Leicester City vs Doncaster Rovers

The model reads this fixture at 11.02 total corners, marginally above the League One baseline of 10.7. That's a thin gap. Both sides carry recent match averages well above that baseline, Leicester at 13.2 and Doncaster at 12.4, yet the model's output doesn't simply track those headline figures, because regression toward the league mean pulls the projected total down.

The explanation sits in the variance within each team's recent sample. Both teams produced an 8-corner match within their last five outings, yet both also registered outputs at the opposite end of the range, Leicester hitting 16, Doncaster hitting 18. That spread is the problem. The medium confidence grade reflects a data window where the floor and ceiling sit far apart, with no reliable resolution of which end this fixture occupies.

Home context adds a further layer the data doesn't resolve cleanly. The King Power carries crowd dynamics that can sustain attacking intent throughout, particularly when the home side controls possession and territory against visiting opposition. That ambiguity matters. How Leicester's structure shifts in front of a full home crowd isn't cleanly captured in a corner-count sample.

Discipline is proportionate to the grade. Spectral's medium-confidence record across 132 settled League One picks sits at 47.7%, a number that reflects how much room for error remains when variance inside a team's recent window is this wide.