Methodology Deep Dive

How CFBTrack calculates Weather Impact.

Weather Impact turns game conditions into readable buckets so fans can see how temperature, wind, precipitation, and indoor context affect scoring and style.

Open related toolBack to methodology hub

Plain-English explanation

The weather pages group games by conditions such as temperature, wind, rain, snow, dome status, and broader adverse-weather scenarios.

The model compares those buckets against neutral or baseline samples, then reports scoring, yardage, pass-run style, and upset behavior where the data supports it.

Inputs used

  • Game-level weather fields tied to venue and kickoff context.
  • Temperature, wind, precipitation, indoor and outdoor status, season, week, opponent, and team scoring.
  • Team-specific adverse-weather games compared with broader FBS baseline buckets.
  • Betting-line coverage for upset rates where that source data exists.

What the model rewards

  • Samples that are large enough to compare without overstating one unusual game.
  • Clear differences between neutral-weather scoring and adverse-weather scoring.
  • Style shifts where pass yards, rush yards, turnovers, or total points move by condition bucket.

What the model does not claim

  • It does not claim weather alone caused a result.
  • It does not treat all rain, snow, wind, or cold games as identical.
  • It does not claim upset rates are complete before betting-line coverage is strong enough.

Example using Florida State

For Florida State, the team weather page compares the program's games against FBS temperature baselines and then calls out adverse-weather games that moved farthest from neutral scoring expectations.

  • A cold or rainy game can appear as an outlier only when the game has enough condition data to classify the scenario.
  • A team can show a passing-yard drop in wind without the page claiming the wind was the only cause.

Known limitations

  • Older games can have thinner weather or betting-line coverage than recent seasons.
  • Venue matching, neutral sites, delayed kickoffs, and indoor context can affect classification.
  • Small scenario samples should be read as directional context, not a universal football rule.