Quote (krova @ 16 Oct 2015 15:27)
that was a cool website
Quote (Mastersam93 @ 16 Oct 2015 19:23)
Could use some kind of sum of Baye's theorems.
Like, chance of winning 1st game with xx win rate + chance of winning 2nd game given you've won the 1st game at xx win rate.
Honestly, I think it'd be much easier to just get data on players grinding arena, and find their 1st game win%, 2nd game win%, etc, and average them out.
That would be more accurate, but a couple of problems here.
#1 win rate fluctuations would be different for different players and playerstyle, so an average wouldn't give a real answer anyways
#2 the formula would be way more complicated