Burnley vs Aston Villa

The model predicts 10.04 corners for this fixture, against a Premier League baseline of 10.5. Burnley's home match total average sits at 9.4, which already trails the baseline, yet their last five home games produced totals of 10, 14, 10, 13, and 12, a sequence that sits consistently above both that average and the model's projection. That tension matters. It doesn't invalidate the read, but it's one clear reason this sits outside the high-confidence tier.

Villa's away numbers are where the uncertainty compounds. Their last five away totals, 10, 7, 10, 12, and 5, average to 9.7, but the range across those games is the operative signal, not the average: a match total of 5 and another of 12 in the same five-game window indicate Villa's away corner profile shifts materially depending on game state. The output reflects that instability as medium confidence rather than anything higher. Dispersion is the issue.

Stake accordingly. The gap between 10.04 and the 10.5 league baseline is narrow enough that any in-game factor shifting the corner production rate can flip the read, which is precisely why medium-confidence outputs call for proportionate exposure. This market isn't forgiving of overconfidence. Resizing to match the confidence tier isn't overcautious, it's correct calibration for the data on the table.

Across 83 settled medium-confidence Premier League corners picks, Spectral's win rate stands at 45.8%.