This post is meant to cover a helpful mental model that I can use for improving my decision-making. Normally, when you're presented with possible solutions to a problem, the default pattern of thought is to find more samples that fit along with the ideas. This is exactly the opposite of what should be done.

If you have a hypothesis, the best way to test the hypothesis is to find counter examples. That way, you get to cover the cases in which your hypothesis would fail - and become invalid.

I remember reading about a really useful example from somewhere else that had to do with a number pattern. It went roughly like the following:

Host: I'm going to present a game. The goal of the game is to find the pattern for a sequence of numbers. I'll give you the first three numbers, and then you can guess as many numbers as you want afterward - until you're ready to guess the pattern.

Guest: Sure, let's hear the first numbers.

: 2, 4, 6Host

Guest: All right. Then what about 0, 2, 4?

Host: That also fits the pattern.

So far it's all good. You have a hunch that the pattern is a sequence of numbers that change by increments of two. Just to be sure, you also try to use a negative number sequence.

Guest: All right. Then what about the sequence -6, -4, -2?

Host: That also fits the pattern.

That's perfect. Now you think you're ready to make your guess.

Guest: OK, I know the pattern. The pattern is that the sequence will change by increments of 2.

Host: Wrong. The pattern was that the sequence is increasing.

What went wrong here? Each and every single one the guesses was perfect - it got you closer to being assured of your hypothesis. The only way to find the real pattern would have been to guess patterns that disprove the hypothesis. Because it feels much better to be correct, it's easier to find evidence that confirms our initial beliefs.