Friday, February 27, 2009

Predictive Luck vs Close Game Luck

John Gasaway’s Friday column looked at the luckiest and unluckiest teams in the nation. His metric is expected wins (based on performance) as compared to actual wins.

You might expect that the lucky teams on that list will have a lot of close wins and the unlucky teams will have a lot of close losses, but that’s not always the case. To be lucky, all you really need is for your margin of victory in wins to be smaller than your margin of defeat in losses. (And technically when luck is adjusted for strength of schedule, you might have no close wins, but if you don’t blow out the bad teams by enough you are lucky.)

But when fans think of luck, they tend to think of the close games, the heart-breaking 1 point losses where you don’t get the tip-in, ect. And that metric presents a different set of teams this year.

Record in games decided by 3 points or less (major conference teams)

Dayton 6-1
Florida State 6-1
LSU 4-0
UAB 5-1
Arizona State 4-1
Georgia 4-1
Michigan 3-0
Villanova 4-1
Xavier 3-0

On the flip side:

Colorado 1-5
Saint Joseph's 1-5
Stanford 0-3
Mississippi State 0-3

Having cheered for Georgetown last year (lucky) and this year (unlucky), I agree with John Gasaway’s sentiment that you’d rather be lucky than unlucky. But when you are lucky, people say you are over-rated. And when you are lucky and you are winning games by 3 points or less, everybody wants revenge the next year. Georgetown travels to Villanova this weekend, but I doubt Villanova will be overlooking the Hoyas despite Georgetown's 5-10 record.