Thursday, November 28, 2002
The Disposable Income/Blackout Hypothesis of Point Spreads-I've been successful in the prognostication game over at Spudlets this season in large part of picking the occasional home underdog (here's this week's predictions, where I've got a lot of home dogs picked) , being able to pick a good number of upsets. At first, I thought Vegas was undervaluing home field advantage, but the following theory came to me. [A quick primer for the sports-illiterate- a point spread is the projected amount of victory for the favorite. In football betting, the favorite has to win by more than the spread in order to win a bet on the favorite. For instance, New England is a six-point favorite over Detroit this afternoon. If the Patriots win 20-17, the people who bet on Detroit would win the bet, for New England won by less than six, failing to "cover the spread."] First, lets review how the point spread is determined-it's not the thinking of a few hotshots in Nevada but the "price" that clears the market, getting half the betters picking one team and half picking the other. That way, the bookmaker pays off the winners with the loser's money and pockets his commission without any risk. This might not reflect the true strengths of the teams, but it's the figure that gets betters ambivalent about placing a bet either way. Assume for the moment that there are two types of betters, the homers and the handicappers. Homers want to bet on their team, while handicappers will bet if the point spread is out of line, regardless of whether he's a fan of the team he's betting on. For the homer, betting on sports is part of his entertainment budget, as are tickets to the home team's games. Let's look at the New England-Detroit game as an example. Joe Southie is a Pats fan who'll go to the game on a regular basis; he's not a season ticket holder, but makes it down for a lot a games. With the Pats out of town, he's got some extra cash to spend this week, so he lays a bet on one of those offshore betting sites (or with the local bookie at the pub) on the Pats. He has a opposite number, Sam Shoprat in metro Detroit. He'll plunk down a 20 on the Lions, but he's heading down to Ford Field before consuming large quantities of strategically-burned dead avian flesh. With his pockets a bit empty from getting the ducats to the Lions game, Sam's going to pass on betting on the game this week. Thus, the road team's homers will have more spending money than the home team's homers; the homer money will thus tend to favor the visiting team. This should move the point spread in the favor of the visiting team. Won't the handicappers step in to correct the spread? Yes, but only when the spread is off enough to make a bet worthwhile. If the bookies get a 5% commission, the spread will need to favor the home team at least 55% of the time for the game to be a good bet for the handicapper. This might throw the point spread off a couple of points in the direction of the visiting team before the market intervenes to take advantage of the mispricing. You might also have an additional away-team factor in that a home-team's games are blacked out in the host city if they don't sell out 72 hours ahead of kickoff. Teams with poor attendance thus will not be able to have the home fans watch the game. Inability to watch the game will tend to lessing demand for gambling on the game, for the fun of seeing whether you team win is lessened if you have to listen on radio. That's my theory, open to anyone who wants a quick quantitative-analysis term paper topic to run the traps on. If true, the discrepancy would be a function of walk-up sales of each team (for season tickets paid for in advance wouldn't have this effect) and whether the game was blacked-out or not.
Comments: Post a Comment