausibility of player 1 remaining in pro contention is legitimate.
This is how I figure it. Everything requires context. Game statistics other than win/loss are a good evaluation of a player when in the context of the competition. Although, win/loss does matter when comparing the results of similar challenges, it does not give relative player strength across greatly diverse challenges. More diversity or difficulty makes the comparison of wins/loss less accurate in assessing player rating. Game stats are more accurate than win/loss record when calibrated by a difficulty factor.
