RABBIT NETWORK · NCAA DIVISION I WOMEN'S VOLLEYBALL
Every team. Every game.
No hindsight.
Independent national rankings, frozen Winner Picks, live scores, verified results and a permanent public grading record across all 348 Division I programs.
Today's Games
0–0 —
No eligible finals yet
On the court now
Risers & fallers
Risers
Fallers
The forecast gets graded
People driving the numbers
Rabby Winner Picks
Loading the Sept. 13 preboard…
Rabby Volleyball 348
The official published ranking entering the day. Rabby Strength is the underlying cardinal team-strength rating; the ordinal rank is the public 1–348 order.
Rabby vs. the rankings world
Compare Rabby's full 1–348 predictive ranking with published polls and independent rating systems. Disagreement is the point: it shows where Rabby sees a team differently from outside consensus.
Rabby vs. AVCA
Rabby edge means Rabby ranks the team higher than the external poll. Poll edge means the external poll ranks it higher. These rankings remain completely outside the production model.
Rabby sees them. AVCA does not.
AVCA sees them higher.
Teams just outside the AVCA Top 25
National team directory
Every Division I program has a Rabby profile. Because “national” is supposed to mean something.
Results & grading
Every Sept. 11 eligible pregame prediction is preserved exactly as frozen. Finals are attached afterward, then scored. The seven late-lock display-only predictions remain ungraded.
Performance lab
Probability calibration, confidence-band performance and the misses that matter. Every certified daily board is graded out-of-sample and preserved without hindsight.
What Rabby is getting right, and what we are auditing
Production stays frozen. Shadows get graded beside it.
No shadow can alter a published pick. Promotion requires a larger forward sample plus better calibration, Brier score and log loss.
The performances behind Rabby’s biggest hits and misses
Forecast probability vs. actual results
Where Rabby was wrong
Statistical Performance Lab
Attack efficiency, point production, first-contact pressure, floor defense and blocking context. These numbers do not alter the active frozen Elo/network probabilities.
Current ingestion audit
Sept. 8: all 31 matches are represented; 60 of 62 team-game performances have the complete five-feature diagnostic set. Butler–Northwestern remains explicitly source-limited. Sept. 9: the expanded context snapshot covers 99 team rows. Sept. 10: the certified production cycle adds a separate player/stat context release for the latest completed day. Missing fields stay marked as missing rather than being fabricated.
Current ingestion audit
Data Center
The operating picture behind Rabby Volleyball 348: source discovery, verified adapters, live-capability coverage and last-success evidence. Discovery is not treated as proof of live support.
What the source mesh can actually retrieve
Team-by-team production status
National providers and verification ladder
Rabby archive
Every major release is preserved. New evidence creates a new release; it does not rewrite an old one.
Latest certified-day trajectory
Players who shaped the board
Players highlighted in official team previews and the Sept. 8 game-day briefing, preserved alongside the results.
Rabby vs Market
When legal/public prices are available, Rabby is compared with the no-vig market probability without changing the frozen forecast.
No usable public WNCAA moneylines populated yet.
The board is ready. When both sides of a market appear, we convert the moneylines to implied probability, remove vig and calculate the percentage-point disagreement.
A prediction system with a memory.
Rabby’s core rule is simple: preserve what the model knew, when it knew it, and grade it later without rewriting history.
Predictive
Active production backbone: Elo/network. This creates the frozen game probabilities and predictive ranking.
Statistical
Attack, point/set, first contact and opponent-adjusted performance remain a distinct diagnostic layer. The ingestion audit now tracks the Sept. 8 game-level recovery and the expanded Sept. 9 team-context snapshot.
Resume
Quality wins, losses, strength of schedule and opponent network are evaluated separately from prediction probability.
Power
The eventual synthesis. Predictive, statistical and resume evidence stay distinct so we can see why a team is rated where it is.
The non-negotiables
- Complete national slate before a daily freeze.
- Every certified game gets a prediction.
- Predictions become immutable before first serve.
- External systems and Vegas are audits, not hidden ingredients.
- Low-confidence games are labeled instead of disguised.
- Historical releases are preserved rather than overwritten.
Sept. 8 briefing sources
Team identity marks use a static Rabby logo manifest. Official ESPN/NCAA marks are attempted first; every Division I program also has a bundled local team badge so no logo position can render blank or broken.

