Anonymous games surely just makes it impossible to know how many games your opponents have played with each other. A good root around on the old profiles is always my first port of call before I join a game.
I can't be bothered to read back and get the right name (think it was Draugnan) but I agree with the suggestion that there should be penalties for unsubstantiated accusations.
It's only practicable though, if a clear set of guidelines as to what does and doesn't constitute meta-gaming can be established, and enfringements confirmed or repudiated.
I don't actually think that's possible, and would instead favour measures to make meta-gaming less profitable rather than going down the banning route.
The half points for draws could be one thing, perhaps in each game the number of games a player has palyed with each of the others could be displayed as a string of coloured numbers.
e.g. Major Malfunction (10) 0 0 0 10 0 0 ; )
Wouldn't take up too much space and should be relatively straight forward to pull the data. Though now I think of it, that might be a ball-acher of a query to get right.
And to wrap up this unintentionally extended ramble...
On the issue of challenges/accusations, I suggest a three outcome system. When a challenge is made it should be raised as a query, rather than worded as an accusation, and on examination it could then be either upheld, unsubstantiated, or deemed to be unreasonable.
Upheld should be points off rather than a ban ( I always think banning just encourages them to come back as a different user), unsubstantiated is no penalties either way but with the onus being on stating clearly that the challenged player has nothing to answer for and hasn't done anything wrong, and an unreasonable challenge should see points knocked off the person who challenged.
I personally felt okay querying the community on the statement Germany made, but I would. If it was generally felt that i'd made an unreasonable assertion I would humbly accept my penalty and ask myself how I got so adrift from the general consensus.
Sorry, end of sermon.