I’ve noticed a growing interest in predictive market sites. The old standard is the Hollywood Stock Exchange, where people can buy “shares” of a star or movie, whose market value rises and falls as they gain fame, notoriety, or invisibility.

Owise seeks to build on this idea by creating a “neural prediction community” of users who give their best estimates about future events, e.g., the likelihood that President Bush’s second Supreme Court nominee is a woman. Owise then calculates the probability of a certain outcome based on the estimates of their members. Members who successfully predicted results in the past are weighed more heavily, so unlike a poll, Owise skews towards what “experts” think, while not relying solely on the analysis of a single individual.

The events that Owise seeks to predict outcomes for, however, are chosen by Owise staff (with suggestions by users). It might be interesting to let people create their own questions, too, using Owise’s backend and a small script to run an Owise “sticker” on their own sites. For example, a bread company could ask “What’s the probability of another low-carb craze this year?” Online polling company Zoomerang has a similar program, whereby individual members can create their own polls for private and small business use, but they also work with corporations to do formal market research using Zoomerang’s database of people who agree to be polled. Letting people create and categorize their own predictions also ties nicely into the whole social networking trend.

There is, as it turns out, an open-source package, Idea Futures, for anyone interested in building their own predictive market. (Thanks to Populicious for the sighting.) Even Google is getting into the act, using predictive modeling to gain insight into business issues such as product launch dates.

I predict a slightly better than even probability that Google will buy Owise. Google obviously sees the value in such technology, and it could give them an interesting add-on functionality for Google News (which may even come out of beta in our lifetime). Reading about the John Roberts hearings? What’s the probability a third Supreme Court seat will come open within the next year? Google could also serve requests for predictions as part of its AdSense feeds, thus easily expanding the number of people predicting, and thus increasing the accuracy of the system. Sure, Google could build their own system (or use Idea Futures), but Owise already has a user base of people who have earned “weight” by their predictions, and this could let Google hit the ground running. (Full disclosure: I have a cousin working for Owise, although he has never discussed with me the likelihood or not of any potential sales or partnerships.)

Will we see more experiments in predictive modeling in the months to come? You can bet on it.