The Trap Game Myth in College Football

Harvard Study Found Teams Perform at Higher Rate in Trap Games

The most rigorous academic analysis came from Harvard Sports Analysis in 2012, which examined every game over a 10-year sample where a winning team played a losing team the week before a major opponent.

The result: winning teams performed at a rate higher than their baseline in so-called trap games, not lower.

Good teams beat bad teams approximately 80% of the time in all contexts, and the week before a marquee game produced essentially the same winning percentage as every other week.

Harvard study findings:

• Examined 10-year sample of trap game scenarios
• Winning teams performed higher than baseline, not lower
• Good teams beat bad teams 80% in all contexts
• Week before marquee game same winning percentage

The difference was statistically indistinguishable from noise, with a t-value of 0.60, well below the 1.96 threshold required to establish significance.

College football is chaos. The Content Lab makes it simple.

FiveThirtyEight Replicated Analysis With 16-Season Sample

FiveThirtyEight replicated this analysis in 2021 with a 16-season college football sample and found that top-10 teams went 143-21 in supposed trap games, a win rate of .872.

In that same sample, top-10 teams went 1,161-149 against all unranked opponents, a win rate of .886.

The difference is 1.4 percentage points, which means a top-10 team playing a supposed trap game loses at basically the same rate as a top-10 team playing any unranked opponent on any given Saturday.

FiveThirtyEight replication:

• 16-season sample, top-10 teams 143-21 in trap games (.872)
• Top-10 teams 1,161-149 vs all unranked (.886)
• Difference is 1.4 percentage points
• Basically same loss rate in trap games vs any unranked opponent

Think you know Saturdays? Prove it in Gridzy. Build your picks. Beat your buddies.

Football Outsiders 25-Year Study Arrived at Same Conclusion

Football Outsiders conducted a 25-year comprehensive study that arrived at the same destination from a different direction.

Not only did teams not lose trap games at a higher rate than their baseline, they actually won slightly more frequently.

The directional evidence consistently runs against the trap game hypothesis, meaning the bias in the data, to the extent any exists, runs opposite to what the narrative predicts.

Good teams win football games against bad teams reliably, whether or not the calendar includes a marquee opponent the following week.

Before rivalry week explodes, check the Content Lab. From Heisman angles to upset alerts, it's all in the Lab.

Myth Persists Due to Confirmation Bias

The reason the myth persists despite the data is a combination of confirmation bias and narrative availability.

When Michigan State loses to an unranked team the week before Ohio State, the "trap game" explanation is immediately available, emotionally satisfying, and prevents the need to engage with the more uncomfortable explanation.

That Michigan State may simply not be as good as their ranking suggested, or that the opponent played exceptionally, or that variance in football produces results that talent differentials alone cannot predict.

Why trap game myth persists:

• Confirmation bias and narrative availability
• Emotionally satisfying explanation
• Prevents uncomfortable explanations (team not as good as ranking)
• Converts random outcome into story about motivation and focus

The trap game label converts a random outcome into a story about motivation, focus, and preparation, human qualities that feel more meaningful than probabilistic variance.

If you're calling an upset, back it up in Gridzy.

Trap Game Has Sliver of Support in ATS Context

The trap game concept has a sliver of empirical support specifically in the ATS context rather than the straight-up result context.

Doc's Sports conducted a study specifically on spread outcomes in trap game scenarios (defined as games where a team plays a significant underdog one week before a major opponent) and found that favorites covered just 42.5% of those games against the spread.

This is a genuinely different finding from the win-loss data, and it has a plausible structural explanation.

ATS context findings:

• Doc's Sports study: favorites covered just 42.5% in trap games
• Genuinely different from win-loss data
• Sportsbooks price game based on public's belief in trap vulnerability
• Creates inflated spreads that fail to cover

Sportsbooks price the game based on the public's belief in trap game vulnerability, which creates inflated spreads for the favored team that then fail to cover because the actual talent differential, while significant enough to win the game, is not significant enough to win by the inflated margin.

Between games? Piggy Arcade's top picks are ready.

Practical Conclusion for Bettors

The trap game does not meaningfully affect which team wins. But it does meaningfully affect whether favorites cover enormous spreads.

Because the public perception of trap game risk drives spreads upward to levels that inflated favorites cannot systematically reach.

The myth is false as a win-loss predictor and partially useful as an ATS fading tool.

Understanding the difference between those two applications is what separates sophisticated analysis from ESPN segment content.

No late kickoff? No problem. Piggy Arcade's live. Switch from tailgate mode to top slots in seconds.

The Bottom Line on Trap Game Myth

The trap game is college football's most beloved analytical fiction (Harvard study found teams perform at higher rate in trap games, not lower, t-value 0.60 well below significance threshold). FiveThirtyEight replication: top-10 teams 143-21 in trap games (.872), vs 1,161-149 against all unranked (.886), difference 1.4 percentage points. Football Outsiders 25-year study: teams won slightly more frequently in trap games. Myth persists due to confirmation bias (emotionally satisfying, converts random outcome into story about motivation). Trap game has sliver of support in ATS context (Doc's Sports: favorites covered just 42.5%, public belief creates inflated spreads). Practical conclusion: doesn't affect who wins, does affect whether favorites cover enormous spreads.

College football is chaos. The Content Lab makes it simple.