🐻⬇️🏀

Team model

Dynamic Bradley-Terry

Time-evolving paired-comparison model with latent team strength drift.

Status implemented Short Name DynBT
Rating Style elo Where Shown teams rankings, team pages, model performance

Model Card

Glicko-style paired-comparison model with team uncertainty that expands during idle stretches and tightens with new games. This is the first explicit temporal paired-comparison model beyond static BT plus recency weighting.

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 0.603 (#9) · 14 games
Rolling Avg Log Loss 0.555 (#8) · 22 windows
Window Wins 0 · 70.5% 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 NCAAW 2025-2026. The focus line is Dynamic Bradley-Terry. Click legend items to hide/show peers.

Log Loss Brier AUC Accuracy

Weekly strict holdout log loss. Lower is better. Showing 6 models across 22 windows.

Recent Windows

Holdout Log Loss Rank Winner
Apr 1 - Apr 5 0.713 #4/16 Elo (0.673)
Mar 25 - Mar 31 0.545 #10/16 Adjusted Context Blend (0.474)
Mar 18 - Mar 24 0.612 #11/16 Adjusted Efficiency (0.431)
Mar 11 - Mar 17 0.665 #10/16 Margin Recency (0.603)
Mar 4 - Mar 10 0.516 #10/16 Recency Ensemble (0.500)
Feb 25 - Mar 3 0.505 #8/16 Recency Ensemble (0.484)
Feb 18 - Feb 24 0.525 #5/16 Recency Ensemble (0.515)
Feb 11 - Feb 17 0.523 #11/16 Core Ensemble (0.485)

Selected Scope Peer Table

Comparison snapshot for NCAAW 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
Core Ensemble - - 0.573 #4 0.499 #1 22
Recency Ensemble - - 0.566 #3 0.500 #2 22
Margin Recency - - 0.539 #1 0.515 #3 22
Margin 0.878 #1 0.588 #7 0.517 #4 22
Points Off/Def 0.870 #5 0.578 #5 0.542 #5 22
Adjusted Context Blend - - 0.631 #12 0.543 #6 22
Points Off/Def Recency - - 0.539 #2 0.551 #7 22
Dynamic Bradley-Terry - - 0.603 #9 0.555 #8 22

League × Season Summary

Cross-scope validation history for Dynamic Bradley-Terry. 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 2025 - - 0.549 #6 0.628 #5 21 1
NCAAWD2 2025-2026 - - 0.506 #8 0.569 #6 14 2
NCAAW 2025-2026 - - 0.603 #9 0.555 #8 22 0
NCAAMD2 2025-2026 - - 0.564 #10 0.597 #10 13 0
NCAAM 2025-2026 - - 0.688 #14 0.595 #10 22 0
NBA 2025-2026 - - 0.590 #3 0.618 #6 26 1
GLEAGUE 2025-2026 - - 0.860 #14 0.691 #13 22 3

Output Coverage

Stored site-output rows for this model.

League Season Rows
WNBA 2025 17
NCAAWD3 2025-2026 584
NCAAWD2 2025-2026 524
NCAAW 2025-2026 592
NCAAMD3 2025-2026 646
NCAAMD2 2025-2026 510
NCAAM 2025-2026 704
NBA 2025-2026 37
GLEAGUE 2025-2026 31