I assume this is Werewolves of Miller's Hollow that we're talking about.
I happened to glance at the results for Werewolves game #52647941, which was concluded earlier today, and saw a glaring discrepancy in the way the changes in ELO rating were distributed to the 9 winners of the game. Most of the winners gained points but the top-rated winner PGpirata lost 27 ELO points while, for instance, winner Architorture gained 46.
What PGpirata's ELO adjustment should have been was to add the 8.78 gained against Werewolf team member cream to the 6.67 and 8.71 gained against fellow wolves Cajunworld and shayan_1047 for a total gain of 24.16 ELO points. Instead, the software lumps in a loss of 51.80 points that he "lost to his 8 teammates because he 'tied with them'
for a net loss of 27 points.
How hard can it be to direct the software to ignore ELO considerations against fellow team members and only compute ELO changes against the players from the opposing team ?
Seriously, how hard is that ?
And then take the code that was hammered out to ensure that ELO calculations in a team game like Werewolves of Miller's Hollow are handled properly and then apply said changes to other team games like Not Alone and 2v2 Innovation ?
Cause i really want to try it!
