From Sabermetrics to Nate Silver, the past decade has become the decade of probabilities as headlines. Serious baseball fans don’t just study the box scores in the paper anymore – they turn to Fangraphs and other websites to see how their analysts have projected probabilities of everything from a team making the post-season to winning the World Series. Each baseball site has its own proprietary analytical formula, which gets tweaked after each season to try to adjust for whatever the difference turned out to be between their projections and the final actual results.
And of course, Nate Silver, who got his start doing probabilities and projections on sports stats, applied the concepts to political polls and created a model that provided probabilities projections for the ’08 and ’12 national elections. Silver’s model’s projections the night before both elections was stunningly close to the actual results.
I follow national politics more closely than does me good (I get super anxious and neurotic about things in a way that isn’t really helping anyone or anything). I am someone who tends to form strong gut feelings about political trends, but I don’t think I have an accurate sense of how often my gut turns out to be right. Regardless of whether I have a very intuitive gut or whether I’m just proven right some of the time and tend to remember those times but not the others, I am, in the end, stuck with my gut feelings.
But something I’ve never done before is express my gut feelings in terms of probabilities. Previously, I’ve just said, “I think so-and-so’s going to win,” and then say why. But I’ve never said, “I think right now so-and-so has a 75% likelihood of winning,” so I thought I’d take a shot at doing that now. I’ll quantify my gut feeling probabilities as UGPs – unscientific gut-feeling probability.
Here goes nothing! Continue reading “Gut Feelings as Probabilities”