Guide • Probability reading • Filters

Practical guide: using a double threshold (probability + confidence)

Published on December 9, 2025 · Updated on December 22, 2025

Probability Confidence index By league Coverage Reliability
Double threshold: probability and confidence index in football prediction

Why this guide exists

A probability (for example, 60%) is easy to misread. Foresportia therefore relies on a double threshold: match probability combined with a confidence index based on league stability.

The goal is to help choose a clear compromise between coverage (number of matches) and reliability (statistical stability).

The double threshold, explained

1) Probability threshold

This is the probability assigned to a given outcome. Increasing the threshold reduces volume but generally improves stability; lowering it increases volume but also fluctuations.

2) Confidence index (league-weighted)

Not all leagues behave the same way. The confidence index reflects observed reliability and acts as a safeguard: a high raw probability in an unstable league should be read with caution.

Choosing your balance: three simple profiles

Explorer

Threshold around 55–58% with exclusion of low-confidence leagues. Useful to learn and explore without excessive filtering.

Selective

Threshold around 60–62% combined with high-confidence leagues. Often the best compromise for consistent analysis.

Highly filtered

65%+ with strong confidence only. Very low volume, which can lead to misleading short-term impressions.

Conclusion

The double threshold encourages a responsible reading of predictions: probability as a match-level signal, and confidence index as a league-level reliability safeguard. The right setup is the one that remains consistent with your objective over time.