Actually, I would like to see for such collective events (senate races, oscars etc), to see a “meta-market” in each exchange:
“How many events is our exchange going to guess correctly?”
In principle, it should be possible to price correctly such contracts using the reported probabilities and their dependencies. This should give some extra attention to the fact that markets are not supposed to get everything correctly to be accurate.
Panos Ipeirotis’-s blog [*] and website
[*] Great name (”-Behind The Enemy’-s Lines”-) —-but where does the name come from?