Lorient vs Le Havre

The model's predicted total of 9.18 corners sits below the Ligue 1 baseline of 9.7, and that suppression warrants attention. The spread is the problem. Lorient's recent average of 8.5 per match pulls in one direction, Le Havre's 9.9 pulls in the other, and the gap between them is narrow enough that neither figure anchors the projection with real authority. When the two averages diverge and the difference is this small, neither tendency dominates the projected total cleanly.

What complicates the picture further is variance across recent matches. Le Havre's last five corner totals read 6, 19, 7, 7, 14, a distribution so wide that the 9.9 average carries limited predictive weight on its own. Lorient's sequence of 6, 16, 8, 11, 11 tells the same story: two matches spiked well clear of their mean, and three clustered low. High-variance inputs produce uncertain outputs. That's not a flaw in the data; it's the data being honest.

Medium confidence reflects exactly that uncertainty. The model ran on complete inputs, the predicted total of 9.18 is coherent, but the confidence tier is calibrated to what the data actually supports. Structurally high variance and a projected total sitting close to the baseline means small shifts in each team's approach can move the count meaningfully. Exposure should match the tier, not optimism about how this particular match plays out.

Spectral's medium-confidence Ligue 1 record stands at 47.8% across 23 settled picks.