Chapecoense vs Fluminense

The read here is cautious. Chapecoense's last five match corner totals, 13, 6, 10, 8, and 8, produce a 9.8 average that sits below the league baseline of 10.1 and signals meaningful game-to-game variance rather than a settled pattern. The predicted total of 10.27 clears the baseline, but only fractionally, and that thin margin is exactly why the model outputs medium rather than high confidence.

Fluminense offer partial support. Their last five away totals, 6, 12, 7, 11, and 14, average 10.3, which is fractionally above the 10.1 league baseline, but the distribution is wide enough to limit what that average actually tells us. A range spanning 6 to 14 corners across five matches is the central uncertainty in this assessment, and the model captures the dispersion rather than suppressing it.

Thin edge. The gap between the 10.1 baseline and the 10.27 predicted total is real but narrow, and small data shifts could close it entirely. On fixtures like this, where the signal is present but not dominant, the variance documented in both teams' recent data hasn't been smoothed away by the model, it's been priced into the confidence tier. Treating medium-confidence picks with the same weight as high-confidence tier picks is a structural error.

Spectral's medium-confidence picks in this league have settled 34 times at a win rate of 55.9%.