sweep: match.total_goals >= 2
← lab · AUC 0.532 (weak signal) · ran 7/4/2026
What this is: Asks which pre-match factors drive one specific outcome, using a walk-forward model and permutation importance.
| Factor | Importance | Direction | Survives all eras? |
|---|
| home__team.elo | 0.0001 | ↑ +0.052 | ⏳ yes |
| away__t3__team.dominance_l5 | 0.0001 | ↑ +0.008 | no |
| away__t3__team.venue_ppg_l5 | 0.0001 | ↑ +0.003 | no |
| away__team.dominance_l5 | 0.0001 | ↑ +0.004 | no |
| ratio__team.shot_diff_avg_l5 | 0.0001 | ↑ +0.026 | no |
| match.elo_gap | 0.0001 | ↑ +0.020 | no |
| market.p_draw | 0.0000 | ↓ -0.083 | ⏳ yes |
| away__t3__team.season_ppg | 0.0000 | ↑ +0.001 | no |
| match.state_gap | 0.0000 | ↑ +0.015 | no |
| ratio__team.venue_ppg_l5 | 0.0000 | ↑ +0.004 | no |
| away__team.defensive_leak_l5 | 0.0000 | ↓ -0.002 | no |
| away__team.elo | 0.0000 | ↑ +0.025 | no |
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 · the reproducible recipe
{
"name": "sweep: match.total_goals >= 2",
"sport": "football",
"target": {
"op": ">=",
"value": 2,
"metric": "match.total_goals"
},
"features": "all"
}