I’ve mentioned a few times that Nate Silver, Baseball Prospectus writer and inventor of the PECOTA prediction system, is putting his methodologies to the test with the election over at fivethirtyeight.com.
Here are his final predictions:
Obama 349
McCain 189
In popular vote terms, that’s a 52.3% to 46.2% win for Obama.
In the Senate, Silver is predicting the Dems end up with 57 seats, the GOP 41. (I don’t think that includes Lieberman of Conn. and Sanders of Vt., independents who caucus with the Dems. Though with Lieberman stumping for McCain, don’t be surprised if he officially goes GOP soon.)
I don’t have time today, but I’d love to find out how 538 and PECOTA are related. Does a strong showing today by 538 give PECOTA more legitimacy? At least superficially, yes. If Silver nails the political predictions, get ready for Baseball Prospectus to gain more respectability — deserved or not — in the eyes of stat-skeptics. Some, not all. But it stands to reason that if the guy can get to reasonable outcomes by crunching complicated historical poll and election data, he should be able to do so with baseball statistics.
I heard NY Times economics writer David Leonhardt on the radio last night. He not only gave a nod to 538, he said something interesting: people don’t like to hear about chances expressed in percentages. Therefore, pundits say things like, “It seems nearly certain Obama will win” or “It doesn’t look good for McCain.” Technically, there’s nothing inaccurate about that, but it leaves less emotional room to acknowledge why the opposite might come true. “Obama has a 95% chance to win” is a lot more bloodless, at least for a populace that thinks math is for nerds and being smart is elitist.
Any thoughts on statistics, political or basebattical? You’re also welcome to continue with the voting preferences or anecdotes as you come back from the polling places. Thanks for keeping it civilized so far.
