Inkling Markets is a startup based in Chicago that lets you set up "prediction markets" to forecast future events and earn a few "inkles" if you are right. There are markets on whether Harry Potter will live or die in the last book, the average March temperature in Chicago, and the Oscars. You can register and set up your own market at no charge, or contract with them to build your own private or public marketplace for your organization.
I'm running a few Ann Arbor specific markets, including one to predict the number of people who are at my weekly Thursday a2b3 lunch at Eastern Accents. (Last week 16, this week predicted 16.5; the poll at lunch last week of attendees was for 13-14). Some folks at the U of Michigan School of Information are running a market on who will be the next dean of SI.
One more specifically of interest to libraries and patrons is forecasting the number of holds on books in order to get some sense for velocity of books returned so that you can guess when you'll get your copy of a busy book.
Whether or not you are a patron of the AADL, you can make a prediction on the number of holds on the revised edition of Freakonomics as of 3/1/07. As of this writing, the actual count was 12 holds on 5 copies in circulation, with one book coming due before the market closes; the prediction is for 16 holds, which I guess takes into account my ability to publicize the market before then and see if people can forecast actual results.
The short term market should be entirely demand driven. In the longer run, the library can be an insider by buying more copies and putting them into circulation. The stated goal is no more than 4 holds per book in circulation, and it takes 2-6 weeks to get new copies into the pipeline once a decision has been made to buy a new copy.
I'm practicing on this market because I want to get things set up to forecast the number of holds on the new Harry Potter - with any luck that will be of actual use to some library somewhere.