Nudges
Small things the replays turned up. Most were first noticed on the 2023–25 seasons, then checked on all 27 replayed seasons, 1999–2025 (6,918 games); each entry says which evidence it rests on, and several reversed on the long run. Elo's own constants (home field, reversion, the quarterback multiplier) can only be tested on the seasons we replay ourselves, so those stay three-season findings. Some are built into the model, some are being watched, some we chose not to act on, and several are myths we can now put down. Each one carries a live tally for 2026, so the season tests it out of sample. Sample sizes matter: a 60-game finding has a standard error of about 6 points. The methods behind these are on the How it works page.
Wind hurts favorites; cold helps them
What we found. Over 27 seasons of outdoor games with weather on record (4,500): in wind of 15+ mph favorites won 62% against a 65% forecast (626 games); at 50F or colder they won 70% against 66% (1,517); above 75F, 61% against 63% (742). Scoring drops about 1.5 points below the total in wind.
What we did. The System shrinks toward 50 by a fifth in 15+ mph wind, adds 2 to the favorite in the cold, and shrinks by a tenth in heat, outdoors only. About +9 points a season, positive in every era. Forecasts come from Open-Meteo at kickoff.
What would confirm it. Favorites in windy games winning a few points less than the market says, again.
Does a track-record-weighted crowd beat the market?
What we found. Nobody has tested this in a public game. FiveThirtyEight showed the raw crowd average, which trailed Elo. Here each player's pick is weighted by their season points (1 + points/50), and the result is scored like any forecaster.
What we did. "The sharp crowd" is a row on every leaderboard and the Predictions scoreboard next to the raw crowd, the market and The System.
What would confirm it. The sharp crowd finishing above the raw crowd over a season would show skill is identifiable; above the market would be news.
When Elo and the market agree tightly, the favorite wins more than either says
What we found. In our 2023-25 replays the favorite won 74% against a 66% forecast in these spots. Over 27 seasons (1999-2025, 2,607 games) the surplus is real but small: 69.8% against 67.7%, consistent across eras. An 8-point bump fitted to 2023-25 cost 40 points a season over the long run.
What we did. The blend pushes 2 points toward the favorite in these games (favorite at least 55%). Cards show a "Tight agreement" marker.
What would confirm it. Favorites winning a point or two more often than the average forecast in these spots; anything much larger would be a surprise.
When Elo and the market disagree by 10+ points, Elo is the one that is wrong
What we found. Over 27 seasons, 1,102 games with a 10+ point gap: Elo's side won 50% while Elo claimed 65% and the market 48%. Our 2023-25 replays matched (49% vs 64%).
What we did. The blend averages the two, so it never follows Elo far from the market. The bold call of the week is shown as the contrarian view it is.
What would confirm it. Elo's side winning about as often as the market says, not as often as Elo says.
Week 18 is where Elo fails worst
What we found. On 2023-25, week 18 was Elo's worst slice: Brier 0.246 against the market's 0.203, with three of the 15 worst misses being clinched teams resting starters. Over 1999-2025 the last regular week is not special at all: home teams won 60.3% against Elo's 59.5% and the market's 57.6% (411 games), and Elo's Brier there was better than its season average. So the week is fine; what is real is the narrower thing, teams with nothing to play for, which the 17-game schedule since 2021 has made more common.
What we did. The lean targets the mechanism rather than the week: the simulation finds teams locked into a seed or eliminated, and their games from week 15 on are forecast on the market's weight. The blanket week 18 treatment rests on three seasons and is being watched rather than trusted.
What would confirm it. The blended week 18 forecast scoring at least as well as the market that week.
Rookie quarterbacks were treated as disasters
What we found. Rookies started at a value of zero against a team value built by the previous starter, so Daniels was -25 in his second game, Dart -40, Ward -32. Three of the 15 worst misses. Fitted on 2021-25 rookie starters: first pick ≈ 40, late picks low 20s, undrafted 15, league mean start 49.
What we did. Rookies start at a draft-slot prior from nflverse draft data. The replay gained about 110 points from this and the warm-up fix together.
What would confirm it. Teams starting a rookie winning about as often as forecast, not less.
Teams change faster than a one-third offseason reversion assumes
What we found. Replaying with half reversion instead of a third gained 37 points over three seasons (55 before the QB fix). Teams Elo lagged the market on most were regime changes: CIN, WSH, NE, NO, ATL, DET.
What we did. Offseason reversion is one-half.
What would confirm it. Elo close to the market in weeks 1-4, where it was worst (Brier 0.230 vs 0.223).
The safest weekly pick is where both agree hardest, not where Elo is boldest
What we found. One pick a week over 27 seasons: strongest agreement went 434-76 (85.1%) against a 79% forecast, steady across eras. Biggest disagreement went 246-264 (48%) while Elo claimed 65%. Our 2023-25 replays ran hotter (59-6) but the long run is the number to expect.
What we did. The lock of the week uses the agreement rule with a 2-point nudge. The bold call keeps the disagreement rule as the contrarian card.
What would confirm it. Locks hitting around 84% over a season; 80% would still be fine.
The bye bonus is fine
What we found. On 2023-25 it looked shaky: home teams off a bye won 51% while Elo said 61% (80 games). Over 27 seasons, using ten or more days of rest as the bye, it is about right: home teams off a bye won 57.7% against 60.3% said (582 games), away teams 46.3% against 44.7% (643).
What we did. Left at +25. The recent shortfall was noise.
What would confirm it. Nothing to confirm; closed by the long backtest.
Monday night home teams underperform
What we found. On 2023-25 Monday night home teams won 49% against a 58% Elo forecast (63 games). Over 1999-2025 there is nothing there: 448 Monday games, home teams won 55.1% against Elo's 56.6% and the market's 55.4%, and both halves of the period agree. Sixty games was a lead; four hundred say it was noise.
What we did. Nothing, and the live tally stays as a check on 2026.
What would confirm it. It would take several seasons in the same direction to reopen this.
Thursday night home teams overperform
What we found. On 2023-25, 64% against a 59% forecast (59 games). Over 1999-2025, 267 Thursday games: home teams won 56.2% against Elo's 55.5%, within a point in both halves of the period. Nothing to act on.
What we did. Nothing.
What would confirm it. A tally is kept for 2026 as a check, not as evidence.
Home field may be smaller than 48 points
What we found. Tuning preferred a home-field advantage of 35 Elo (about 1.4 points of spread) over 48, worth 13 points across three seasons, inside the noise. Home teams won 55% overall; Elo said 56%.
What we did. Left at 48, 538's number. Silver's ELWAY uses 1 to 3.5 points varying by team.
What would confirm it. Home teams winning a point or two less often than Elo says across the season.
The QB adjustment might deserve more weight
What we found. Turning the QB adjustment off costs 357 points over three seasons, so it clearly matters. Raising the multiplier from 3.3 to 4.5 gained 13 points, inside the noise.
What we did. Left at 3.3. The side favored by a 25+ point QB adjustment won 56% of the time.
What would confirm it. Games with a big QB adjustment resolving toward the adjusted side more than Elo already expects.
Elo beats the market in the playoffs
What we found. Playoffs 2023-25: Elo Brier 0.199 vs market 0.212 over 39 games. Over 287 playoff games since 1999 Elo is ahead (2,038 points to the market's 1,917), and the blend at its normal weight matches Elo there, so nothing extra is needed.
What we did. Nothing. A 75% Elo weighting in January was added on the 39-game result and removed when the long run disagreed.
What would confirm it. Only worth revisiting if Elo keeps beating the market in January for several more seasons.
Home field is smaller inside the division
What we found. On 288 divisional games in 2023-25 nothing stood out. Over 2002-2025, 2,198 divisional games: home teams won 54.6% while Elo said 56.9% (standard error 1.1), against 57.2% at home outside the division. The market is closer, at 55.6%. Favorites are fine either way: 70%+ Elo favorites won 78.4% in division against 78.3% said. So the finding is not that divisional games are upsets, it is that the home edge is about two points smaller when the visitor knows the building.
What we did. Nothing yet. The System already weights the market three to one, and the market carries most of this; the leftover is worth about a point a season.
What would confirm it. Divisional home teams under Elo's number again in 2026, and a Kalshi or Polymarket price that agrees with the books on it.
The worst misses are surprises, not model errors
What we found. On 2023-25, half of Elo's 60 worst misses were decided by a field goal or less. Over 1999-2025 the worst 1% (69 games) were only a little closer than average: median margin 7 against 8, 32% within three points against 23% overall. The telling number is the market: on those same games it had the loser at 78% on average, Elo at 85%. The worst misses are games everyone got wrong, by a bit less. There is nothing to learn from them except to keep the slider off 90%.
What we did. Coin flips are their own spot on the Predictions page with advice to stay near 50.
What would confirm it. Nothing to confirm; it is a reminder.