[stigmata]

I thought a more interesting hypothesis to test would be the idea of online collusion.

E.g. players working in teams. I think this has a far greater chance of turning up something interesting than checking if the sites are cheating (which they clearly arent).

However this is not as easily achieved....

I guess the way I would do this is:

1. Generate a large pokertracker database. This could be achieved in two ways: A) Ask nicely on this forum and combine peoples databases B) Watch multiple tables on Party, and import Hand Histories to POke Tracker before you close down.

2. Query database to find multiple instances of players playing at the same time. This limits you to players who are either Friends or Colluders (or both).

3. Test playing strategy for two things A) Trapping players between the two hands. E.g. The two (or more) players show significantly more aggression when in a pot togethor than would be expected. B) Best hand plays e.g. The two players only play the best hand between them.

I guess the problem with my suggestions are that it is a far bigger project and is not as statistical (more of a database/programming) exercise. So not much use to you, although it does have a statistical element....