Member-only story
World Cups and the Curse of Low Sample Sizes
“14 games is a fairly low sample size. If you want to determine who the best player is, you need more matches. If you want to find the best player, you need to change the format to increase the chance of finding the best player.”
Magnus Carlsen, World Chess Champion
The German team was kicked out of the football world cup in Qatar in the preliminary round for the second time in a row after the same thing happened in 2018 in Russia.
Putting all the justified controversy around the tournament in Qatar and the ridiculous amount of corruption inside FIFA aside (“A cartel-like group of scumbags and assorted criminals who occasionally put on soccer matches.”), what struck me in all the outrage of German media about the German team dropping out (“a disgrace, a shame, a dishonor”) is that no one ever mentions the ridiculously low sample size that led to Germany dropping out of the tournament.
People love looking for patterns, structural deficits, and profound explanations, and in no place more than football, an infinite canvas for projections for millions and billions of fans around the world.
But sometimes, when it’s only three games, or, actually only one game that decides the fate of a tournament, it could also just be noise. Even though it is hard for our human minds constantly on the lookout…
