Model-driven · Fully graded · No cherry-picking

Every pick, every week,
graded in public.

Sports EDGE (EMPIRICAL DATA-DRIVEN GAME ENGINE) predicts every NCAA football game straight up, against the spread, and on the total — from a market-anchored blend (10% model / 90% market) sealed before kickoff — then publishes the result, win or lose.

—%
Moneyline Accuracy
—
—%
Against The Spread
—
—%
Totals Accuracy
—
Retrospective slate

Matchup Board

View full board →

Season Performance

Moneyline accuracy by week. Retrospective replay until graded live games exist.
Below average At or above average

Teams

2025 retrospective replay · top teams by moneyline accuracy
All D-I teams →

How It Works

Plain-English version. The model runs on structured data, not intuition.
01

Team strength

A market-anchored blend (10% model / 90% market), updated only from completed games before each kickoff — no same-day leakage.

02

Consensus lines

Presence-aware consensus for moneyline, spread, and total. Picks are sealed before kickoff.

03

Public grading

Each game is graded on the line of the last revision sealed before kickoff. Replay numbers are labeled and never mixed with the live ledger.

For entertainment and informational purposes only. Replay results are not live skill and do not claim edge — this is a statistical model, not betting advice.