Quiniela de la Semana

Each match's probabilities come from three independent datasets: the home team's last 10 home matches in their domestic league, the away team's last 10 away matches, and head-to-head results at the home venue across all seasons. We compute three estimates of P(1) / P(X) / P(2) and combine them via a sample-size-weighted average.

For ticket-level numbers (the 14/14 and ≥13/14 probabilities), we assume per-match outcomes are independent and compute the Poisson binomial distribution across all 14 (or 7) matches. The independence assumption is approximate — weather, schedule fatigue, and tournament dynamics introduce correlations our model doesn't capture.

Small samples can produce extreme probabilities. When a team has fewer than 5 total historical matches across all three datasets, we default to a roughly uniform 33/33/33 prediction and flag the match as "Limited data" or "No data". This avoids overconfident estimates from thin data. The optimizer naturally favors converting these matches to triples (covering all three outcomes) where the rules allow.

The model uses regular-season domestic league data only — no cups, friendlies, or playoff matches. Real-world hit rates over time may run lower than the model predicts.

Goletrics is not a gaming operator. We provide statistical analysis and probability estimates for football matches and quiniela packages. We do not accept wagers, hold funds, or facilitate plays. The recommendations on this page are informational; they are not guaranteed outcomes and do not constitute financial or gambling advice.

To play Progol or Progol Revancha, visit Pronósticos para la Asistencia Pública or any authorized agent. Pronósticos is operated by Lotería Nacional, a Mexican federal entity. Goletrics has no commercial affiliation with Pronósticos or Lotería Nacional.