Hi,
here's an idea. When you look around all the hereos matchups you will see it's basicly the same list that the general ranking of heroes in terms of W/L ratio.
Knowing that eg. Spirit Breaker's worst matchup is Centaur says realy nothing becouse of Centaurs insane W/L ratio in general.
Why do not clear the matchup data comparing it with the hereos W/L ratio?
I mean lets look how is Cent's percentage of win diffrent when he plays against Spirit breaker then from his general W/L ratio?
Normaly Cent got 62,92% win ratio.
Against SB 61,63% win ratio
That gives us (atm) +1,29% (lets call it the Diffrence Ratio) of winning with Cent if we play not a static random hero but SB.
When we count DR like that for all matchups, we order it and get a list of hereos that playing SB changes most the chances of wining or losing --> worst and best versus (some might say). This way i assume we make a more reliable ranking of matchups for all hereos.
I know the system I suggest is not perfect and it depends on the interpretation to say if it's a valid system or not, but you can give your own suggestion how it can be counted or imporovments, any way in my opinion the problem of nothing-saying matchup ranking should be dealed with.
Hi,
here's an idea. When you look around all the hereos matchups you will see it's basicly the same list that the general ranking of heroes in terms of W/L ratio.
Knowing that eg. Spirit Breaker's worst matchup is Centaur says realy nothing becouse of Centaurs insane W/L ratio in general.
Why do not clear the matchup data comparing it with the hereos W/L ratio?
I mean lets look how is Cent's percentage of win diffrent when he plays against Spirit breaker then from his general W/L ratio?
Normaly Cent got 62,92% win ratio.
Against SB 61,63% win ratio
That gives us (atm) +1,29% (lets call it the Diffrence Ratio) of winning with Cent if we play not a static random hero but SB.
When we count DR like that for all matchups, we order it and get a list of hereos that playing SB changes most the chances of wining or losing --> worst and best versus (some might say). This way i assume we make a more reliable ranking of matchups for all hereos.
I know the system I suggest is not perfect and it depends on the interpretation to say if it's a valid system or not, but you can give your own suggestion how it can be counted or imporovments, any way in my opinion the problem of nothing-saying matchup ranking should be dealed with.