Sports Betting as a Prediction Market

January 6, 2008
Posted by Jay Livingston

There’s been some discussion of prediction markets at sociology blogs like Statistical Modeling and Scatterplot. Prediction markets are like stock markets: in stock markets, investors are betting on the ultimate price of a stock; in prediction markets they are betting on the outcome of events like elections, Academy Awards, American Idol, or when the US will start withdrawing troops from Iraq.

The collective wisdom of the investors is reflected in the price of the different options. If you want to buy shares in Obama as the Democratic nominee, you have to pay a higher price now than you would have three days ago.

The question is whether that collective wisdom has more predictive power than do the so-called experts.

Sports betting is essentially a prediction market. The odds or the betting line reflects the collective wisdom of the bettors (or “investors” – the term applied to people who bet on stocks). I blogged about this a year ago, and I came down on the side of the experts – the bookies and oddsmakers who set the initial line.

Yesterday’s NFL games offered a good example (i.e., anecdotal evidence) of what I meant. The oddsmakers made the Seahawks a 4-point favorite over Washington. But the public bet Washington, and the line came down to 3. The Seahawks covered easily, winning by 21. The crowd lost.

The Steeler game was an example of the worst-case scenario of following the “wisdom of crowds.” The Jaguars opened as 1 point favorites. The public bet them heavily, and by game time, the line was 3. If you had bet the Jaguars early in the week, you would have given up one point. But suppose you had waited to see wisdom of the crowd. You see the line going up, you see that the collective wisdom is heavily in favor of the Jaguars, and on Saturday you put down your bet, giving the Steelers 3 points.

The Jaguars won but by only two points. Following the wisdom of the crowd turned a winning bet into a losing bet. (That was cold comfort for us Steeler fans, who may have won our bets but saw our team lose a heartbreaker.)

In today’s games, the 3-point line on the Bucs and Giants hasn’t moved, at least not as of this morning. But in the Chargers-Titans game, the public has been favoring the Chargers. The line opened at 9 and is now up to 10 at most bookmakers. If you believe in the wisdom of crowds, you bet the Chargers. If you think the oddsmakers are smarter than the public, you bet the Titans plus ten points.

4 comments:

Anonymous said...

A very good article and one that could benefit a number of sports bettor's out there. I have been sports investing for three years and agree with your analysis. A contrarian investment strategy is often times the best approach. "Fading the steam" is one of the major angles that I incorporate in my decision making process. Following the herd can be dangerous in this market!

OilyDave

http://microsportsinvestingformicroloans.blogspot.com

jeremy said...

The hard counterargument, though, is that if the experts are better, why don't they go clean up on the prediction markets? Aren't experts interested in making money? Or is the problem that prediction markets aren't high enough stakes yet?

Jay Livingston said...
This comment has been removed by the author.
Jay Livingston said...

1. Sports betting as a prediction market has the interesting structural feature that the “experts” – the oddsmakers and bookies – are also the brokers. They do clean up, at least enough to stay in business, though of course much of that cleanliness comes from the vig.

2. It’s possible that the smart money does clean up, especially by taking contrarian advantage of line shifts. It wouldn’t surprise me if some smart bettors took the Jags -1 on Monday expecting the line to move, then hedging back with the Steelers +3 on Saturday. They don’t hit a middle like that very often, but they can use the line-shift to hedge a position, with the possibility of catching a middle or an “edge” (a win on one side, a push on the other).

3. Sports betting may be different from betting in other prediction markets. If, as you say, the price on Obama started rising on Thursday before the poll results were known, I’d be very curious to know who those Obama bettors were and how they decided to buy those shares. Perhaps it’s a case of the smart money moving the market and then many others following figuring that someone must know something. That sometimes happens in sports betting. But generally, when you see a line shift in football (NFL or big NCAA games) the day of the game, you’re looking at the “wisdom” of the crowd, not the smart money.

(p&e)