[{quoted}](name=Miku Hime sama,realm=EUW,application-id=3ErqAdtq,discussion-id=mPKN90v2,comment-id=00010002000000000000,timestamp=2015-01-23T06:49:52.461+0000)
I feel we're on the same page, but win-rates are amongst the LAST stats they should bother observing for information. It has extremely low relation to a champion's overall health. Only an EXTREME outlier is of any real significance, that champion has maintain both absurdly high win-rates AND pick-rates simultaneously in all elos, especially amongst the higher ones. Of which there currently are NO CHAMPIONS that come remotely close.
Even if someone so absurdly powerful were to appear, it would become painfully apparent for the players long before it showed in their win-rates. Ultimately making the point moot before it even has chance to prove itself.
Win-rates do nothing more than fool people into making false uneducated judgments.
not saying win rate should be the sole or primary factor considered, but win rate to me says a lot about a champion's state. there are a large number of champions, who are played in only 1 role, that have under 47% win rates. when a champion wins 47 for every 53 he loses over the course of several million games, something's wrong with him. and the thing is, their underwhelming states are apparent in game and are also reflected in their pick rates. consider eve, tf, twitch, elise, zac. their inability to influence games is obvious, at least in my games, and is reflected in both their win rates and pick rates. so i'm saying, since in game weakness can be reflected in win rates, a substantial amount of consideration has to be given to win rates.
trying to optimally balance a moba with well over 100 champions in 5 different roles is understandably incredibly difficult. the best way to go about it is probably by trial and error and observing results. win rates are the clearest and most direct results available to the balance team, so they should be used, but it doesn't feel like enough consideration is being given to those results.