Catcher framing is the skill of receiving pitches in a way that gets called strikes more often. Elite framers steal 15+ runs per season from opposing offenses. Framing is one of the most undervalued inputs in MLB betting and shows up directly in pitcher ERA.
Framing measures how often a catcher converts pitches IN the strike zone (called for strikes correctly) AND pitches just OUTSIDE the strike zone (stolen strikes that should have been balls). Statcast tracks every pitch's location and the umpire's call, then assigns each catcher a per-pitch framing run value.
Elite framers (Yasmani Grandal historically, JT Realmuto, Sean Murphy in modern era) post +10 to +20 framing runs over a full season. Poor framers post -10 to -20. The gap between best and worst framer is 30-40 runs, equivalent to 3-4 wins.
Framing directly impacts pitcher outcomes:
The market often evaluates pitcher stats without crediting the catcher. A starter who looks like 3.50 FIP with a Realmuto might be closer to 3.85 with a replacement catcher. Conversely, a 4.20 FIP starter with a poor framer might be closer to 3.90 with a good one.
Framing matters more for some pitcher types:
Framing dwarfs all of these. A +15 framing catcher saves more runs through framing alone than the best blocker and caught-stealer combined.
The model tracks each starting catcher's framing runs (rolling 30-game window) and applies a per-pitcher adjustment based on the catcher's framing skill. Elite framer behind the plate → small downward adjustment on starter's expected ERA tonight. Poor framer → upward adjustment. The brain has learned framing as one of the higher-weighted inputs in the current iteration (weight 1.43 in analytics/brain_weights.json), reflecting how much it actually predicts outcomes.
Theory is one thing — using the concept on tonight's slate is the value. These hubs are the practical application: