Our NCAAB totals model prices every college basketball game from tempo data up, before books adjust for public action.
Each layer targets a distinct inefficiency that raw totals, and books catering to the public, systematically miss.
We measure true pace: possessions per 40 minutes, adjusted for opponent tempo. Fast teams playing slow teams distort raw totals. We account for it.
Offensive and defensive efficiency ratings, adjusted for strength of schedule. A 70-point team in the Big 12 scores differently than in the Sun Belt.
Short-term variance in shooting and turnovers obscures true scoring rates. We regress toward stable tempo and efficiency signals to filter noise.
From raw tempo data to a betting edge. Five stages, no black boxes.
Raw game totals are meaningless without pace context. We calculate adjusted possessions per 40 minutes for every D1 team, weighting recent games more heavily via exponential decay.
A fast-tempo team playing a slow opponent won't get to run its offense. The game pace converges toward the slower team's preferred rhythm. Our model projects the matchup pace directly, not each team's season average in isolation.
Points per possession, adjusted for opponent defense. We separate volume scoring from efficient scoring. 85 points at 60 possessions is far more dangerous than 85 points at 75.
A team with a gaudy point total that plays in a fast conference faces more possessions per game and naturally scores more. Raw totals conflate volume with quality. Our efficiency metric normalizes for pace before any comparison is made.
Every team's offense faces an opponent's defense. We model the matchup directly: how does this offense's efficiency map against this defense's specific weaknesses?
Averaging each team's season efficiency numbers gives you a rough estimate. Our matchup model directly applies each offense's strengths against each defense's measured vulnerabilities, in the half-court, on transition, and from three. The matchup card shows the result.
College basketball teams play brutal schedules. A team on their third game in five days, 1,200 miles from home. Their offensive execution suffers in measurable ways.
These adjustments are applied to the model total after the base projection is established. They are not applied to teams individually. They represent net expected total shifts given the game context, calibrated from historical results in comparable situations.
When our model total diverges from the opening line by a meaningful margin, and the line hasn't moved toward us yet, that's our window. We track 67% CLV across all plays.
CLV confirmation means the closing line moves toward our number after we release, independent evidence that our model identifies real edge, not noise. 67% is the highest CLV rate of any Whizard model.
Every play logged against closing line value. Lines moving toward our number after release is real confirmation of edge, not just win/loss noise.
First full season tracked. 67% CLV, the highest of any Whizard model, confirms consistent edge over closing line.
Conference tournaments and March Madness are the most mispriced stretch of the college basketball calendar. Public money floods the market. Books shade toward popular teams. That is exactly when this model has the most edge.
Get CBB TotalsTotals picks released when lines drop, positioned ahead of line movement to maximize CLV, with edge magnitude and lineup context reviewed before each release.
Picks are posted when lines first become available, not on a fixed morning schedule. Getting on before public money moves the number is the entire CLV strategy. Timing is the edge.
Every release shows the model total, the book opening, the direction of our play, and the size of the edge in points. No ambiguity. You see exactly why the play clears our threshold.
Major injury reports are reviewed before each release. Lineup scratches that meaningfully affect pace or efficiency are factored in. We don't release stale model outputs when significant information has changed.
All picks land directly in the Whizard subscriber Discord. Each post includes the model output card, edge breakdown, and unit recommendation. Season runs through the NCAA Tournament in April.
NCAAB Totals picks released when lines drop. Season runs through April.