Everton vs Sunderland

Variance dominates this fixture. Everton's last five corner match totals, 5, 14, 7, 12, 9, show the kind of swing that complicates any confident position, and Sunderland's recent figures, 5, 9, 9, 14, 11, are no tighter. Both team averages sit below the Premier League baseline of 10.5, Everton at 9.6 and Sunderland at 8.9, and the model's predicted total of 9.92 reflects that shared directional lean. Directional alignment across both teams doesn't collapse the dispersion risk embedded in the data.

What's driving the medium-confidence grade is the gap between model output and edge conviction. That gap matters. The model-level confidence is high, but the edge assessment lands at medium, and that separation reflects the dispersion visible in both teams' recent data. When a five-game sample contains both 5 and 14 for the same side, the average is a useful reference point but a limited guide to any individual match.

There's no signal in either team's recent profile that corners are converging to a tighter band. Direction is clear. The data-driven picture has both teams averaging below the league baseline of 10.5 and the predicted total of 9.92 sitting below it too, but coherence in direction and confidence in outcome are different things. Structural variance at this level is the primary reason the confidence grade sits where it does.

Stake accordingly. Medium-confidence plays require a smaller footprint than high-confidence ones, and the dispersion profile here reinforces that discipline rather than overriding it. Premier League medium-confidence record: 101 settled, 46.5% win rate.