This may or may not be of interest to this audience, but there is a very interesting discussion going on right now over at my blog regarding the use of game theory to understand the current dynamics in Iran. A colleague of mine, Andrew Little, wrote down a very simple model of election fraud that has sparked some good debate, so I wanted to alert your attention to it should the topic be of interest.

Here is a brief on Andrew's model:

In this game there are two players an incumbent I and an opposition O. These players are bargaining over some policy space (we’ll say it’s the [0,1] interval, though it is not particularly important), which represents the amount of concession the incumbent is willing to give to the opposition (sound familiar?). The game unfolds in the following manner:

1. Nature reveals private information to I about their strength should conflict occur.
2. An election occurs. Nature reveals the results of the election to I.
3. If I loses, they choose to reveal the true results to the opposition or commit fraud to make it appear that they won.
4. The incumbent makes an offer x to the opposition.
5. If fraud was committed, the opposition detects the fraud with probability pf.
6. The opposition chooses to accept the offer or reject it, leading to conflict.

The opposition’s decision to accept the offer from the incumbent or have a conflict is based both on what they observe (a victory or loss for the incumbent) and their belief about whether the incumbent is a strong or weak type (i.e., their repressive capacity). The election results matter because they shape the opposition belief about this type. Andrew went on to discuss the pooling and separating equilibrium of this game, and while this discussion is a bit too detailed for a blog post, the consequences of these findings are very interesting.
For more on the interpretation of this model for Iran, and the ensuing debate please check out the full post.

Hope you enjoy!