Chilean Primera Division

Corners prediction uses a Poisson regression trained separately for Chilean Primera Division. Every league has its own tempo, refereeing style, and variance profile, so a model calibrated on one doesn't transfer cleanly to another — we retrain per-league and recalibrate as new matches settle.

Once the model produces a probability for each corners line, we compare it against live bookmaker odds. Where the implied probability is meaningfully lower than ours (the edge), we flag a value bet. Pick sizing follows a Kelly fraction scaled by our confidence tier, capped for bankroll safety.

The full write-up — features, calibration method, overdispersion handling, and league-specific bias correction — lives on the methodology page. Historical accuracy for every league Spectral covers is tracked publicly on /track-record.