Stats
Every graded play the board listed, sliced however you like. Beating the closing line (CLV) is the signal; win-loss and ROI swing wildly on small samples, so slices under 30 plays are greyed out.
By sport and market
+EV picks (2–15%) matching the filters above.
| Plays | Avg CLV | Beat close | Record | Units | ROI (95% range) | |
|---|---|---|---|---|---|---|
| 89 | +0.3% | 63% | 53-36 · 60% | +17.85u | +20.1% ± 21% | |
| 74 | — | 100% of 3 | 38-36 · 51% | +1.65u | +2.2% ± 23% | |
| 51 | -0.1% | 52% of 23 | 39-12 · 76% | +9.65u | +18.9% ± 19% | |
| 64 | -0.3% | 45% | 31-33 · 48% | -2.11u | -3.3% ± 25% | |
| 74 | -0.1% | 56% | 29-45 · 39% | -16.25u | -22.0% ± 22% | |
| 95 | -0.4% | 45% | 64-31 · 67% | +2.91u | +3.1% ± 15% | |
| 210 | 0.0% | 47% | 113-97 · 54% | +15.52u | +7.4% ± 14% | |
| 224 | +0.4% | 61% | 105-119 · 47% | -13.87u | -6.2% ± 13% | |
| 30 | +0.2% | 56% of 27 | 15-15 · 50% | -5.83u | -19.4% ± 30% | |
| 34 | -0.1% | 38% | 9-25 · 26% | -16.45u | -48.4% ± 29% | |
| 50 | +0.2% | 41% | 21-29 · 42% | -8.13u | -16.3% ± 28% | |
| 83 | -0.2% | 48% | 46-37 · 55% | +0.69u | +0.8% ± 20% | |
| 23small | -0.5% | 35% of 20 | 9-14 · 39% | -6.75u | -29.3% ± 37% | |
| 61 | -0.1% | 42% | 25-36 · 41% | -9.78u | -16.0% ± 26% |
MLB · Moneyline+0.3% · 53-36
MLB · Totals— · 38-36
NFL · Moneyline-0.1% · 39-12
NFL · Spread-0.3% · 31-33
NFL · Totals-0.1% · 29-45
NCAAF · Moneyline-0.4% · 64-31
NCAAF · Spread0.0% · 113-97
NCAAF · Totals+0.4% · 105-119
NBA · Moneyline+0.2% · 15-15
NBA · Spread-0.1% · 9-25
NBA · Totals+0.2% · 21-29
NHL · Moneyline-0.2% · 46-37
NHL · Puck line-0.5% · 9-14
NHL · Totals-0.1% · 25-36
Does more EV mean more edge?
The same filters, split by EV band (the band filter is ignored here). If thin edges beat the close as often as +EV picks, the 2% floor isn't filtering anything; over-cap plays are hidden from the picks as likely bad lines.
| Plays | Avg CLV | Beat close | Record | Units | ROI (95% range) | |
|---|---|---|---|---|---|---|
| Thin edges (0–2%) | 459 | +0.1% | 55% | 258-201 · 56% | +4.50u | +1.0% ± 8% |
| +EV picks (2–15%) | 1162 | 0.0% | 50% | 597-565 · 51% | -30.90u | -2.7% ± 6% |
| Over cap (15%+) | 117 | +0.7% | 65% | 52-65 · 44% | -10.81u | -9.2% ± 19% |
Thin edges (0–2%)+0.1% · 258-201
+EV picks (2–15%)0.0% · 597-565
Over cap (15%+)+0.7% · 52-65
Where our prices came from
The source with the best price for each pick when it was made. Unmatched: the price moved between the pick and the game's last pre-game pull, so we don't guess.
| Plays | Avg CLV | Beat close | Record | Units | ROI (95% range) | |
|---|---|---|---|---|---|---|
| Kalshi | 463 | +0.2% | 50% | 253-210 · 55% | +19.08u | +4.1% ± 9% |
| NoVig | 367 | -0.1% | 49% | 186-181 · 51% | -9.96u | -2.7% ± 10% |
| ProphetX | 189 | -0.3% | 52% | 89-100 · 47% | -25.05u | -13.3% ± 13% |
| Pinnacle | 23small | +0.3% | 45% of 20 | 9-14 · 39% | -5.52u | -24.0% ± 40% |
| ESPN BET | 17small | +0.5% | 70% of 10 | 13-4 · 76% | +8.53u | +50.2% ± 41% |
| DraftKings | 15small | -0.2% | 54% of 13 | 7-8 · 47% | -2.05u | -13.7% ± 49% |
| Caesars | 7small | +0.9% | 57% of 7 | 2-5 · 29% | -2.98u | -42.5% |
| FanDuel | 6small | -1.3% | 0% of 3 | 4-2 · 67% | +1.22u | +20.3% |
| BetMGM | 6small | -1.0% | 17% of 6 | 3-3 · 50% | -0.35u | -5.9% |
| BetRivers | 4small | -0.6% | 33% of 3 | 2-2 · 50% | -1.30u | -32.6% |
| Unmatched | 65 | +0.1% | 51% | 29-36 · 45% | -12.50u | -19.2% ± 23% |
Kalshi+0.2% · 253-210
NoVig-0.1% · 186-181
ProphetX-0.3% · 89-100
Pinnacle+0.3% · 9-14
ESPN BET+0.5% · 13-4
DraftKings-0.2% · 7-8
Caesars+0.9% · 2-5
FanDuel-1.3% · 4-2
BetMGM-1.0% · 3-3
BetRivers-0.6% · 2-2
Unmatched+0.1% · 29-36
Calibration
When the model says 60%, does it win about 60%? Dots on the dashed line are honest probabilities. Hollow dots: under 20 plays.
CLV over time
Is the edge holding up or fading? Above zero = beating the close.
MLB totals CLV is measured in runs, so it's left out of the averages.
Sportsbooks
Moneyline prices at our daily pull, across every game we priced — not just our picks. Only the sport and date filters apply here. Shopping every source beat the average price by 1.52% in win probability across 2,972 sides.
| Source | Best price | Avg gap to best | Sides quoted |
|---|---|---|---|
| NoVig | 46% | 0.64% | 2,646 · 89% |
| ProphetX | 46% | 0.68% | 2,916 · 98% |
| Kalshi | 40% | 0.41% | 2,780 · 94% |
| BetRivers | 4% | 2.39% | 1,798 · 60% |
| FanDuel | 2% | 1.92% | 2,932 · 99% |
| ESPN BET | 2% | 2.05% | 1,750 · 59% |
| DraftKings | 2% | 1.96% | 2,904 · 98% |
| Pinnacle | 2% | 1.35% | 1,706 · 57% |
| BetMGM | 1% | 2.09% | 2,836 · 95% |
| Caesars | 0% | 2.06% | 1,646 · 55% |
NoVig46%
ProphetX46%
Kalshi40%
BetRivers4%
FanDuel2%
ESPN BET2%
DraftKings2%
Pinnacle2%
BetMGM1%
Caesars0%
Best price: share of the sides it quoted where it had (or tied) the best price. Avg gap: how much worse its price was than the best, in win probability. Exchanges and Pinnacle don't limit winning bettors; retail books often do.