Omaha 2.9.0
Released 17 October 2026
A backup who plays and records nothing now counts, both in the averages our projections lean on and in how we score ourselves. Backup tight ends and receivers were projected for noticeably more than they got, and that is mostly fixed. The figures on this page now use the stricter yardstick, so they are not comparable with earlier versions.
How to read these numbers
Every figure below compares the simulation against a player’s own season average so far, scaled to how much of the game is left, roughly what a reader could work out without us. Zero means we add nothing over that. Positive means we beat it; negative means we are worse than it.
Measured across the 2024 season, weeks 1–18, rebuilding the model each week using only what was known at the time. 400 simulations per situation.
Accuracy by stat
| Stat | vs. season average | Change from 2.8.0 |
|---|---|---|
| Receiving touchdownsrare event | +21.0% | — |
| Rushing touchdownsrare event | +18.4% | — |
| Receiving yards | +15.4% | — |
| Rushing yards | +13.2% | — |
| Pass attempts | +12.2% | — |
| Passing yards | +10.4% | — |
| Completions | +9.0% | — |
| Passing touchdownsrare event | +6.6% | — |
| Rush attempts | +5.1% | — |
| Receptions | +4.6% | — |
| Targets | +4.0% | — |
| Average across all figures | +10.9% |
The average is an unweighted mean across the figures above. It is a summary of how we are doing, not a single score for the model: yards and counts are not measured in the same units, so combining them any more cleverly than this would just let passing yards decide the answer.
Each change carries the range the measurement can actually support. A change whose range crosses zero is shown in neutral rather than as a gain: we cannot tell it apart from no change, and we are not claiming it as one.
What changed
Before kickoff each projection is pulled partway toward the player's own season average. That average only counted games where he recorded something, so a backup who played in eight games and caught passes in three was averaged over three. His average, and the projection pulled toward it, came out well above what he really does in a typical game. The average now counts every game he took a snap in, at zero when he did nothing.
Backup tight ends were projected for 1.33 times the targets they got in 2024 and 1.24 times in 2025; now 1.02 and 1.00. Backup receivers go from 1.10 to 0.98 (2025: 1.16 to 1.03). Scored the same way before and after, backup tight ends' target forecasts improve by 4.6 points in 2024 and 3.0 in 2025, and receptions by 3.5 and 2.0. For players with nothing yet this season, targets improve by 19 points and 9. All of these ranges are clear of zero.
We used to score a player only in games where he recorded a stat. That quietly dropped exactly the games where a backup dressed, played and did nothing, which is the miss an over-projection produces, so our own scoring could not see the error above. Every player who takes a snap is now scored, at zero when he records nothing, and the season average we are measured against counts those games too.
Both sides of the comparison changed, so the figures on this page measure a slightly different question from earlier versions'. That is why this release shows no change column. The improvements above are stated on one fixed yardstick instead.
The simulation is unchanged: every simulated game is identical to 2.8.0. What moved is the season average projections lean on before kickoff, and who we score. Starters play nearly every game, so their averages barely change, and their forecasts move by less than 0.15 points.
What this version still gets wrong
- Not comparable with earlier versions. Some figures here are lower than 2.8.0's (passing attempts, for example) because more players are now scored and the season average we compare against got better, not because a projection got worse.
- This corrects the average projections lean on, not the simulation. On its own the simulation still projects backup tight ends about 40% high, and the correction fades as a game goes on and the projection leans on the simulation instead.
- Backup running backs' receptions now come out slightly low in 2025, at 0.90 to 0.95 of what they caught.
- Rushing figures move inside the normal run-to-run variation.
- Touchdown figures move only because more players are scored, not because touchdowns are projected differently.