Ask anything. The AI references our math.

A +5% post-tax edge means the bookmaker is offering odds that imply a probability ~5% below what Pinnacle's sharp fair odds suggest. Concretely: if a market gives you 2.20 and Pinnacle's de-vigged fair odds are 2.10, you're being paid ~5% more than the true expected price. For sizing, the Kelly Criterion would suggest a stake fraction of roughly: kelly = (book_odds × fair_probability − 1) / (book_odds − 1) ≈ (2.20 × 0.476 − 1) / (2.20 − 1) ≈ 0.039 → 3.9% of bankroll (full Kelly) Almost no one bets full Kelly because the variance is brutal. BetEdge defaults to Quarter Kelly (~1% of bankroll on this bet), which captures ~88% of the long-term growth rate with about 75% lower drawdown. The edge is the math. The sizing is the discipline. Both matter.

Win rate over a short sample is noise. A bettor with a real 5% edge will still lose 47% of the time at even odds — and might hit a 12-bet losing streak in their first 100 picks just by random variance. Closing Line Value (CLV) measures whether your entry price beat Pinnacle's final closing line. Since Pinnacle's close is the sharpest publicly available probability estimate, consistently beating it proves your process finds real edges — independent of which specific bets won or lost. Pro syndicates track CLV as their primary KPI, not ROI. ROI converges to your true edge only after thousands of bets. CLV signals it within hundreds. We capture closing-line snapshots automatically (every 15 min) and compute CLV per settled pick. Average CLV across our public track record is the honest answer to "is this guy actually sharp?"