Model rank 1
Marcus Vale
77.7%
anytime-TD probability
Touchdown intelligence
Anytime-touchdown probability from four public factors, weighted in the open and pushed through a stated logistic. Hand the board to a skeptic and they can recompute every number without trusting us.
What we will not do: show you a hit rate we have not earned on our own data, or print an “edge” against a market price we do not have a license to read. Where the board is unpriceable it says so.
Goal-line intelligence
Every score below is recomputable by hand: a weighted composite of the four published factors, pushed through a stated logistic.
prob = logistic((score − 0.5) × 6)
Model rank 1
Marcus Vale
77.7%
anytime-TD probability
Model rank 2
Deon Pryce
75.0%
anytime-TD probability
Model rank 3
Tariq Bell
63.9%
anytime-TD probability
Edge is model probability minus the vig-stripped market probability. Rows without a market price show an em-dash — an unpriceable board is reported, never dressed up.
Educational modelling on an illustrative player pool — the names are fictional, so no probability is ever pinned to a real person. Past results do not predict future results, and nothing here is a promise an outcome. Building the proof layer? See the public CLV report and the calibration report.