sure-thing survey
β lab Β· Β· ran 7/17/2026
What this is: A one-off study β its numbers below are exactly what the run reported.
cells
| n | band | wins | market | roi pct | avg price | selection |
|---|
| 1,218 | 70-80% | 901 | Match result | -0.81 | 1.343 | away side win |
| 2,982 | 70-80% | 2,230 | Match result | 0.42 | 1.344 | home side win |
| 479 | 80-90% | 402 | Match result | 0.92 | 1.202 | away side win |
| 2,081 | 80-90% | 1,739 | Match result | -0.93 | 1.187 | home side win |
| 19 | p>=90% | 16 | Match result | -8.41 | 1.088 | away side win |
| 447 | p>=90% | 405 | Match result | -1.65 | 1.086 | home side win |
| 1,557 | 70-80% | 1,107 | Over/Under 2.5 goals | -3.7 | 1.357 | Over |
| 398 | 70-80% | 266 | Over/Under 2.5 goals | -7.6 | 1.384 | Under |
| 222 | 80-90% | 186 | Over/Under 2.5 goals | 0.71 | 1.203 | Over |
| 2 | 80-90% | 1 | Over/Under 2.5 goals | -37.5 | 1.25 | Under |
| 3 | p>=90% | 3 | Over/Under 2.5 goals | 10.5 | 1.105 | Over |
Reading the columnswhat each number actually means
| AUC | predictability: 0.50 = coin flip, ~0.70 = ceiling for sports |
| Importance | how much the model leans on this factor (permutation importance) |
| Direction | sign of the raw correlation with the outcome |
| Survives all eras | effect points the same way in every historical era |
Spec Β· the reproducible recipe
{
"markets": [
"1x2",
"ou_goals_2_5",
"btts"
],
"threshold": ">=70% implied"
}