edge hunt: one +EV leg at market price, train/test validated
β lab Β· n survivors 2 Β· segments scanned 217 Β· ran 7/29/2026
What this is: A one-off study β its numbers below are exactly what the run reported.
| note | survivor = +EV in train AND test at opening prices; gold = also +EV at closing (beats the sharpest line). Expect a few false positives from multiple testing even so. |
| n survivors | 2 |
| segments scanned | 217 |
gold
| sel | band | comp | market | test n | train n | test roi |
|---|
| away side win | priced 1.00β1.80 | PPL | Match result | 211 | 672 | 6.56 |
| away side win | even | SA | Match result | 319 | 1,414 | 5.61 |
survivors
| sel | band | comp | market | test n | train n | test roi |
|---|
| away side win | priced 1.00β1.80 | PPL | Match result | 211 | 672 | 6.56 |
| away side win | even | SA | Match result | 319 | 1,414 | 5.61 |
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
{
"kind": "edge_hunt",
"bands": [
"odds_on",
"even",
"mid",
"long",
"xlong"
],
"cutoff": "2022-07-01",
"min_test": 100,
"min_train": 200,
"price_agg": "max",
"test_min_roi": 1,
"train_min_roi": 3
}