← Maine XC Index

Regional previews

Saturday 24 October · South at Twin Brook, Cumberland · North at Troy E Howard, Belfast

Graded before it was published. Run against the 2025 Southern Maine regionals — a championship none of this had seen, using only races run before it — this method called 91.0% of runner pairs in the right order and placed teams within 0.61 of a finishing position on average. Everything below is ten thousand simulated races, not one guess.

How these are built, and what they cannot know

The field is everyone rated in that class and region. A roster is not something this can know, so a runner who stopped racing in September is still in the field. A coach reading their own team will spot that faster than any rule could.

Improvement is credited by school year, not by runner. Underclass runners beat their index slightly; upperclass runners come in just under it. Fitting each runner's own trend and extrapolating it was tested against four alternatives and came last, so it is not used.

The course factor does not affect team predictions at all. Scoring depends on who beats whom, and a course factor moves everyone together. It is applied to the projected times, which is the only place it changes anything.

Two things were tried and rejected on the 2025 holdout: shrinking thin ratings toward the field (0.775 of a place against 0.630 — clearly worse), and ranking teams by median place rather than expected points.

A regional is one race. The spread on each line is the honest width of what could happen, not decoration. Where two teams' ranges overlap, the order between them is close to a coin toss.