Peter Watts Is An Angry Sentient Tumor: Revenge Fantasies and Essays

by Peter Watts

Here’s the trick: it was all exactly the same data. All Kahan et al did was switch the labels. What they discovered was that your ability to correctly interpret these data comes down to how statistically smart you are, regardless of political leanings—but only when you think you’re dealing with rashes and skin creams. If you think you’re looking at gun control data, suddenly politics matter. If you’re a numerically-smart conservative, you’ll have no trouble parsing the data so long as they show that gun control results in increased crime; but if they show that gun control reduces crime, suddenly your ability to read those numbers drops to the level of a complete innumerate. You’ll only be able to interpret the data correctly if they conform to your pre-existing biases. Smart liberals are just as stupid as smart conservatives, but in the opposite direction. Show a numerically-savvy left-winger that gun control reduces crime, they’ll be all over it; show them the opposite and, once again, their performance drops to the point where they might as well not have any statistical smarts at all. Note that this is not a case of people rejecting a pattern they don’t agree with. This is more subtle, and more pernicious: this is people literally not being able to perceive the pattern to begin with, if it’s too threatening to their beliefs. Ideology compromises your ability to do basic math.