Adjusted Context Blend
Experimental context-heavy win model blending strong team components with rest and venue context.
| Status | implemented | Short Name | Adj+ |
|---|---|---|---|
| Rating Style | probability | Where Shown | model performance |
Model Card
Context-heavy second-stage win model that blends our strongest component ratings with rest and venue context. This remains an exploratory team baseline: useful in the bakeoff and model pages, but not yet clean enough for the main public team-output surfaces.
Validation on this page uses stored season-level full-fit rows plus strict weekly rolling holdouts. Rolling ranks are based on average weekly log loss in the selected league/season. Current scope: Ranked Games. Only games where at least one team had an AP Poll rank on game day. All models are re-evaluated on that same subset, and AP Poll joins the comparison set here.
Selected League / Season
Use the selected scope to inspect this model against peers and across rolling weekly holdouts.
| Full Fit AUC | - · 0 games |
|---|---|
| 7-Day Holdout Log Loss | 0.463 (#1) · 5 games |
| Rolling Avg Log Loss | 0.464 (#2) · 22 windows |
| Window Wins | 6 · 77.7% rolling accuracy |
Full-fit AUC is only available where the older stored-rating batch produced that model for the selected scope. Recency and ensemble families can still have valid 7-day and rolling holdout results even when the full-fit columns are blank.
Rolling Holdout Curves
Selected model plus comparison peers for NCAAM 2025-2026. The focus line is Adjusted Context Blend. Click legend items to hide/show peers.
Weekly strict holdout log loss. Lower is better. Showing 6 models across 22 windows.
Recent Windows
| Holdout | Log Loss | Rank | Winner |
|---|---|---|---|
| Apr 1 - Apr 6 | 0.463 | #1 | Adjusted Context Blend |
| Mar 25 - Mar 31 | 0.603 | #5/16 | Log Adjusted (0.591) |
| Mar 18 - Mar 24 | 0.498 | #7/16 | Adjusted Efficiency (0.427) |
| Mar 11 - Mar 17 | 0.591 | #10/16 | Margin (0.550) |
| Mar 4 - Mar 10 | 0.453 | #1 | Adjusted Context Blend |
| Feb 25 - Mar 3 | 0.659 | #10/16 | Points Off/Def Recency (0.592) |
| Feb 18 - Feb 24 | 0.684 | #14/16 | Points Off/Def Recency (0.609) |
| Feb 11 - Feb 17 | 0.563 | #7/16 | Points Off/Def (0.517) |
Selected Scope Peer Table
Comparison snapshot for NCAAM 2025-2026. Rows are ordered by rolling log-loss rank; full-fit AUC may be blank for models that do not have legacy full-season rows stored for this scope.
| Model | Full AUC | AUC Rank | 7d LogLoss | 7d Rank | Rolling LL | Rolling Rank | Windows |
|---|---|---|---|---|---|---|---|
| Margin | 0.878 | #3 | 0.494 | #4 | 0.458 | #1 | 22 |
| Adjusted Context Blend | - | - | 0.463 | #1 | 0.464 | #2 | 22 |
| Margin Recency | - | - | 0.568 | #6 | 0.465 | #3 | 22 |
| Points Off/Def | 0.877 | #4 | 0.507 | #5 | 0.471 | #4 | 22 |
| Core Ensemble | - | - | 0.606 | #9 | 0.472 | #5 | 22 |
| Adjusted Efficiency | 0.880 | #1 | 0.493 | #3 | 0.472 | #6 | 22 |
| Log Adjusted | 0.880 | #2 | 0.490 | #2 | 0.473 | #7 | 22 |
| Recency Ensemble | - | - | 0.629 | #10 | 0.476 | #8 | 22 |
League × Season Summary
Cross-scope validation history for Adjusted Context Blend. Rolling ranks are by average weekly log loss.
| League | Season | Full AUC | AUC Rank | 7d LL | 7d Rank | Rolling LL | Rolling Rank | Windows | Wins |
|---|---|---|---|---|---|---|---|---|---|
| NCAAW | 2025-2026 | - | - | 0.461 | #1 | 0.368 | #7 | 22 | 3 |
| NCAAM | 2025-2026 | - | - | 0.463 | #1 | 0.464 | #2 | 22 | 6 |