Bwog resident nerd Zach van Schouwen fakes his way through the game theory of Google Chat.

When GChat rolled out its “Go Invisible” feature last week (which has been in other instant messenger clients for awhile), the whole landscape of online chat shifted a little. While invisibility on AIM is one thing, most GMail users have a window open at all times. “Go invisible” takes away the responsibility of being constantly connected.

The problem is that the ability to go invisible is predicated on the assumption that no one else behaves the same way. If everyone else naively leaves their status lit up in green, you can pick-and-choose the conversations you want to have, an optimal outcome for you, and a betrayal for everyone else. However, if they’re rational, they’ll realize the advantages of invisibility takes part.

Of course, people don’t behave naively in practice. In fact, almost everyone else is probably as awkward and antisocial as you are. As a result, invisibility “strictly dominates” participation, particularly since being the only non-invisible one means you’re being watched constantly. In the short run, the equilibrium is for every user with any misanthropic tendencies to withdraw to invisibility mode. Fortunately, chat is an iterative problem, so eventually the social dead zone will correct itself, finding a new equilibrium where more people opt into the system after it bottoms out.

Yesterday I logged in to find only two people online on my list, which traditionally had about twenty signed in at that time of the afternoon. Am I just unpopular? Maybe. But I’m pretty sure game theory is taking over.