NFL turnovers in bad weather

The popular version is wrong. Across 3,474 outdoor games from 2006 to 2025, turnovers in wet or cold games are almost indistinguishable from dry, mild ones. Steady rain adds +0.30 turnovers per game — real in direction, too small to plan around. Interceptions do not move at all. The one clear effect is loose balls: when rain and cold arrive together, total fumbles rise by +0.73 per game. Meanwhile scoring falls sharply, which is where the weather actually shows up.

Before weather can be blamed for anything, the era has to be removed. Outdoor games averaged 3.60 turnovers in 2006 and 2.22 in 2025 — a fall of about 38% with no help from the weather. Any study that pools two decades without this correction will find whatever it wants.

Every number in the tables below is each game measured against its own season's average, so this decline cannot leak into the weather comparison.

The smaller figure in each cell is the season-adjusted gap against the control group, with its t statistic. Anything under roughly 2 is noise. Bands under 40 games are marked thin and greyed out — the heaviest rain band in particular rests on a couple of dozen games and should be read as direction only.

• Interceptions do not move. Not in trace rain, not in steady rain, not in freezing cold. This is the cleanest negative result in the study and the one most at odds with the broadcast script.
• Fumbles rise, mostly when rain meets cold. Games with measurable rain below 40°F produce +0.73 fumbles per game against the control, comfortably the strongest effect here.
• Fumbles lost lag behind fumbles. More balls hit the ground but recovery stays close to a coin flip, so the turnover column absorbs only a fraction of the extra chaos.
• Scoring is the real weather variable. Steady rain costs -4.7 points against the control — an effect several times larger, relative to its own noise, than anything in the turnover columns.

• Sample: 3,474 outdoor regular-season NFL games, 2006–2025, each with a complete public boxscore and resolved venue coordinates. Domes and closed roofs are excluded entirely rather than used as a control.
• Weather: hourly reanalysis at the stadium's own coordinates. Rainfall is summed across the three hours from kickoff, not read off a single hourly snapshot; temperature and wind are taken at kickoff.
• Turnovers: interceptions thrown and fumbles lost come from team boxscore totals. Total fumbles are summed from individual player lines, because team totals only publish fumbles lost.
• Control group: under 0.1 mm of rain and a kickoff temperature between 50 and 75°F. Games that are both wet and cold are held in their own band so they cannot contaminate the rain-only or cold-only rows.
• Not controlled: wind beyond its reporting, team quality, pass-run balance, surface type, and whether either side had already clinched. Snow is not separated from rain — reanalysis reports both as precipitation.
• Closing lines are not attached to these games, so this study reports rates and scores, not Over/Under or prop hit rates. We will not publish hit rates without the lines.

Data generated 2026-09-07. Numbers refresh as the archive tops up; corrections are dated and kept visible.