The conventional wisdom says catcher framing is dead in the ABS era. If the batter can challenge any call, what's the point of making a ball look like a strike? The data tells a more interesting story.
We identified every borderline pitch from the 2025 season — pitches that landed within 1.5 inches of the zone edge — and grouped them by the catcher who received them. The sample: 74 catchers with at least 200 borderline pitches received.
The overall called-strike rate on borderline pitches across the league: 42.6%. That's the baseline. Now look at the range.
The Framing Gap
| Metric | Value |
|---|---|
| Catchers analyzed (200+ borderline pitches) | 74 |
| Overall borderline called-strike rate | 42.6% |
| Best framer | 51.7% (+9.1pp) |
| Worst framer | 33.0% (-9.7pp) |
| Total range | 18.7pp |
The best framer in the league earns a called strike on 51.7% of borderline pitches — nearly 10 percentage points above the league average. The worst framer gets just 33.0%. That's an 18.7 percentage point gap on the pitches that matter most in every game.
These aren't random fluctuations. At 200+ borderline pitches per catcher, the signal is robust. The best framers consistently present the ball in a way that makes the umpire's visual impression match the call they want.
The ABS Paradox: Why Framing May Gain Value
Here's where the analysis gets counterintuitive. The obvious prediction is that framing loses value in the ABS era — if the batter can challenge a wrong call and get it overturned, then tricking the umpire is pointless.
But that logic misses a critical step: the batter has to decide whether to challenge.
A challenge is not free. Teams have a limited number of challenges per game. Using one early means you might not have one when a 3-2 call goes wrong in the 8th inning. The decision to challenge is a judgment call made in real time, and it's influenced by how the pitch looked coming in.
This creates the strategic tension:
- Good framing makes the umpire's call look correct. If the catcher receives the ball smoothly and presents it in the zone, the pitch looks like a strike — both to the umpire and to the batter.
- Batters are less likely to challenge pitches that looked like strikes. Even if the pitch was technically outside the zone, a well-framed presentation creates doubt. Was it really outside? Is it worth burning a challenge?
- The challenge system doesn't eliminate framing value — it redirects it. Instead of directly influencing the call, framing now influences the challenge decision. The mechanism changed, but the advantage persists.
Umpire Accuracy by Pitch Speed
Framing effectiveness doesn't exist in a vacuum. The umpire's baseline accuracy varies by pitch velocity — faster pitches are harder to track, which creates more opportunities for framing to influence the call.
| Pitch speed | Umpire accuracy |
|---|---|
| Under 85 mph | 93.9% |
| 85 – 90 mph | 93.9% |
| 90 – 95 mph | 91.8% |
| 95 – 100 mph | 91.7% |
| 100+ mph | 91.9% |
Umpire accuracy drops about 2 percentage points when pitch velocity crosses 90 mph. Interestingly, accuracy doesn't continue declining above 95 mph — the umpires seem to have adapted to elite velocity, or those pitches tend to be more clearly in or out of the zone.
The implication for framing: the 90–95 mph range is where framing has the most room to operate. These pitches are fast enough to create uncertainty but common enough to represent a large share of borderline calls.
How Well Can Location Predict Called Strikes?
We also built a simple classification model to see how much of the called-strike decision is explained by pitch location alone. Using only the horizontal and vertical position of the pitch relative to the zone boundaries, the model achieves an AUC of 0.980.
That's nearly perfect — pitch location explains almost all of the variance in called-strike decisions. But the remaining 2% is where framing lives. On borderline pitches, that small residual is the difference between a called strike and a ball, and the data shows catchers vary systematically in their ability to capture that margin.
What This Means
Catcher framing is not dead. The 18.7 percentage point gap between the best and worst framers is real, robust, and likely to persist in the ABS era — because framing's mechanism has shifted from influencing the initial call to influencing the challenge decision.
For front offices evaluating catchers, framing remains a legitimate skill with measurable impact. For ABS challenge strategy, the interaction between framing quality and challenge likelihood is a new dimension worth modeling. CalledThird will track framing metrics alongside challenge data throughout the 2026 season.
Methodology
Data source: 2025 MLB Statcast data (729,827 pitches, 379,155 called pitches). Accessed via pybaseball.
Borderline definition: Pitches where the ball center was within 1.5 inches of the nearest zone edge (horizontal or vertical). This captures the zone where umpire accuracy is most variable.
Catcher filter: Minimum 200 borderline pitches received. 74 catchers met this threshold.
Called-strike rate: Number of called strikes divided by total called pitches (balls + called strikes) on borderline pitches, grouped by catcher.
Location AUC: Logistic regression predicting called strike vs. ball using only plate_x (horizontal position) and plate_z (vertical position) relative to zone boundaries. AUC = 0.980 on held-out test set.
Umpire accuracy by speed: Accuracy computed in 5 mph bins. Only called pitches included. Speed measured via Statcast release_speed.
If you find an error, tell us — we'd rather be corrected than wrong.