Bologna vs Inter Milan

Thin data on both sides. Bologna's home corner average of 12.7, drawn from three matches, is shaped significantly by a 17-corner game that carries outsized weight in the mean. Inter's two-game away sample produced 6 and 14, a range that tells you more about sample instability than about any stable away corner tendency.

The model's predicted total of 9.53 sits fractionally below the Serie A league baseline of 9.9. Both samples are thin. That gap, combined with the divergence between Bologna's home average of 12.7 and Inter's away average of 10.0, means the data doesn't cleanly resolve which tendency is more predictive at Stadio Renato Dall'Ara. No dominant signal.

Small samples cut both ways. Bologna's last three home totals of 7, 17, and 14 show genuine variance, and Inter's two-match away sample doesn't compress it. When the predicted total lands this close to the baseline and the underlying data is this dispersed, the medium-confidence rating is an accurate read of the model's certainty, not a conservative hedge.

Stake discipline matters. Medium confidence is not a discounted version of high confidence; it's a structurally different tier where the model's certainty is materially lower, and treating the two tiers identically ignores the information that the confidence rating is designed to convey. The league record at medium confidence is the relevant benchmark.

Serie A medium-confidence record: 58 settled, 44.8% win rate.