A pitch crosses the plate at the knees, clips the edge of the strike zone, and freezes the hitter. Whether the umpire’s arm goes up or stays down is, on its face, the umpire’s call. But watch the catcher’s glove on the replay — the way it doesn’t jerk, the way it settles a hair toward the middle, the way the whole receive looks like the pitch was always going to be a strike. That quiet bit of theater has a name. It’s called pitch framing, and for a decade it was one of the most valuable skills in baseball that almost nobody outside front offices could see.
Framing is the catcher’s art of coaxing borderline pitches into the strike column — and, just as importantly, of not losing strikes that were already there. It sounds like a rounding error. It is not. Spread across the thousands of pitches a catcher receives in a season, a fraction of a strike per game compounds into one of the larger defensive contributions in the sport. Here’s how the skill works, how the metrics put a number on it, and why the looming robot umpire may quietly erase it.
Why one borderline call matters so much
Start with the unit of currency: the difference between a called strike and a called ball. It looks trivial, a single tick in the count. In run terms it is anything but. A pitch that goes from ball to strike doesn’t just add a strike; it moves the entire at-bat to a more pitcher-friendly count, and counts drive outcomes. A hitter in a 1–1 count is a fundamentally different threat than the same hitter in 0–2, and a 3–1 hitter is a different animal from a 2–2 one. The expected run value of a plate appearance swings hard with the count, which is exactly the kind of state-by-state accounting that powers win probability models and run-expectancy tables.
Analysts have generally pegged the average value of flipping one ball into one strike at somewhere in the neighborhood of a tenth of a run — small, but real. Now do the multiplication. A regular catcher receives well over a hundred pitches a game, and a meaningful slice land in the ambiguous shadow zone at the edges where the call could go either way. Win a handful of those per game across a full season and the runs pile into a total that rivals what a good defender saves with his glove in the field. Framing was hiding in plain sight precisely because each instance was too small to notice.
What a good framer actually does
The craft is mostly about stillness and timing. The best receivers are described, almost universally, as having “quiet hands” — the glove arrives at the ball with minimal travel and no last-instant lunge. A catcher who stabs at a low pitch and yanks the glove upward is broadcasting doubt; an umpire reads that body language and rings up a ball. A catcher who lets the pitch settle, catches it on the inner half of the glove, and holds it a beat presents the same pitch as a strike.
A few sub-skills recur. Receiving the low strike is the marquee one: the bottom edge of the zone is the hardest border for umpires to judge, and a catcher who can stick a pitch at the knees without dropping his glove harvests the most strikes. Working the corners matters too — catching an outside pitch so the glove finishes over the plate rather than drifting off it — as does a quiet, balanced setup that doesn’t require violent movement to reach the target. The one-knee-down stance that swept the sport in recent years was adopted largely because it helps receivers stabilize the bottom of the zone. None of this is deception in any sinister sense; it’s presentation, the same way a confident answer earns more benefit of the doubt than a hesitant one.
How framing gets measured
You cannot eyeball framing across a season — the signal is buried under too much noise — so the metric is built from pitch-tracking data, in three steps.
First, build a model of the strike zone. Using the recorded location of every pitch and how it was actually called, you estimate the probability that a pitch in a given spot — conditioned on the count, the batter’s handedness, and the pitch’s height and width — gets called a strike. A pitch down the pipe sits near 100 percent; a pitch a foot off the plate near zero. The interesting pitches live in the shadow zone, where the probability might be 40 or 60 percent. Second, compare the catcher to that expectation pitch by pitch: if a pitch had a 45 percent chance of a strike call and got one, the catcher is credited with the strikes he gained above expected; if a likely strike is called a ball, that’s a debit. Summed across every borderline pitch, that yields strikes gained or lost versus a league-average receiver.
Third, convert strikes to runs using that per-call run value and you arrive at framing runs. In Statcast’s public terminology this lives under Catcher Framing, expressed as a strike rate on shadow-zone pitches and as runs above or below average. The throughline with our other Statcast work is exact: just as expected stats ask what an outcome should have been given the inputs, framing asks how many strikes a catcher should have gotten given where the pitches landed — and credits the gap.
An illustrative example makes the scale tangible. Suppose a catcher is, by the model, worth a quarter of a strike more per game than an average receiver — roughly one extra borderline pitch turned into a strike every four games. Over a full season that drips into tens of extra strikes, and at a tenth of a run apiece those add up to a double-digit run contribution. Numbers in that range have, in the better-documented seasons, been enough to make an otherwise modest-hitting catcher a genuinely valuable everyday player — and a poor framer a quiet liability no matter what he does at the plate. (Treat the figures as illustrative of the magnitude, not any specific player’s line.)
The caveats: it’s a three-body problem
Framing runs are a real measurement, but they measure a relationship, not a solo act. At least three parties shape every called strike. The pitcher matters: an arm with pinpoint command who lets the catcher set a target and hit it is far easier to frame than a wild one whose catcher is constantly reaching, so some of a framing line is really the staff’s command. The umpire matters: umpires have individual zones and susceptibilities, and a catcher’s numbers partly reflect which ones he drew. And the count matters enormously — umpires call a visibly wider zone on 3–0 and a tighter one on 0–2, which is exactly why the model has to condition on it. Good framing metrics adjust for these, but adjustment is not elimination. Read a single season of framing runs as a strong estimate with real error bars, not as gospel.
The robot umpire is coming for it
Here is the existential threat. If a machine calls balls and strikes, framing is worth exactly nothing — a pitch is a strike or it isn’t, and no glovework changes the verdict. The Automated Ball-Strike system, the “robot umpire,” has been tested for years in the minor leagues in two flavors: a full ABS mode where the computer calls every pitch, and an ABS challenge system where human umpires call the game but teams can challenge a handful of calls to the machine. The challenge version has been moving toward the major-league stage, tested publicly in spring settings.
The two flavors threaten framing differently. Full ABS would zero it out completely. A challenge system is gentler — most pitches are still called by a human, so framing on the un-challenged majority still matters, but the most egregiously stolen strikes become recoverable. Either way the trend points one direction. The skill that front offices spent a decade learning to value and pay for is, by design, the skill an automated zone makes obsolete. Framing may end up the perfect case study in a metric that measured something real right up until the rules legislated it away.
The bottom line
Pitch framing is the rare baseball skill that was invisible to the naked eye, enormous in aggregate, and entirely dependent on tracking data to prove it existed at all. A great receiver steals a fraction of a strike per game, and a fraction of a strike per game — compounded over a season and converted to runs — is worth real wins. The metric is honest about being a relationship rather than a solo feat, tangled up with the pitcher, the umpire, and the count. And it may be living on borrowed time: the moment a machine calls the zone, the catcher’s quiet hands stop mattering. For now, when you watch a borderline pitch get rung up, look at the glove. Somebody worked for that strike.
Sources & Further Reading
- Baseball Savant — Statcast Catcher Framing leaderboards and the shadow-zone strike-rate methodology.
- FanGraphs Library — explainers on framing runs and catcher defense.
- MLB.com — coverage of the Automated Ball-Strike system and the ABS challenge experiments.