[Valerius]

I do think it makes sense to break large games (maybe 12+ players?) out separately from smaller games.

For one thing, it just seems fair. Take two players of equal ability, equal winning percentage and putting equal time into the game but one prefers small games and the other large ones. Assuming they both face an equal level of competition the one playing small games will record more wins because his games play faster.

I also think large games are a better measure of overall skill. Knowledge of game mechanics and tactical and strategic sense are certainly tested in small games but there's some factors that I think get tested much more in larger games:

* In terms of planning, in a large game you will have to prepare for and deal with a larger variety of situations. For instance, in small games there's more of a chance of not having to face heat/cold nations early or maybe there aren't any heavy bless nations in the game, etc.

* Also, in a smaller game there might only be two nations that complete research and are wielding the full array of endgame magic, sometimes you might not even get to this stage. In a large game you will likely have several opponents capable of doing this and so it more fully tests your ability in all phases of the game.

* In a small game a single alliance can be all it takes to get those nations into a winning position. In a large game you'll likely need a coalition (obviously assuming it isn't a no diplo game) and if you can keep that coalition intact while you are the one benefitting the most, well that's an accomplishment right there.

* Micromanagement. People don't really mention this but, perhaps because it's my worst weakness, I think being able to focus and not let your play get sloppy when you've got dozens or even hundreds of commanders and thousands of troops to manage in battle is a real advantage. Zeldor's megagame win isn't impressive just because of his skill as a player but because of the insane tolerance for micro it implies.

I'm sure there's more but that's just some stuff that comes to mind. Of course it's not perfect - nobody is going to think winning an 8 player game filled with Micah clones is less impressive than winning a game twice that size filled with Valerius clones. But unless you're going to audit each game for the players involved, perhaps even the circumstances of the win, this seems like a reasonable solution that hopefully is possible with the information you've already compiled so there's no extra work involved.