Backtest Results - Monday, February 23, 2026
SPREAD PICKS SCORECARD
| Pick | Result | Outcome |
|---|---|---|
| HOU -12.5 | Jazz 105 - Rockets 125 | Covered by 7.5 |
| MEM -3.5 | Kings 123 - Grizzlies 114 | Failed to cover by 12.5 |
| SAS +1.5 | Spurs 114 - Pistons 103 | Covered by 12.5 |
Spread Record: 2-1 (+0.9u)
OVER/UNDER PICKS SCORECARD
| Pick | Result | Outcome |
|---|---|---|
| SAC @ MEM UNDER 232.5 | Kings 123 - Grizzlies 114 (Total: 237) | Over by 4.5 |
| UTA @ HOU UNDER 229.5 | Jazz 105 - Rockets 125 (Total: 230) | Over by 0.5 |
| SAS @ DET OVER 232.5 | Spurs 114 - Pistons 103 (Total: 217) | Under by 15.5 |
O/U Record: 0-3 (-3.3u)
PARLAY RESULT
| Leg | Pick | Outcome |
|---|---|---|
| 1 | Houston Rockets -12.5 | UNMATCHED |
| 2 | Sacramento @ Memphis UNDER 232.5 | UNMATCHED |
Parlay Result: LOSS (-0.5u)
DAILY SUMMARY
| Category | Record | Units |
|---|---|---|
| Spreads | 2-1 | +0.9 |
| Over/Under | 0-3 | -3.3 |
| Parlay | LOSS | -0.5 |
| Daily Total | -2.9 |
QUALITATIVE ANALYSIS
WHAT WORKED
- SAS +1.5 WIN: Model correctly identified Detroit's back-to-back fatigue as a major factor, with the Spurs winning outright by 11 points
- HOU -12.5 WIN: Properly weighted Utah's massive injury situation (5 key players out) against a motivated home team
WHAT MISSED
- MEM -3.5 LOSS: Severely underestimated Sacramento's ability to score despite injuries - Kings put up 123 points with depleted roster, suggesting model overvalues injury impact on offense
- All totals MISSED (0-3): Complete failure on over/under predictions, with games going opposite direction by significant margins (4.5, 0.5, 15.5 points)
This appears to be a systematic methodology flaw - the model's pace/scoring projections are fundamentally flawed.
MODEL CALIBRATION NOTES
- Confidence ratings missing: No confidence levels were provided for most picks, making calibration assessment impossible
- Injury weighting bias: Model appears to overweight injuries for totals (expecting lower scoring) but underweight them for spread purposes when bad teams exceed expectations
METHODOLOGY ADJUSTMENTS TO CONSIDER
- Revamp totals methodology: Current pace/scoring model is broken - need to incorporate team-specific offensive efficiency metrics and recent scoring trends rather than just BPI ratings
- Refine injury impact scaling: Differentiate between injury impact on offense vs defense, and consider that bad teams with injuries might still have individual players step up offensively
Record: 2-4 spread/totals, 0-1 parlay. Major systematic issues with totals predictions need immediate attention.
CUMULATIVE SEASON RECORD
| Category | Record | Win % | Units |
|---|---|---|---|
| ATS | 37-31 | 54.4% | -2.3 |
| O/U | 35-36 | 49.3% | |
| Parlays | 4-22 | ||
| Season Total | -2.3u |