Where this data comes from
Schedule/score data from ESPN's public scoreboard feed (start times shown in your local timezone). Covers NFL, NBA, MLB, NHL, MMA (UFC), Tennis (ATP/WTA), Golf, Soccer, and college ball. Expand an in-season NBA/WNBA/NFL/MLB/NHL game (▸) to load its live Kalshi player props. For anything else — or to price a line no book is offering — use the Line Odds tab.
Line Odds Model
Type any player, pick a stat, enter a line (e.g. Vinícius Júnior · Shots On Goal · over 1.5). The model pulls their real recent game-by-game numbers from ESPN and estimates the fair probability and fair (vig-free) odds — works for lines no sportsbook is even offering. Paste a market price to see your edge.
Price a screenshot
Upload a screenshot of your slip and it reads the players, lines and over/under straight off the image, then prices every pick.
| Player | Pick | Odds / EV | L10 | Edge |
|---|
Click a player to expand the full picture — L5/L10/L15/season hit rates, streak, recent average, model vs implied probability, the market's vig and no-vig price, the last 12 games, and where the line came from.
What edge and EV mean here
Edge = model probability minus the book's implied probability; positive means the model likes it more than the price does. EV is what that edge is worth: expected profit per $1 staked at the posted price. Hit rates are history, not forecasts — they don't adjust for opponent, minutes, or injuries.
Two things EV deliberately won't tell you. PrizePicks quotes no per-pick price
— their -110 is a placeholder and the payout comes from the entry
multiplier — so their EV reads n/a here and only appears on a
Pre-Built Slip. And a price outside ±400 shows its EV
struck through: those quotes reach +19 per
dollar, but they're order books with nobody in them, not edges.
Top picks single lines, best first
Pre-Built Slips same book, sports mixed
How these are picked
Slips are grouped by book, not by sport — every pick'em app lets you combine an MLB leg with a WNBA one, so where you place it is the only thing that actually constrains a slip. Legs use distinct players, need 8+ games of history, and must beat the price on their season record as well as recent form.
The headline percentage is the conservative edge: the lesser of what recent form says and what the season record says. Ranking on the model alone is adverse selection — it surfaces the legs where the model is most extreme, and extreme and wrong are largely the same set.
PrizePicks pays by entry size and publishes no payout in their API, so EV there uses the standard Power Play multiplier — override it in Filters if your app differs. Underdog and Sleeper quote a real price per pick, so their EV comes straight from those. Kalshi is left out: it's an exchange with a thin order book, not a slip product.
Your Slip
| Player | Stat | Line | Dir | Odds | Model % |
|---|
Boost
Applying a sportsbook promo recomputes your slip's EV with it. A profit boost multiplies winnings; an odds boost replaces the price; a bonus bet refunds the stake on a loss.