Manchester UnitedvsNottingham Forest

This fixture warrants caution. Manchester United's last five corner totals span from 5 to 15, a range that resists any clean directional read, and Forest's figures over the same window run from 7 to 16. When the distribution is that wide, a single match can look like an entirely different team's output. That unpredictability isn't noise around a stable trend; it is the trend, and it shapes the confidence assessment before anything else.

The model projects 11.2 corners against a league baseline of 10.5. The gap is real. Manchester United's home match average is 9.6 and Forest's away average is 10.0, both below the baseline and short of the projection. Getting to 11.2 requires both sides to clear their recent averages on the same night, in a fixture where neither team has shown consistent output over the last five. The projection is defensible; the interval around it is wide.

Variance is the primary driver of the medium-confidence rating. A single defensive spell can compress the corner count well below any central estimate, and nothing in the last-five records establishes a floor for either team's output. That's an honest accounting of what the data contains, not a hedge. Stake discipline applies here. Medium confidence doesn't warrant an extended position, and the recent range in both sides' outputs is reminder enough.

In the Premier League at medium confidence, Spectral has settled 77 picks at a 45.5% win rate.