Omaha 1.1.0
Released 4 September 2026
Play calling now reads the clock as well as the scoreboard. The gain is small overall and much larger late in games, which is exactly where it should be.
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 1.0.0’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 1.0.0 |
|---|---|---|
| Receiving touchdownsrare event | +20.4% | −0.03−0.13 to +0.21 |
| Rushing touchdownsrare event | +17.6% | −0.02−0.57 to −0.06 |
| Receiving yards | +12.8% | −0.09−0.05 to +0.09 |
| Rushing yards | +10.7% | −0.14−0.29 to +0.03 |
| Passing yards | +9.6% | +0.63+0.24 to +1.07 |
| Pass attempts | +9.6% | +1.72+1.27 to +2.24 |
| Passing touchdownsrare event | +7.4% | −0.32+0.93 to +2.23 |
| Completions | +7.1% | +1.25+0.81 to +1.73 |
| Receptions | +1.3% | −0.02+0.03 to +0.24 |
| Rush attempts | +1.2% | +0.14−0.11 to +0.39 |
| Targets | −0.4% | +0.47+0.42 to +0.69 |
| Average across all figures | +8.8% |
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
Until now the simulation knew the score but not the time, so a two-score lead in the first quarter and the same lead with ninety seconds left were the same situation to it. They are nothing alike. A team two scores ahead throws on 62% of snaps in the first half and 13% inside two minutes.
Across the clock, within a single margin band, real pass rates move between 0.17 and 0.49, more than the 0.21 they move across the margin bands we already modeled. Pass attempt accuracy improved 1.7 points, completions 1.3, passing yards 0.6, targets 0.5.
Knowing there are ninety seconds left is worth nothing in the first quarter and a great deal in the fourth, so the benefit is concentrated where games are actually watched.
Averaged across all eleven figures, accuracy in the final quarter improved from 15.0% to 16.4% better than a player's season average, nearly five times the improvement across the game as a whole (8.5% to 8.8%).
Still deliberately flat. It is a running down whatever the clock says, and the adjustment is applied in a way that moves an already-extreme situation very little.
What this version still gets wrong
- Targets, at −0.4%, remains the one figure slightly worse than a player's own season average. It improved by half a point and is very close.
- The overall gain is 0.3 points, which is small. Situational play-calling detail is reaching the end of what it can add. The remaining gaps are about how often a specific player touches the ball, not about what his team does.
- Our numbers remain weakest before kickoff and strongest late. That is inherent: before a game there is no situation to read.
- We still do not model WHERE a run goes. Runs to the edge gain 5.3 yards on average and runs between the guards 4.1, and teams differ enormously and consistently in which they prefer. This is the next thing being built.