Does Weather Actually Change College Football?

Every broadcast has a version of the same line: "this rain is really going to hurt the passing game." We had a database of nearly 18,000 games and their weather, so instead of taking the announcer's word for it, we checked. Some of it holds up. Some of it doesn't.

Rain and snow make it a lot harder to throw

This is the clearest thing we found, and it's not close. The wetter or snowier it gets, the fewer passing yards teams put up — even after accounting for the fact that good offenses play in bad weather sometimes too. Here's what an average team's passing output looks like as conditions worsen:

Wind adds to this on its own, separate from rain or snow — a gusty 20 mph day costs a passing offense somewhere around 17 yards by itself, even under a clear sky.

Wind makes field goals a little tougher, too

Kickers lose some accuracy as the wind picks up — about 2 percentage points over a windy 20 mph day. It's real, but it's a much smaller effect than the "no way he makes this" reaction a wind-blown miss usually gets on TV. Rain, on its own, didn't move the needle on kicking at all once we accounted for wind.

But it doesn't cause more fumbles or turnovers

Here's where the conventional wisdom breaks down. "Sloppy conditions, sloppy game" is a broadcast staple, but once we control for the fact that bad teams and good teams don't play in bad weather equally often, there's no real turnover bump from rain or wind at all. Whatever's driving that highlight-reel image of guys slipping and fumbling in the mud, it isn't showing up as an actual rate effect across thousands of games.

So why doesn't the final score change?

This was the most surprising result, so it's worth slowing down on. If passing offense really drops in bad weather, shouldn't teams score less — and shouldn't that occasionally flip the winner of a close game? It turns out there's a subtle reason it doesn't work that way: weather doesn't pick a side. It suppresses both teams' offense by roughly the same amount, in the same game. Here's a simplified example of what that looks like:

That's exactly the pattern in the real data. A game's final score total (what sportsbooks call the over/under) really does respond to weather. But the margin — and with it, who wins — barely budges, because both teams are getting rained on. We checked this two different ways (a statistical model and a machine-learning model trained to predict winners), and they agreed: weather explains essentially nothing extra about who wins, once you already know how good the two teams are.

Snow doesn't just slow the game down — it changes the plan

Snow is a special case. It's not just "more weather" than rain — teams visibly change strategy for it, leaning on the run instead of trying to force throws into the wind and cold:

The time we almost got it wrong

We want to be upfront about something, because it's a good example of why you should be suspicious of results that look too clean. Our first attempt at testing "does weather affect the final margin" came back with an answer of exactly zero — not "close to zero," but zero down to the fourteenth decimal place. That's not what a real "no effect" looks like; real findings are noisy. It turned out to be a math trick playing games with us: we'd accidentally set the analysis up in a way that was guaranteed to produce zero regardless of the truth, sort of like asking "what's the average of a number and its own negative" and being impressed that it's always zero. Once we fixed the setup, the real (still not-significant, but now trustworthy) answer came out. We're mentioning it because it's exactly the kind of mistake that's easy to make and easy to miss.

What's next

This isn't a one-time report — the dashboard updates through the season, and we're actively using these findings to build a model that predicts a game's total score before it's played. Read about how that works, and see it in action against this week's real games, in the next post.