Monday, November 5, 2012

One Offs: MA Question Two and Nate Silver F/U

Normally at the Billy Rubin Blog we like giving a good 12 paragraphs to explore the intricacies of an idea, but no time for that today. Before the election, though we want to follow-up on two themes about which we have been writing this year: Physician-Assisted Suicide and, more recently, the phenomenon of Nate Silver.

In Massachusetts this year, Question Two on the ballot proposes a legalization of PAS. It is often described as a "liberal" issue. I don't think it is, and I'm voting No--as emphatically as one can, given that voting consists of coloring an oval black, but there you have it.

The reasons to oppose PAS, especially from a lefty point of view, are twofold:

a. Hospice is wildly underutilized; and
b. PAS should only be practiced in a place in which all citizens have equal access to good health care, and even with the Affordable Care Act we are still a long way from that.

PAS has the feel of a hot-button issue--strong opinions on both sides with heated rhetoric and fierce stands based on core moral beliefs. But unlike, say, abortion, PAS is mostly a smoke-and-mirrors debate where there's very  little "there" there. Even in Oregon, the PAS pioneer, there have been only about 600 "prescribed" suicides in 18 years since legalization. But PAS has substituted for a more substantive discussion about how we will treat end-of-life issues in the US. At a time when we need to examine how we spend money on our health because we cannot sustain our current model (which isn't a good model anyway), that's lousy politics for the left, right, and center.

Onto Nate Silver.

We wrote that being "against" Nate Silver generally showed a contempt for science and mathematics, and we still stand by that claim. However it's important to note that Silver isn't a stand-in for Truth, and that if his success in political prediction matches that of his baseball analysis acumen, the Dems may be in for a very rocky night tomorrow and the Billy Rubin Blog staff will have one phenomenal hangover Wednesday morning. That is, Silver's statistical baby for baseball, named "PECOTA", hasn't performed significantly better than other models predicting player performance and has done so using a cranky, Rube Goldberg-like statistical contraption. Colby Cosh of Maclean's (a Canadian publication) heaps reams of skepticism on the Silver phenomenon here.

Cosh's writing is really good and provides some fresh insight from a guy who appears to be steeped in the numbers, although that said I think he's mostly missing the point. As can be found here and here (and talked about over here), there are number-crunchers who can provide plausible scenarios of why we might wake up trying to familiarize ourselves with the phrase "President Mitt Romney". But this is a question of "what is the underlying reality of the campaign, and how do we find data to help make an accurate prediction?" That question of late has frequently become morphed with "which guy do you want to win?", and Silver, who is almost completely a numbers geek with very few overt partisan leanings, has gotten pegged by conservatives as being in the tank for the Obama cause: right wing paranoia if ever there was any. This is why people have been giving pushback on Silver attacks.

Indeed, there were plenty of Dem-leaning commenters on the Cosh/Maclean's piece that welcomed the critique, because they endorsed the idea of data-driven analysis and not cult-of-personality devotion to Silver. Said one commenter: "I think Silver is a good thing for journalism, but it is misleading to call him a statistician or a scientist. He's something else entirely: a data journalist. He's a very bright guy to have spotted the gap in the market which opened up thanks to the easy availability of data, which mainstream journalists have no training or inclination to use." That sounds right to me, and is the best explanation of King Nate's popularity. Indeed, Silver's book Signal and the Noise is mostly a journalist's account, and it's a really good read.

We will wake up Wednesday morning and will know, given many Senate and House races in addition to the 50 state Presidential races, whether Rasmussen's polls, which have always favored Republicans by about 2 points this year, are more accurate than Quinnipiac's or anyone else's. If so they're probably doing something right and have a better model. Sure, some on the left have foolishly conflated support for Silver with support for liberal political issues, but crying foul against libs who Don't Get It is an exercise in false equivalence. The vast majority of the right wing screeching about Silver has not had to do with an opposition to Silver's possibly errant calculations, but rather with a resistance to any data that does not support one's ingrained assumptions.

More than anything else, this is why we find the Republican Party as currently constituted a menace to society, and until this problem is fixed, we have a very deep political problem in this country.

No comments:

Post a Comment