I was very interested in a recent post here on Gamasutra: Avoiding the "Tagalong Trap", by Catharina Bøhler. Catharina was interested in identifying what are the components that create a meaningful cooperative experience in cooperative games.
My interest was piqued by the fact that I've been researching this very same topic, but from a different perspective: I've focused on cooperative boardgames.
Catharina's post shows that our findings converge in several ways. I believe that there is one other significant difference between our complementary perspectives, besides the difference between digital games and boardgames. She is looking for components that make up a "good co-op experience", whereas I was looking for components that would introduce cooperative dynamics even in competitive games.
Thus, Catharina focuses on games where all players are in the same team. But cooperative dynamics are also present in multi-team games — which may also offer the possibility of players changing teams mid-game, either covertly or not ("traitors").
There are two components that have great impact on how cooperative dynamics unfold in a game: information about the game state and communication between the players. Controlling the display of the game state is commonplace in digital games, but enforcing communication restraints is almost impossible. But enforcing communication restraints in a boardgame is also impossible: even if the rules say something like "you may not speak to your partners", there is no game police beside the table, and players can speak to their partners.
One of the crucial difference between digital and non-digital games is very present here. In digital games, the game rules are enforced by a neutral party (the software); in non-digital games, the rules are enforced by the players themselves. I have the feeling that there are more players willing to cheat against the software than there are players willing to cheat against themselves.