WSJ: Bialik: Can the Red Sox Sweep Again? Furniture Buyers Want to Know

What do you get when Carl Bialik, Jim Albert, Phil Birnbaum and Jim Lackritz put their heads together on the furniture puzzle? Uhh...a leg up on the competition?

So I turned to several sports statisticians, whose answers for the probability of a Red Sox sweep are higher than you might think: Somewhere between 2% and 5%. They all pointed to sports books that had the Sox as among the favorites to win the American League pennant, at about one in four or five. (See, for instance, TradeSports or BetFair.) That reflects their dominant championship last year and their high level of talent. Then the probability that the Red Sox sweep the Series is equal to about 0.2 — the chance they make the World Series — multiplied by their probability of sweeping the Series. If each game is a toss-up, that’s one in 16: 1/2 multiplied four times, for the four wins needed for a sweep. That translates to one out of 80 that Boston will sweep the Series. But sports-book odds suggest that the Red Sox are likely to be better than their World Series opponent, because the odds they’ll win it all are greater than half their pennant odds. That can nudge the probability of a Sox sweep up to one in 50.

In reality, individual games aren’t identical; even at the simplest level, the home team should get an advantage. (I went into more detail about winning probabilities in baseball in this 2005 column.) That doesn’t significantly change the likelihood of the sweep, at least not within the fuzziness of making such a calculation in late April. One statistician also suggested that sweeps are more likely than statistics would suggest because a team down 2-0 and 3-0 is likely to feel demoralized and have a lower chance of winning subsequent games.