Today I want to talk about a fascinating concept I just learned about: regression to the mean. This occurs if you get an extraordinary result the first time around, and a the second/future result which is a lot closer to the average or vice versa. Let's say I'm playing a match of chess against a friend. We both have no experience in chess. If my friend were to beat me in chess, knowing what we know about us, we could probably agree that my friend won become of luck. Still, who do you think would win in a second match up, me or my friend? Before I learned about this fallacy, I'd probably say my friend, maybe they just have some innate chess talent I never knew about. If we take regression to the mean into account we'd know that, on average, it's essentially pure luck on who wins and loses, it should be about a 50-50 chance either way. My friend probably has no advantage, even if it may seem like they do. I was just unlucky.
It's easy to believe that my friend would win because they seemed like a better chess player. It's one I'd believe too, but an extraordinary result the first time probably isn't going to be repeated. In fact, it'll probably get worse. To be fair, it is possible that my friend is a chess genius, but it's much more likely that I was just unlucky that first time. I think this phenomenon gives us something to think about and shows a little bit of the beauty of statistics. Until next time!