No replies to this topic

[Symbiodinium]

I'm seeing a problem where there is almost always a disconnected (or missing) player at the start of the game. More often than not, the team that's down a mech loses. If we're going to base player matchups on Elo scores that only take into account team win/loss records (and not individual metrics), it seems to defeat the purpose if disconnected players destroy the validity of the games. It's just randomly throwing wins and losses onto your record depending on which team is short one (or more) mech(s), homogenizing everyone's Elo score.

Solution: If a match starts with an uneven number of disconnected players, the effects of that match on players' Elo score is invalidated. If both teams have the same number of disconnects, the match is valid. Alternately, only invalidate the result for the losing team.

Yes, I know that this opens a possible avenue to manipulate one's Elo, that's why it only counts if the disconnect is before the start of play.