Ranking and Learning

Fuzzy ranking

Result ranking blends:

• ordered match quality
• contiguous run quality
• boundary boosts
• early-match preference

Interaction learning

Hop Launcher learns from your interactions:

• apps and windows you select more often receive boosts
• frequent selections for similar fuzzy queries rise higher over time
• behavior stays query-sensitive, not just globally popular

This keeps repeated workflows fast and stable without losing useful fuzzy matching.