Every total-corners and total-cards (booking-points) Over/Under where a book has posted a line: our projected probability, the best posted price and what that price implies once the margin is removed.
Cards projections are referee-adjusted where the official is known (0 fixtures) — a strict ref raises the cards line, a lenient ref lowers it. We publish no edge figure on these: no graded corners or cards ledger exists to test one against. 67 model-outlier lines flagged and sunk below.
🟨Lazio v AC Milan⚠ outlierCards Over 5.5 · proj 3.4 · 2026-09-12
13%
2.9Pinnacle
33%
🟨Le Mans v Lens⚠ outlierCards Over 5.5 · proj 3.2 · 2026-09-13
10%
2.97Pinnacle
32%
🟨Paris FC v Lyon⚠ outlierCards Over 5.5 · proj 3.0 · 2026-09-12
8%
2.97Pinnacle
32%
"Model" is our projected probability the line goes over; "Implied" is what the best posted price implies once the margin is removed. We used to print the gap between them as an edge and shade the positive ones green — both are gone, because no graded corners/cards result set with prices attached exists anywhere on this site to test that gap against.
Referee adjustment: once ESPN exposes the appointed official (kickoff onward), the cards projection is nudged toward that referee's strictness vs the WC baseline (yellows + 2×reds per game), shrunk by sample size and clamped; corners get a smaller ref nudge. Projection only where no book line is posted (shown on the match page, never as a fabricated edge).
Single-side O/U prices are de-vigged with a flat overround haircut (an approximation). Model outliers (a projection beyond a believable corners/cards gap) are flagged and sorted last. Not advice — information only.
For entertainment and informational purposes only. 21+. Please bet responsibly. Model probabilities are estimates, not guarantees. How the model works →