Bradley-Terry Recency
Static Bradley-Terry with exponential recency weights on newer games.
| Status | implemented | Short Name | BTr |
|---|---|---|---|
| Rating Style | elo | Where Shown | teams rankings, team pages, model performance |
Model Card
Same paired-comparison form as Bradley-Terry, but newer games receive more weight. This is the simplest bridge between static BT and fully dynamic team models.
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: All Games. All scored games in the selected league and season. AP Poll is excluded 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 | 0.792 (#10) · 3 games |
| Rolling Avg Log Loss | 0.633 (#6) · 33 windows |
| Window Wins | 1 · 64.5% 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 NBA 2025-2026. The focus line is Bradley-Terry Recency. Click legend items to hide/show peers.
Weekly strict holdout log loss. Lower is better. Showing 6 models across 33 windows.
Recent Windows
| Holdout | Log Loss | Rank | Winner |
|---|---|---|---|
| Jun 3 | 0.687 | #2/16 | Elo (0.636) |
| May 27 - Jun 2 | 0.885 | #14/16 | Home Team Baseline (0.714) |
| May 20 - May 26 | 0.588 | #12/16 | Adjusted Context Blend (0.551) |
| May 13 - May 19 | 0.655 | #2/16 | Elo (0.615) |
| May 6 - May 12 | 0.585 | #13/16 | Log Adjusted (0.495) |
| Apr 29 - May 5 | 0.706 | #15/16 | Points Off/Def (0.654) |
| Apr 22 - Apr 28 | 0.634 | #13/16 | Adjusted Context Blend (0.589) |
| Apr 15 - Apr 21 | 0.709 | #2/16 | Home Team Baseline (0.646) |
Selected Scope Peer Table
Comparison snapshot for NBA 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 Recency | - | - | 0.695 | #2 | 0.622 | #1 | 33 |
| Points Off/Def Recency | - | - | 0.694 | #1 | 0.622 | #2 | 33 |
| Recency Ensemble | - | - | 0.749 | #4 | 0.623 | #3 | 33 |
| Core Ensemble | - | - | 0.771 | #6 | 0.625 | #4 | 33 |
| Avg Margin Baseline | 0.757 | #1 | 0.767 | #5 | 0.633 | #5 | 33 |
| Bradley-Terry Recency | - | - | 0.792 | #10 | 0.633 | #6 | 33 |
| Log Adjusted | 0.747 | #6 | 0.809 | #13 | 0.633 | #7 | 33 |
| Adjusted Efficiency | 0.747 | #5 | 0.810 | #14 | 0.633 | #8 | 33 |
League × Season Summary
Cross-scope validation history for Bradley-Terry Recency. Rolling ranks are by average weekly log loss.
| League | Season | Full AUC | AUC Rank | 7d LL | 7d Rank | Rolling LL | Rolling Rank | Windows | Wins |
|---|---|---|---|---|---|---|---|---|---|
| WNBA | 2026 | - | - | 0.652 | #4 | 0.648 | #1 | 2 | 1 |
| WNBA | 2025 | - | - | 0.441 | #2 | 0.619 | #2 | 21 | 0 |
| NCAAWD2 | 2025-2026 | - | - | 0.527 | #11 | 0.581 | #8 | 14 | 0 |
| NCAAW | 2025-2026 | - | - | 0.639 | #14 | 0.596 | #14 | 22 | 0 |
| NCAAMD2 | 2025-2026 | - | - | 0.580 | #12 | 0.610 | #13 | 14 | 0 |
| NCAAM | 2025-2026 | - | - | 0.727 | #16 | 0.643 | #15 | 22 | 0 |
| NBA | 2025-2026 | - | - | 0.792 | #10 | 0.633 | #6 | 33 | 1 |
| GLEAGUE | 2025-2026 | - | - | 0.933 | #15 | 0.682 | #9 | 22 | 0 |
Output Coverage
Stored site-output rows for this model.
| League | Season | Rows |
|---|---|---|
| WNBA | 2025 | 17 |
| NCAAWD3 | 2025-2026 | 730 |
| NCAAWD2 | 2025-2026 | 697 |
| NCAAW | 2025-2026 | 592 |
| NCAAMD3 | 2025-2026 | 781 |
| NCAAMD2 | 2025-2026 | 603 |
| NCAAM | 2025-2026 | 704 |
| NBA | 2025-2026 | 37 |
| GLEAGUE | 2025-2026 | 31 |