Jerome has implicitly cited that this PAM has a good track record. However, it's important to know what that track record consists of. In particular, predictions about some kinds of events are not necessarily indicative of predictive power about other kinds.

The big problem with predictive markets is all about how many people (and of course on some issues, who) are actually doing the buying.

Thus, in this case, are enough people actually buying and selling on the Iran issue to make the prediction worthwhile? (Perhaps a comparison with trading volumes on the Iraq issue?)

Likewise, just getting the Iraq issue right is not in itself massively predictive. At what point did it get it right? Was it really more predictive than $observer? etc.