...and if you're right you don't win money--you get a ticket to the game itself?
Read about it. (NY Times)
Joe Kocott and Ian McKinley are both die-hard Pittsburgh Steelers fans. In the past week, Mr. McKinley urgently searched eBay, Craigslist, StubHub and other sources for a ticket to Super Bowl XL, eventually buying one for $2,500, while Mr. Kocott secured tickets for himself, his wife and seven children, and nine others in early January by paying an average of $350 for 18 futures contracts that promised him tickets if the Steelers made it to the championship game.
Futures markets have existed for decades for commodities like pork bellies and oil, whose prices fluctuate over time. The advantage of a futures market is that it allows the buyer and seller to lock in a price, and in that way reduce uncertainty.
There is great uncertainty surrounding sporting events like the Super Bowl, including whether a fan's team will make it to the game and the price of tickets on the secondary market. Stephen K. Happel and Marianne M. Jennings of Arizona State University proposed a futures market for tickets to major events to reduce risk in 2002. Their idea is finally coming to fruition.
As the article by Alan Krueger explains, one of the difficulties of a market like this is that you need to be able to trust the marketmaker to be able to deliver on the tickets. Krueger suggests that sports leagues make some tickets available through these markets.
You would think that someone would have thought of this before. (When I read the article, I admit I had the immediate "why didn't I think of that" response.) I'll bet a few people thought of it and quickly dismissed the idea because of the problems associated with gaining the trust. Trust is a costly thing to establish. But if people don't have any way of knowing that the marketmaker can come through on his end of the bargain, they are going to be very reluctant to put down hundreds of dollars on such a deal.
It's a good lesson for economics students (or anyone) to see the importance of trust and enforceability of contracts.

Can we build implied probabilities of team success in reaching the Super Bowl along the same lines as the interest rate graphs that David Altig puts up at MacroBlog --- Implied Probability of Steelers to Super Bowl at TimeX = $SteelersOptionTimeX/(Sum(PRICE Each Time at Time X) --- this would be a pretty interesting prediction market if it is deep and trusted enough.
Fester
Interesting question. I don't think it would be quite the same however. Super Bowl tickets, unlike fed funds futures, have intrinsic value that could even depend on who is playing.
Las Vegas odds would probably do a better job of giving you the implied probabilities. So I'd turn it around... If you could get the implied probabilities from Las Vegas, and you had the futures prices, you might be able to back out the willingness to pay for tickets under different outcomes. Now that would be interesting.