Monday, February 25, 2008

Updated MLB Projected Win Totals

Baltimore 71.6
Boston 89.1
New York 96.6
Tampa Bay 86.7
Toronto 78.5

Chicago 78.8
Cleveland 91.6
Detroit 88.8
Kansas City 73.7
Minnesota 74.3

Los Angeles 86.9
Oakland 78.7
Seattle 73
Texas 74.7

Atlanta 83.3
Florida 75.2
New York 92.9
Philadelphia 85.6
Washington 75.9

Chicago 87.8
Cincinnati 80.6
Houston 71.7
Milwaukee 85.6
Pittsburgh 73
St. Louis 75

Arizona 85.2
Colorado 80.3
Los Angeles 86.8
San Diego 77.8
San Francisco 70.5

Remember, these are still preliminary, especially because rosters can change significantly between now and opening day.

Quick take: Yeah, I know what you're thinking. "This guy has Tampa at 86 wins? I'll take my business elsewhere." Go ahead and laugh, I could use less competition at the betting window. For what it's worth, the correlation between these and the Baseball Prospectus numbers is .98.

These totals aren't adjusted for strength of schedule, although that wouldn't change things around too much. At most, perhaps the Orioles would lose 1.5 wins and the Angels would pick up 1.5.

Market Bias

When trying to predict the standings in an upcoming season of any sport, it's important to check and make sure the total projected wins and losses equal out, so the league as a whole is .500. After all, sports are a zero-sum game. It's easy to get into the mindset of looking at each team's positives without considering their shortcomings, like the media is doing with the Mariners right now. If one makes this mistake, he could make projections where the average team is 82-80, which throws everything off.

A little ways back, another blogger suggested that sportsbooks wouldn't make this common error, because there's too much money at stake. At the time of that comment, the MLB season wins over/unders at major bookmaker CRIS did actually average to exactly 81. In a little more than a week since then, the lines have been bet up substantially: Now the total number of wins for all 30 teams is 10.5 wins too high.

That may not sound like a lot, but the bookie moves the line half a win at a time, and only when he receives very one-sided action. So there's been enough betting on the overs to generate 21 different line moves upward.

The bookie may have done his homework, but he's not really responsible for setting the market odds on a sporting event; the public is. And the public has demonstrated time and again that they're willing to be overoptimistic about their team's chances in the upcoming season. CRIS probably could have made more money if they had built the bias into the opening lines, but I have a feeling they'll do okay anyway.