Saturday, June 21, 2008

Nate Silver has a political blog

Last night I ran across a link to a post at, a political blog by Nate Silver. The same Nate Silver who wrote for Baseball Prospectus and designed the industry-leading PECOTA projection system? The same! I'm quite excited. Apparently he blogged anonymously until May 30 of this year, when he unmasked himself as a necessary precursor to getting wider distribution by some big media enterprise or another.

For those who don't know, BP's core product is a quantitatively-oriented baseball analysis, and when you click on one of their player pages you see a lot of charts like the one above.

That's from Milton Bradley's page, and shows the distribution of historical seasons for the players he most closely ressembles, with the y-axis representing playing time (plate appearances) and the x-axis representing quality (in the form of a catch-all offensive statistic they call Equivalent Average).

His political analysis gives the same treatment to polling data. As he put it in the aforementioned post: "What we do over there and what I'm doing over here are really quite similar. Both baseball and politics are data-driven industries." And it's very true, at least if you're just talking about horserace politics. Policy seems to have a way of being less data-driven, but in baseball as in horserace, the key question is Win Expectancy. All subcategories (home runs, steals, white women, hispanic seniors) are all fodder for the overarching question of how likely a given team or candidate is to win.

Having been a BP reader for many years, I have a lot of faith in his quantitative game, and he apparently gives Obama a 76.3% chance of victory. That's slightly higher than I would have guessed, but I don't know how much it fluctuates with new polling data. If it's in there, the new Newsweek poll that has him up by a whopping 15 may be juicing that a little bit.

He's also an Obama supporter, which means to whatever extent there's an editorial slant (and my guess is it will be minimal) it will probably make the blog more enjoyable. So he's in the ol' RSS reader already.

This morning he did a post on yesterday's FISA vote, and found that the 32 House Democrats in districts designated as competitive voted overwhelmingly for capitulation while the rest of the caucus voted 120-82 against. That in itself is not at all surprising. Districts that were competitive in 2006 tend to be pretty darn reddish, and the blue seats tend to be held by Blue Dogs like Shuler, Carney and Ellsworth. The kind of Democrats that electoral pragmatists would have us live with and not challenge from the left.

The more interesting point was relating the votes of this subgroup to Obama's decision to support the bill: "Was he really going to run afoul of the Blue Dogs when they are probably his swing voters in passing some version of national health care legislation?" He also points out that Pelosi herself was very much on board with capitulation, and she's no Blue Dog. So there's the intramural explanation for his position that focuses not on Obama's reluctance to alienate swing voters in November but Congressional swing voters in early 2009, and it gets more appealing as a justification the more you discount November swing voters' awareness of the FISA issue at all.

My problem with it is that it could be used in any circumstance to justify agreement with the right wing of the Democratic party. On any unambiguously liberal legislative effort, the Blue Dogs are going to be the whiny ones who worry about being called Communists. That goes for health care, taxes, Social Security, carbon emissions... you name it. How big a Congressional majority will he need before he'll be comfortable telling the Blue Dog Caucus to get bent? Because that's what it comes down to, for me.

No comments: