Omaha 0.6.0
Released 3 September 2026
Running back workloads are noticeably closer to reality, because the simulation now decides whether a play is a run or a pass before it decides who is on the field for it.
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.
The change column is simply this version’s accuracy minus 0.5.0’s, each measured across everything that version covers, so the two columns you can see always account for it. We also measure the change on shared situations alone, which is a more sensitive test; where the two disagree we publish this one.
Accuracy by stat
| Stat | vs. season average | Change from 0.5.0 |
|---|---|---|
| Receiving touchdownsrare event | +15.0% | −0.11−0.67 to +0.39 |
| Rushing touchdownsrare event | +13.1% | +0.54+0.50 to +2.50 |
| Passing touchdownsrare event | +7.6% | +0.75−0.38 to +1.62 |
| Passing yards | +7.4% | +0.49−0.85 to +1.83 |
| Completions | +3.2% | +1.92+0.32 to +3.53 |
| Pass attempts | +3.0% | +0.95−0.81 to +2.66 |
| Receiving yards | −4.2% | +0.29+0.11 to +0.53 |
| Rushing yards | −11.1% | +1.32+0.88 to +1.88 |
| Receptions | −21.4% | +0.20−0.14 to +0.43 |
| Targets | −27.1% | +0.31−0.11 to +0.58 |
| Rush attempts | −38.8% | +2.72+1.98 to +3.54 |
| Average across all figures | −4.9% |
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
A blocking back is on the field because the play is a pass. The simulation used to choose personnel without knowing the play type, so it fielded the pass-protection back on running downs and then handed him the ball. Lead backs were given about 62% of the carries they really get, and the carries went to the wrong teammate. Play type is now drawn first, and both the personnel on the field and the play itself follow from it.
Rush attempt accuracy improved by 2.8 points (95% interval 2.0 to 3.5). Measured on one team, a lead back went from 9.9 simulated carries per game to 11.4 against an actual 12.0, while the pass-down back fell from 6.1 to 3.6 against an actual 2.7.
Attributing carries to the right back moves the yards and scores with them. No change was made to how far a run gains or how often it reaches the end zone.
Rushing yards improved 1.4 points (0.9 to 1.9); rushing touchdowns 1.5 points (0.5 to 2.5).
Knowing the play type before choosing personnel also puts the right receivers on the field for passing downs. This was not the change's purpose and the effect is smaller, but it is real.
Receptions improved 1.9 points (0.3 to 3.5). Simulated pass attempts moved from 35.6 per game to 37.8, against an actual 38.0 for the teams measured.
Effectively unchanged. Their intervals include zero, so we are not claiming a movement in either direction.
What this version still gets wrong
- Rush attempts remain the least accurate figure we publish, at −38.8% against a player's own season average. This release improved it; it did not fix it.
- Play calling still does not respond to the score. Real teams pass on about 78% of snaps when well behind and 33% when well ahead; the simulation stays near 57–60% regardless. Late-game situations are the weakest part of the engine.
- When the play type is known, the simulation still estimates the quality of the two units using a blend of run and pass ability rather than the one that applies. This is measurable and is the next thing being worked on.
- Receiver and tight end carries remain under-simulated (0.6 per game against an actual 1.5). The volume is small and it has not been prioritized.