discovery v2: full feature factory
← lab · 590 pairs tested · 327 survive every era · ran 7/3/2026
What this is: Theory-free sweep: correlates every pre-match factor with every outcome, corrects for the thousands of tests, and keeps only effects stable in every era.
Theory-free sweep: every pre-match factor against every outcome. Only correlations that stay significant after multiple-testing correction AND point the same way in every era are shown. Strength guide: 0.02 tiny · 0.1 small · 0.2+ notable for sports.
| Factor | Predicts | Correlation | By era |
|---|
| match.elo_gap | goal_margin_home | +0.436 | + + + + |
| match.elo_gap | home_win | +0.356 | + + + + |
| match.elo_gap | away_win | -0.349 | − − − − |
| match.elo_gap | home_goals | +0.336 | + + + + |
| home__team.elo | goal_margin_home | +0.314 | + + + + |
| match.elo_gap | away_goals | -0.305 | − − − − |
| away__team.elo | goal_margin_home | -0.300 | − − − − |
| match.form_gap | goal_margin_home | +0.285 | + + + + |
| home__team.elo | home_goals | +0.280 | + + + + |
| away__team.elo | away_win | +0.268 | + + + + |
| home__team.dominance_l5 | goal_margin_home | +0.263 | + + + + |
| home__team.shot_diff_avg_l5 | goal_margin_home | +0.262 | + + + + |
| home__team.elo | home_win | +0.260 | + + + + |
| away__team.elo | away_goals | +0.253 | + + + + |
| home__team.season_ppg | goal_margin_home | +0.250 | + + + + |
| away__team.shot_diff_avg_l5 | goal_margin_home | -0.247 | − − − − |
| away__team.dominance_l5 | goal_margin_home | -0.247 | − − − − |
| away__team.elo | home_win | -0.241 | − − − − |
| match.style_clash_corners | goal_margin_home | +0.238 | + + + + |
| away__team.season_ppg | goal_margin_home | -0.232 | − − − − |
| match.form_gap | home_win | +0.231 | + + + + |
| home__team.goal_diff_avg_l5 | goal_margin_home | +0.228 | + + + + |
| home__team.dominance_l5 | home_goals | +0.227 | + + + + |
| match.form_gap | away_win | -0.226 | − − − − |
| home__team.shot_diff_avg_l5 | home_goals | +0.225 | + + + + |
| home__team.elo | away_win | -0.222 | − − − − |
| away__team.shots_for_avg_l5 | goal_margin_home | -0.221 | − − − − |
| home__team.shots_for_avg_l5 | goal_margin_home | +0.218 | + + + + |
| away__team.dominance_l5 | away_win | +0.218 | + + + + |
| away__team.shot_diff_avg_l5 | away_win | +0.216 | + + + + |
| away__team.shots_for_avg_l5 | away_goals | +0.216 | + + + + |
| match.form_gap | home_goals | +0.215 | + + + + |
| home__team.season_ppg | home_goals | +0.215 | + + + + |
| away__team.goal_diff_avg_l5 | goal_margin_home | -0.214 | − − − − |
| home__team.dominance_l5 | home_win | +0.213 | + + + + |
| home__team.shot_diff_avg_l5 | home_win | +0.212 | + + + + |
| home__team.venue_ppg_l5 | goal_margin_home | +0.212 | + + + + |
| home__team.shots_for_avg_l5 | home_goals | +0.211 | + + + + |
| home__team.form_points_l5 | goal_margin_home | +0.208 | + + + + |
| home__team.corner_diff_avg_l5 | goal_margin_home | +0.208 | + + + + |
Reading the columnswhat each number actually means
| Factor | a pre-match metric (leak-free: computed only from earlier games) |
| Predicts | a post-match outcome |
| Correlation | Pearson r; sports live in 0.02–0.4 territory |
| By era | sign of the effect in each historical era — must agree everywhere to be listed |
Spec · the reproducible recipe
{
"kind": "discovery",
"sport": "football"
}