The Texas Sharpshooter

Oh yes, we’re doing logic again. A little bit, anyway. It’s been too long since I wrote about philosophy, and I’m shaking the dust off my bones. My brain was growing moss. I want to talk about more fallacies, because fallacies are incredibly fun to talk about. Besides, we already talked about formal logic a year ago. This won’t be a weekly thing like it used to be, because I’m doing too much other writing to make that happen. Also, did you know that having a real job crunches your time way more than grad school? True fact. Anyway, today is about the Texas Sharpshooter fallacy.

When you want to find something out, the thing to do is form a hypothesis and then test it. This is as true in your kitchen as it is in particle physics. If I want to know if I have time to jump into the shower while my eggs are cooking in the morning, I might think about it and then test it by trying it out. Maybe it’ll work. Maybe I’ll start a fire in my kitchen and force my roommate to run out with the fire extinguisher while I sing “My Heart Will Go On” in blissful ignorance. Either way, with enough testing I’ll find out if my hypothesis “I have enough time to shower while breakfast is cooking” stands up.

When I’ve gathered this data over a hundred days, little clusters of data will emerge. Sometimes it might be three weeks between fire extinguisher incidents (though he refers to them as Celine Dion incidents, only because his bedroom is right next to the bathroom). Sometimes there’s a number of Celine Dion incidents in a row. To really find out if it works, I have to look at the whole set of data and compare it to my hypothesis. If, out of the total number of days that I tested it, there were more fires than not, I probably don’t have enough time.

The Texas Sharpshooter masquerades as the scientific method, but falls a bit short. The classic example is that of a gunslinger (presumably a Texan) firing a few rounds into a barn and drawing a target around a few that have clustered together, then proclaiming he’s a sharp shooter. Pro tip son, you ain’t. The Sharpshooter fallacy involves designing the hypothesis to fit around a cluster of data, and ignoring the rest. I’d be guilty of it if I tried to explain my propensity for synchronized cooking and cleanliness to my roommate by saying

“Look at this chart. There are three weeks where I didn’t set the kitchen on fire. This works, and it saves me a ton of time in the morning.”

Not my roommate. Actual fake sharpshooter.

My roommate, who is rather sharp, would rightly point out all the times when he had to go and put the fire out, insisting that I should take into account all of the information available. If we look at the whole chart, we see that my love for Celine Dion keeps me singing in the shower far too often, not to mention that fry pans cost money.

I find the Sharpshooter fallacy arises in conversation as a sort of defensive posture. “If you look only at the set of data that I want you to look at, you’ll see that I’m right.” This can range from finding prophecies in the works of Nostradamus (If we ignore all the irrelevant things it’s amazing what we can find out he knew about our slightly distant past) to people who are jerks arguing that they’re not, as long as you look at all these times when they weren’t jerks. At the end of the day, a good hypothesis needs to come before the data is gathered, and needs to be tested against the largest amount of information available, not the amount that’s most convenient.