Here's their article: https://fivethirtyeight.com/features/upsets-happen-in-college-football-trap-games-not-so-much/ They compare "trap game" winning percentage vs "all games against unranked opponents," and conclude that because the win percentages are similar, there is no evidence of a trap game. Anyone see the obvious broken logic here? They did not look at the independent variable... independently. There is no control group. This is a site I frequent often for sports and other things, and RELIABLY find fallacies in their work such that it seems likely that they attempted to work backwards from a conclusion to support it. If you separate trap games from non-trap games, you get 143-21 (.872) for trap games, and 1018-128 (.888). Does this significantly change the picture? No, not really. But their methodology didn't make sense. What about their interpretation? Is the difference between 87.2 % and 88.8 % in a 16-season sample meaningful? Would controlling for other meaningful factors such as P-5 opponents, annual rivals, etc change the picture? After all, I have never heard an FCS opponent be called a trap game seriously, let alone a non P-5 unranked team. And what about Top 5 teams only? Or Top 15? Top 25? Perhaps the "trap" effect becomes more notable looking at different thresholds for being the trappee? I don't have the time right this moment to dive into that, but I think a more complete definition of a "trap game" besides "unranked opponent" would be interesting to look at.
Well they looked at Top 10 teams who played an unranked opponent right after a ranked opponent. I'm saying that isn't really a good definition of a trap game. The "trap" isn't just anyone, it would need to be a credible unranked team.
That there is even the slight edge towards unranked opponents after a ranked game for top 10 teams leads me to believe there IS something to trap games, if one narrowed what constituted a trap game to something more specific.
That's how I have always defined it in my head. I'd call what they call a trap game a "letdown game".
Ya, which one would expect to not be super prevalent compared to just playing an unranked team in general, hence why it is so damn disappointing.