CLV.gg covers six sports across more than a dozen books, split into three classes: sharp reference books, prediction-market exchanges, and major US soft books. The sports are basketball, baseball, hockey, soccer, tennis, and American football, and the soccer coverage includes the 2026 World Cup. The reason the coverage is built this way is simple: edge detection compares a soft price against a fair line, and a fair line needs sharp anchors to build it. Cover only soft books and you have nothing to measure against. Cover only sharp books and there is no mispriced side to bet. You need both, which is why the book mix matters more than the raw count.
| Class | Examples | Role in detection |
|---|---|---|
| Sharp reference | Pinnacle, Circa | Anchor the no-vig fair line |
| Exchanges | Polymarket, Kalshi, sx.bet, Betfair | Fair-probability inputs, order-book depth |
| US soft books | DraftKings, FanDuel, BetMGM, others | Where the mispriced edges appear |
What sports does CLV.gg cover?
CLV.gg covers six sports: basketball, baseball, hockey, soccer, tennis, and American football. Within each, the detection runs across the markets that carry real liquidity: moneyline or three-way result, spreads and Asian handicaps, totals including push-eligible and quarter-lines, draw-no-bet, and futures. Soccer coverage includes the 2026 World Cup, which gets its own guide section because the expanded 48-team field is the densest stretch of soft pricing in the four-year cycle. The point of breadth is not a longer list. It is that more sports means more markets priced at different speeds by different books, and every speed gap is a place an edge can open.
Which books does CLV.gg track, sharp and soft?
CLV.gg tracks more than a dozen books across three classes, and the split is the whole design. Sharp reference books like Pinnacle and Circa run low-margin, high-limit models and welcome sharp action instead of limiting it, so their prices absorb information fast and sit close to fair. Exchanges like Polymarket, Kalshi, sx.bet, and Betfair price through an order book, where the mid between the best back and lay is a clean fair-probability read. Soft books like DraftKings, FanDuel, and BetMGM profile and restrict winning accounts and update their lines slower, so their prices drift further from fair. That drift is the edge. The detection builds its fair line from the sharp side and exchanges, then flags any soft book whose price beats that fair line by enough to clear the threshold.
Why does CLV.gg need both sharp and soft books?
Because detection is a comparison, and a comparison needs two sides. A fair line is the no-vig probability the sharp market converges on, and it only exists if you track the books that produce it. An edge is a soft price that pays more than that fair line implies, and it only exists if you track the books that misprice. A platform that covers only soft books is guessing at fair value; a platform that covers only sharp books has nothing mispriced to bet. CLV.gg covers both classes for the same reason a scale needs two pans: the measurement is the gap between them, not either side alone.
What does CLV.gg track and grade?
CLV.gg detects +EV, arbitrage, middle, low-hold, and steam signals across all six sports and the exchanges, and grades every detection against the sharp closing line. The grading is the part that makes the coverage honest: a signal is not trusted because it was published, it is scored against where the market closed, and the record sits on the public sample edges page. The historical layer underneath is 1.31 billion-plus rows of odds across six leagues at 15-minute, 5-minute, and 1-minute grain, available to partners through the API for backtesting. The full method, from fair-line construction to CLV grading, is written up at /methodology.