Very Important Players

For me the ideal formula (let's call it [i]gamma-predominance[/i]) should be on this way [b]Number of opponents played[/b] in the last session (2 or 3 months) OP = (NEW_NUM_9P_GAMES - OLD_NUM_9P_GAMES)*8 + (NEW_NUM_8P_GAMES - OLD_NUM_8P_GAMES)*7 + ... + (NEW_NUM_3P_GAMES - OLD_NUM_3P_GAMES)*2 + (NEW_NUM_2P_GAMES - OLD_NUM_2P_GAMES)*1 It's like counting the games played in the last session but in this way a 9-player game counts 8 times more than a 1v1 game. I think it's fair, isn't it? [i]Note: These data are shown on everyone's profile page but (I guess) there is no record of them, the structure behind needs to be created. For those who have passed 2500 games these data are currently incomplete and useless for this task.[/i] So finally, sqrt(OP) tells me how active a player is, RATING tells me how good a player is. ([i]Current rating[/i] or better [i]last known rating[/i]) Let's just multiply them: GAMMA_PRED = RATING * sqrt(OP) I don't care to normalise it into a specific range, because the ranking doesn't change, but yes, it can be done. I would run it every 2-3 months just after a D12 list is run, because it's when players' ratings become public. So I think the ideal is to have a button for running the [i]gamma-predominance[/i] when we think it's the right moment. Note: all numbers of games played (grouped by number of players), of all users, have to stay recorded to the next session!