EvertonvsSunderland

The model sets a predicted corner total of 9.92 for this fixture, running below the Premier League baseline of 10.5. Everton's home average of 9.8 and Sunderland's away average of 9.2 both sit under baseline and point consistently in the same direction, but the gap between those team averages and the league figure isn't wide enough to push output into high-confidence territory. Medium confidence it is. That designation affects how this fixture should be approached.

Recent match totals complicate the picture. Noise is real. Everton's last five home corner tallies ran 14, 7, 12, 9, 11: a spread that makes the 9.8 average feel more like a gravitational centre than a reliable forecast anchor. Sunderland's road figures read 9, 9, 14, 11, 10 across the same window. That 14 in each dataset is doing substantial work in pulling the averages toward the league baseline, and either side reproducing a high-corner-count performance would shift the predicted total into contested territory.

Medium confidence means the directional case is supportable but the variance is genuinely present. The model isn't misfiring; it's pricing in real uncertainty, and that distinction matters. Size this accordingly. Treating this fixture with the same conviction you'd extend to a high-confidence pick is a sizing error, full stop. The range of outcomes across these recent games is wider than the averages suggest.

In the Premier League, Spectral's medium-confidence corners picks have settled 80 times at a 46.2% win rate.