ST Model Accuracy

Out-of-sample backtests of the ST Model's calibrated engine — projections vs. what actually happened, and vs. Vegas closing lines
Player projections
Beat the naive baseline by 16% on per-game accuracy · p10–p90 coverage 81% (2025 backtest, 7,247 preds)
Team margins vs Vegas
Model MAE 10.72 · in-season 10.11 vs Vegas 9.84 (2,127 games)
Player Accuracy
per-game rate vs naive · 7,247 predsskill 16%p10–p90 cov 81%
StatNModel/gBase/gSkillCov
Passing
Pass Yards3731.0237.55+17%73%
Pass TDs370.30.47+36%78%
Pass INTs370.20.24+15%92%
Pass Att373.34.27+23%78%
Pass Cmp372.353.01+22%70%
Rushing
Rush Yards1299.4910.95+13%82%
Rush TDs1290.130.15+14%85%
Rush Att922.212.72+19%76%
Receiving
Receptions3350.670.75+12%79%
Rec Yards3357.899.29+15%81%
Rec TDs3350.10.12+21%85%
Targets3350.921.01+9%81%
Defense
Sacks6540.080.1+18%80%
QB Hits6540.140.16+11%79%
Tackles for Loss6540.130.16+18%80%
Solo Tackles6540.540.63+15%81%
Assist Tackles6540.490.57+15%82%
Interceptions6540.040.04+16%87%
Passes Defended6540.130.15+10%74%
Forced Fumbles6540.040.05+25%87%
Kicking
FG Made280.410.66+38%79%
FG Att280.430.73+41%86%
XP Made280.530.84+37%82%
Punting
Punts280.650.59-10%86%
Punt Yards2831.6530.02-5%86%
Fantasy Board — Preseason Backtest vs Market
reconstruction · 10 seasons
PositionRank qualityADPTop-12 hitsADPBustsADP
Overall0.5700.4604.14.45.35.2
QB0.5050.3657.47.11.31.4
RB0.5600.4836.06.72.92.7
WR0.5770.4825.75.93.83.8
TE0.6300.3126.86.62.63.0
SeasonOverallQBRBWRTE
20250.461 / 0.3850.207 / 0.0690.518 / 0.5530.375 / 0.3390.536 / 0.149
20240.517 / 0.4500.523 / 0.3770.619 / 0.5820.503 / 0.4300.500 / 0.559
20230.630 / 0.5060.476 / 0.2890.523 / 0.4090.681 / 0.6240.755 / 0.198
20220.613 / 0.5500.741 / 0.4530.641 / 0.5320.776 / 0.5810.612 / 0.350
20210.611 / 0.4990.704 / 0.6770.564 / 0.5090.490 / 0.4150.643 / 0.297
20200.547 / 0.4700.622 / 0.5140.506 / 0.5120.589 / 0.5070.681 / 0.454
20190.565 / 0.5490.373 / 0.3220.681 / 0.7120.570 / 0.3990.714 / 0.379
20180.585 / 0.4470.414 / 0.3700.428 / 0.3430.718 / 0.6290.509 / 0.228
20170.617 / 0.4550.459 / 0.2160.665 / 0.5180.567 / 0.4790.635 / 0.240
20160.550 / 0.2890.529 / 0.3660.453 / 0.1570.498 / 0.4140.716 / 0.270
Board vs the market's consensus ADP, each graded against realized PPR finishes. Seasons before 2026 are reconstructions; from 2026 each row grades the board as published.
Game Model — Out-of-Sample Backtest
backtest
SeasonGamesBrierBaseWin%Margin MAETotal MAEVegas MAEBeat%
20252720.22220.248666%10.159.7251.3%
20242720.20170.248970%9.739.6159.7%
20232720.22390.247067%10.259.9050.4%
20222710.22080.246765%9.158.7447.9%
20212720.22300.249862%10.9010.7851.9%
20202560.22280.250065%10.409.8351.6%
20192560.21560.249866%10.3710.2152.8%
20182560.20800.240567%10.009.9851.4%
Overall2,1270.21730.248566%10.119.8452.1%
The model's pre-game win probability is scored by Brier against every completed game's outcome; a Brier below the home-field Basemeans it adds signal beyond "home teams win." Margin/Total MAE measure the projected score. Vegas MAE / Beat% benchmark the projected margin against the closing line — the model never sees the line, it's stored only to grade honestly.