IFK Göteborg vs Malmö FF

The model's predicted total of 10.45 sits marginally below the SE league baseline of 10.6. That gap is tight. When predicted total and league baseline converge this closely, medium confidence is an honest assessment, not a placeholder.

IFK Göteborg's last five corner totals, 9, 10, 8, 17, 16, are the central problem. The final two readings drag their home match total average to 12.0, yet the three before them sit well under the baseline. Malmö FF's sequence is different: 9, 11, 10, 10, 11, a narrow band that holds a match total average of 11.3 without material outliers. Both averages sit above the predicted total of 10.45. Two teams. Two entirely different distributional shapes feeding into the same output.

That contrast is what compresses confidence. The home side's match total average of 12.0 is pulled by those two spikes, and there's no clean way to dismiss them as noise given how recent they are. The model has to reconcile this with a visitor whose readings cluster in a narrow band and whose match total average of 11.3 is similarly above the predicted output. Neither profile is illegible. Both are genuinely difficult to combine into a single high-conviction output at a predicted total that sits below each team's own average.

Size accordingly. Medium-confidence picks carry more output variance than the high-confidence tier, and this fixture illustrates the structural reason why. In SE, Spectral's medium-confidence model record stands at 73.3% across 15 settled picks.