Can eBay Predict Superbowl Champ?


Greg Isaacs, who runs eBay’s developer program sent me an email this morning with this most interesting link. “One (loose) predictor of who may win the Super Bowl could be the value of Seahawks versus Steelers merchandize on eBay,” he writes.

Mpire, one of eBay’s developers has developed a little web tool, that takes eBay market data and tries to correlate current activity on the online auction house with the forthcoming Superbowl championships. So far, Seahawks are getting better pricing, versus Steelers, hinting that the team from the Northwest might actually win this time around.

I think, an ideal mash-up would use income-and-Internet adjusted market data. After all with all the Microsoft up in Seattle, you can easily expect more spending power and more Internet penetration. Pittsburgh fans, I bet are a bit more old school, and likely to do their buying in the analog world.


Om Malik


my favorite sport happens to be in New York, not in India. and yup, I agree, the population and economic posperity cannot match the “on the ground” reality.

ebay-mash-up is a different take on the age old trick – calaculating the odds, though i figure in the end that doesn’t make sense.

if that did, then the Yankees would never lose, but do, more often than the numbers suggest they should. sigh..



I agree with you on the fallacy of this thinking. Otherwise, I suppose we all can will the Sun to rise in the west. Or to use Om’s favorite sports, India will win every match, on the thoery that its fan base is more populace.

Matt Rogers

A perfect example of the utter fallacy of economic thinking. Regardless of how people economically value a team what counts is the hundred million unknowable variables that happen on the ground and that are unknowable until the game is played. Things like how are all the players feeling? Will someone trip and have some sort of accident, will the weather favor one team or other, blah, blah. Economists live in models that have little connection with reality.

As the dot com bubble showed the vaunted “market” is only all too fallible, probably the housing bubble will be the the next example of overinflated markets wrongly predicting the future.

I welcome comments at raven200 at

Om Malik

Of course, no one can argue with those spread makers. i agree gambling sites do tens to be more accurate. still, i though this was an interesting experiment worth sharing with others.

Sam O

The Superbowl is an odd example for this data because there is already a direct market for predicting the winner of the superbowl: gambling sites. A site with high volume, and person-to-person instead of a person-against-company, like, will be a far better predictor.

Comments are closed.