A lot of professional blabbermouths are on the news channels saying the upcoming election is a toss-up and that all predictions about what might happen are equally without merit. This line of argument has started to piss off professional statistician Nate Silver of the polling analysis site fivethirtyeight, which currently has Obama a strong favorite to win. He suggested one critic, Joe Scarborough, should be willing to put his money where his mouth is. If he was so sure fivethirtyeight was calling it wrong, he ought to be willing to bet a thousand dollars on Romney.
This led to a certain amount of blowback from Silver’s employer, The New York Times, which I won’t go into here, especially since the subject has been well covered elsewhere. For myself, however, I rather agree with a Mr. Alex Tabarrok when he said that “a bet is a tax on bullshit.” Which, I guess, is how I wound up in one of these bets myself. I am either about to be taxed on my bullshit, or someone else is.
The other day, over on Twitter, novelist Walter Kirn advanced a Scarborough-esque argument that the polls are wrong and professional pollsters are headed for a kind of Waterloo. I went little-Nate on him and said let’s bet. So we did. And here we are.
The wager: we bet $50, to go to the charity of the winner’s choice, on the outcome of the election. His money is riding on Romney. Mine… well I’m betting on Nate Silver (more about that in a second), which means a bet on an Obama victory.
But wait, there’s more: the loser also must post a photograph to Twitter of themselves, dressed only in their underwear, and eating metaphorical crow (a leg of chicken). Believe me, the humiliation will be way more satisfying than the money. Bet to be decided when the election is.
The rationale: So why am I ready to bet some money, my pants, and my dignity on Obama? Well, as I said above… I’m not really betting on Obama so much as I am on the science of polling and statistical analysis, which just so happens to be a bet on Obama.
But my confidence in Nate Silver’s math does, oddly, provide a backdoor explanation for why I’ve voted a straight Democratic ticket for the last six years, and will again in Tuesday’s vote.
I believe in the importance of a reasonable, agile opposition party to provide checks and balances on the other. (Really, I think the country would be a whole lot better off if there were several parties to choose from, including a Green Party, a No-Tax Party, a Science Party, and a Keg Party.) So in theory, I have no problem splitting my vote between small government Republicans and community-first Democrats.
But the current incarnation of the G.O.P. is defined by dizzying and indefensible anti-science views which make it impossible for me to support them. Pick the science of your choice: evolutionary science, environmental science, the science of biology (especially in regard to female human biology and homosexuality), economics… and now, statistical analysis. If the science backs findings that are contrary to G.O.P. orthodoxy, the rational person would say the orthodoxy might need a tweak; the G.O.P., however, furiously insists that the science itself is wrong, and anyway, science itself is only another “belief system” no better than any other. Also, Stalin employed scientists. Or something.
Senator Dan Moynihan famously said that “everyone is entitled to his own opinion, but not his own facts.” The modern day Republican party thinks otherwise (Romney’s own campaign manager amusingly claimed that their campaign would not be run by “fact-checkers”). I humbly think truth is truth; it cannot be denied or debated out of existence. Or, to quote someone else – David Mitchell, in Cloud Atlas – “Truth is singular. It’s ‘versions’ are mistruths.” The whole idea of the scientific method is that ideas can be tested and retested and found to be true or not.
Nate Silver is not in the bag for Obama. He is in the bag for statistical analysis… the idea that if you know how to look at them, numbers tell us something about the world, and can somewhat predict human behavior in certain fixed situations. Nate Silver is in the bag for science. So am I. Fivethirtyeight has modeled the polling data, and compared it to polling data and outcomes from the last century of elections, and they think the current president is a better than 4-in-5 favorite to be the next president. Basically, the best bookie in the country says there’s a less than 20% chance I’ll be scarfing Kentucky Fried in my skivvies next Wednesday.
I’ll take them odds.