Catcher framing explained, and what happens to it now

The best play a catcher makes all night is one where nothing happens. A fastball arrives just off the outside edge, the glove is already there, it closes, and it does not move afterwards. The umpire's arm goes up. The hitter walks back to the dugout believing he was robbed, and he is about half right.

Catcher framing explained in one sentence is this: it is the skill of not moving, performed under a foul tip, in a crouch, several thousand times a season, in a way that alters an official's perception of a boundary he cannot see.

For most of baseball's history it was not a skill at all, in the sense that nobody could tell you who was good at it. Scouts talked about soft hands, and the phrase meant something, but it meant it vaguely. Then pitch tracking arrived, the vague thing became a measured thing, and within a few years it was being described as the largest single component of a catcher's defensive value. Now the game has installed a machine behind the umpire, and every honest account of framing has to reckon with what that does to it.

What framing actually is, and what it is not

Start with the misconception, because it is nearly universal and it is the opposite of the skill.

Framing is not pulling the ball into the zone. A catcher who receives a pitch six inches outside and yanks his glove towards the plate has done two harmful things. He has shown the umpire, with a large and obvious movement, exactly how far off target the pitch was. And he has made himself look like a man trying to get away with something, which is the fastest available way to lose the benefit of the doubt on the next twenty borderline pitches. Umpires notice. They are not required to be neutral about being played.

Framing is the management of everything that happens between the ball leaving the pitcher's hand and the umpire committing to a call, with the goal of making the moment of decision as favourable as it legitimately can be. That breaks down into a handful of separable mechanical tasks.

Getting there early. The glove should arrive in the neighbourhood of the pitch before the ball does, so that the reception is a small adjustment rather than a lunge. Late hands mean a large movement at the critical instant, and a large movement is the thing the umpire sees.

Catching the outer half of the ball. On a pitch at the outside edge, the catcher wants the ball to enter the glove on its far side, so that the glove finishes closer to the plate than the ball was. On a low pitch he wants the thumb underneath, so the glove finishes above the ball rather than below it. This is the closest thing to a trick in the whole discipline, and it is entirely legal: the glove is not being moved, it is being oriented before contact.

Absorbing without recoiling. A ball at high speed pushes the glove backwards and outwards on impact. A catcher with genuinely soft hands takes that energy into the arm rather than letting it travel through the glove, so the glove finishes where the ball was rather than three inches behind and below it. Nearly all of what old scouts meant by soft hands is this.

Sticking it. Hold the reception still for a beat. Not theatrically, and not so long that it becomes a statement. Long enough that the last visual information the umpire receives before deciding is a stationary glove in a plausible location.

Staying out of the way. The umpire is working in the slot, over the catcher's inside shoulder, and his sightline to the plate runs past the catcher's head and body. A receiver who sits up too tall or drifts across blocks the view of the very pitch he is trying to sell.

None of that involves deception in any sense the rulebook recognises. The ball went where it went. The catcher is simply declining to supply the umpire with a large moving object suggesting otherwise.

• 17Width of home plate, inches
• 2.9Diameter of a baseball, inches
• 5.9Total depth of the shadow band, inches
• 2Challenges a club starts a game with

The shadow zone is defined as running from one ball's width inside the rulebook edge to one ball's width outside it. Ball diameter is derived from the rulebook specification of 9¼ inches maximum circumference.

The band where the job exists

Framing does not operate on every pitch. It operates on a defined ring around the edge of the strike zone, and everything outside that ring is beyond reach.

A pitch down the middle is a strike whatever the catcher does with it, short of dropping it. A pitch a foot outside is a ball no matter how still the glove is, because the umpire is not confused about a foot. The entire discipline lives in the band where the call genuinely could go either way, and that band has a name and a definition: the shadow zone, running from one ball's width inside the rulebook boundary to one ball's width outside it.

Put numbers on that. The rulebook fixes the ball at no more than 9¼ inches in circumference, which is a diameter of a shade under 2.95 inches. So the shadow band is about 2.9 inches deep on each side of the boundary, and just under six inches deep in total, wrapped around the whole perimeter of the zone.

Horizontally that means a corridor running from roughly 11 inches wide at its inner boundary to roughly 23 inches wide at its outer one, centred on a 17-inch plate.

Inner boundary of the shadow band11.2in

Home plate17in

Outer boundary of the shadow band22.8in

Derived from the plate width in Rule 2.02 and the shadow zone definition, using a ball diameter of just under 2.95 inches. Pitches inside the narrow figure are strikes anyway; pitches outside the wide one are gone.

Show the numbers

The corridor a catcher is working in, side to side

Item	Value
Inner boundary of the shadow band	11.2in
Home plate	17in
Outer boundary of the shadow band	22.8in

Read those three bars against each other and the size of the opportunity becomes concrete. A catcher's influence extends over a corridor almost six inches wider than the plate and, at the same time, almost six inches narrower. He is not stealing strikes from wherever he likes. He is contesting a defined ring, and the ring is about the width of two baseballs.

That is also, incidentally, why framing and command are inseparable. A pitcher who cannot put the ball inside the shadow band gives his catcher nothing to work with, and the modern approach to designing pitches that finish on the edges exists partly to feed the receiver material. The strikes get credited to the catcher by the metric. Some meaningful share of them were earned by the man who threw the pitch.

Why the trick works at all

The mechanism is perceptual, and understanding it removes most of the mystique.

The rulebook asks whether any part of the ball passed through any part of a solid whose footprint is the whole five-sided plate, with vertical limits read off the batter's stance. That is the question. The umpire, however, cannot answer it directly, because he is not watching an object cross an invisible boundary in isolation. He is watching a sequence: a ball approaching, a glove moving, an impact, a resting position. His judgement forms across that whole sequence, and the sequence ends with the glove.

That is not a failure of professionalism. It is how visual judgement of a fast event works. When the definitive moment is too brief to resolve cleanly, the brain fills in from surrounding information, and the most salient surrounding information available is a large brown object that has just stopped somewhere.

The other half of the mechanism is timing. The call is made after the reception, always, because there is no way to speak before the ball arrives. Any information that lands in the gap between crossing and calling is inside the decision window. A catcher who moves in that window is adding evidence. A catcher who is perfectly still is adding nothing, which is exactly what he wants: with no new evidence, the umpire falls back on his read of the flight, and the read of the flight is a coin toss inside the shadow band.

There is a structural reason this is unusually easy to exploit at the top and bottom of the zone. The horizontal boundaries of the strike zone are anchored to a white slab of rubber that never moves. The vertical boundaries are anchored to a man who has just adjusted his stance. The rulebook's vertical limits are a line at the hollow beneath the kneecap and a line halfway between the shoulders and the waistband, both of which have to be re-estimated for every hitter. Uncertainty is where framing profits, and the vertical edges carry more of it.

1. 1The pitcher releases the ballThe catcher already knows the pitch and the target, because he called it. That advance knowledge is what lets his glove arrive early rather than react late.
2. 2The ball enters the shadow bandRoughly six inches deep, wrapped around the rulebook boundary. Inside this band the call is genuinely uncertain, and only inside it is framing worth anything.
3. 3The glove closes on the ballOrientation is set before contact: outer half on a pitch off the plate, thumb underneath on a pitch at the knees. The glove is not moved after the catch, it is positioned before it.
4. 4The reception is held stillFor a beat, no longer. The last thing the umpire sees before committing is a stationary glove, which is the whole product.
5. 5The umpire calls itHis judgement is final under Rule 8.02 and nobody on either side may argue it. That finality is what made the skill worth money in the first place.
6. 6Somebody decides whether to challengeUnder the challenge system the pitcher, catcher or batter may appeal immediately. This step did not exist before, and it is the one that caps what framing can be worth.

The window a catcher can influence runs from the moment the ball enters his field of action to the moment the umpire commits. Nothing before it or after it is available to him.

How the metric is built, and why the numbers disagree

Framing was invisible for a century because there was nothing to compare a catcher to. Once every pitch had a recorded location, a comparison became possible, and the construction is worth understanding because it explains both the metric's power and its limits.

The first step is a model of the umpire, not of the catcher. Using the recorded location of every taken pitch and what was actually called, you estimate the probability that a pitch in a given spot gets a strike call, conditioned on the things that demonstrably move it: the height and width of the location, the batter's handedness, and the count. That produces an expectation for every pitch. A pitch in the heart has an expectation near certainty. A pitch four inches off the plate has an expectation near zero. A pitch on the edge has an expectation somewhere in the middle, which is where all the information is.

The second step is to compare the catcher against that expectation, pitch by pitch. A pitch with a modest chance of a strike that gets called one is a credit. A pitch that should have been called a strike and was not is a debit. Sum across a season and the result is strikes gained or lost against a league-average receiver.

The third step converts strikes into runs. A called strike and a called ball send the plate appearance into different counts, and those counts have different expected outcomes; the difference between them, averaged appropriately, is what a stolen strike is worth. Park and pitcher adjustments are applied on top, because catching in front of one umpire crew, one pitching staff and one park is not the same task as catching in front of another.

1. 1Model the umpire, not the catcherEstimate the probability of a called strike for every location, conditioned on count, batter handedness, and the height and width of the pitch. This is the baseline everything else is measured against.
2. 2Restrict attention to the shadow bandPitches that were always strikes and pitches that were never strikes carry no information about receiving skill. The signal lives entirely on the edges.
3. 3Score every taken pitch against its expectationA strike where a strike was unlikely is a credit; a ball where a strike was expected is a debit. Add them up across a season to get strikes gained above average.
4. 4Convert strikes into runsA called strike and a called ball put the batter into different counts with different expected outcomes. The gap between them is what one stolen strike is worth.
5. 5Adjust for park, pitcher and umpireThe same receiving skill produces different raw totals in front of different staffs and crews. Without these adjustments the metric partly measures who a catcher happened to catch.

The published Statcast method in outline. Every provider follows this shape; they differ in the model at step one and the run values at step three, which is why their totals do not match.

Two consequences follow, and both are routinely ignored by people quoting the numbers.

The first is that framing figures are model outputs, not counts. A home run is a fact. A framing run is an estimate that depends on how the umpire model was specified, which conditioning variables were included, and which run values were used. Different public providers produce different totals for the same catcher in the same season, sometimes by a wide margin, and none of them is wrong in the way a miscounted statistic is wrong. This puts framing in the same category as wins above replacement: a genuinely useful construct that people cite with a false air of precision.

The second is an attribution problem the method cannot solve. When a pitch on the edge is called a strike, the credit is shared between the pitcher who put it there, the catcher who received it, the umpire who called it and the hitter who declined to swing. The model assigns all of the residual to the catcher because the catcher is the thing it is trying to measure. That is a defensible convention and it is still a convention.

The umpire is being graded too, and that squeezes from the other side

There is a second force working against framing, and it has been operating quietly for far longer than the challenge system.

Umpires are evaluated on their ball-strike accuracy against the same tracking data that produces the framing metric. Every pitch a crew calls is recorded, compared against the measured location, and fed into a performance review. That is a straightforward consequence of installing cameras in every park: once you can measure a catcher against the umpire, you can measure the umpire against the ball.

The incentive this creates points directly at framing. An umpire whose grade depends on matching a measured zone has a professional reason to resist the very perceptual effect the catcher is exploiting, and to resist it hardest on the pitches furthest from the boundary, which are the ones the grading system will punish him for. He is not doing this out of spite. He is responding to the same feedback loop the catcher is.

The result is a two-sided compression that predates automation entirely. Catchers got better at receiving, which pushed framing value up. Umpires got better at ignoring receiving, which pushed it back down. Both improvements were driven by the identical technology, arriving at the same time, and the net effect on the measured spread between receivers was to narrow it from both directions.

This is worth stating clearly because it changes the story people tell about framing's decline. The challenge system did not kill framing on the day it arrived. It accelerated something that had been happening for years, and the underlying cause in both cases is the same: a skill that depends on an unmeasured judgement gets smaller every time somebody measures the judgement.

What happened to the market, and why the edge closed

The economic story of framing is the cleanest example in sport of a measured skill being competed away.

The sequence went roughly like this. A skill existed and nobody could see it, so nobody paid for it. Measurement arrived and revealed a large spread between the best and worst receivers, large enough that a good framer with a mediocre bat was worth considerably more than the market thought and a poor framer with a good bat considerably less. For a short period, clubs that had done the work could buy that value at a discount, because the players themselves were priced on the visible skills: arm strength, blocking, and offence.

Then the information spread, as information does. And unlike most defensive value, framing turned out to be teachable. It is a set of mechanical habits, not a physical gift. Receiving position, glove orientation, when to move and when not to, all of it can be drilled, and video makes the feedback loop immediate.

So it was drilled. Every organisation now teaches it from the low minors upwards. A catcher who receives badly in the majors is far rarer than he was, because the ones who received badly were either fixed or filtered out on the way up. That compression is the expected end state for any measured, coachable skill, and it means the spread between an elite receiver and an ordinary one has narrowed from what the early public research found. The value did not vanish; it moved from being an inefficiency to being a competence, and competences do not get you a discount.

The clearest visible artefact of that teaching wave is the stance.

The one-knee stance, and what it costs

Catchers used to squat. Many now catch with one knee on the ground, usually the throwing-side knee, in situations that allow it.

The mechanical logic is about the low pitch. A catcher in a traditional squat receiving a pitch at the knees has to reach down for it, which means the glove is travelling downward at the moment of contact and finishes below where the ball was. That presents the pitch as lower than it arrived. Dropping a knee lowers the whole base of the body, so the same pitch can be received with the thumb underneath and the glove finishing level with or above the ball. Same pitch, better presentation, no movement required.

The knee down also stabilises the torso. A squat is an active balance position; a knee is a fixed one, and a fixed base means the small absorptions of a high-speed reception travel into the arm rather than the whole body. It also lowers the catcher's head, which clears the umpire's sightline to the plate.

The costs are real and they are why nobody does it on every pitch. Blocking a ball in the dirt from one knee is harder, because the catcher cannot drive laterally with both legs. Throwing to second from one knee requires an entirely different sequence and takes longer to load. With a runner on and a breaking ball called, those costs bite immediately.

So the stance became situational, which is a more interesting outcome than either side of the original argument predicted. Catchers use the knee-down position with the bases empty or in counts where a low strike is the whole point, and revert to a squat when the running game or a dirt ball is the live threat. Teaching a catcher which is which, pitch by pitch, is now part of the job.

The value that never shows up as a called strike

The metric counts strikes. A good part of what a receiver is worth never appears as one.

Consider what a pitcher does when he trusts the man catching him. He throws to the edge more often, because the downside of missing marginally off the plate is smaller when the marginal miss is likely to be called. He throws his breaking ball at the bottom of the zone in a count where a ball would be expensive, because he believes the pitch will be received rather than dropped or dragged. Both of those are behaviour changes, they both produce better pitches, and neither is a framing credit.

The reverse runs the same way. A pitcher who does not trust his catcher aims at the middle of the plate, which is the safest target and the worst one. Nothing about that shows up as a lost strike, because the pitch that would have been stolen was never thrown.

So the honest description of a good receiver's value has three parts. There are the strikes he gains, which the metric captures. There are the strikes he does not lose through stabbing, recoiling and blocking the sightline, which the metric captures as an absence and which nobody notices. And there is the effect on where his pitchers are willing to aim, which the metric cannot see at all and which may well be the largest of the three.

That third component is also the one least affected by automation, for the plain reason that it operates on the pitcher rather than the umpire. A pitcher who believes his catcher will handle a curveball in the dirt will throw it in a full count whether or not a machine is standing by to adjudicate the ones that stay up.

The honest part: automation is a ceiling on all of this

Any article about framing that does not deal squarely with the automated strike zone is describing a world that no longer exists.

The majors now run an automated ball-strike challenge system. The plate umpire calls every pitch as before, and his call stands unless the pitcher, catcher or batter immediately asks for a review, with each club starting the game holding two challenges and keeping the ones it wins. When a challenge is made, the pitch is re-adjudicated against a flat rectangle at the midpoint of the plate, seventeen inches wide, with its vertical limits taken from the batter's measured height. The full mechanics are set out in the piece on how the challenge system works.

Three things follow for framing, and they point in different directions.

Most framing survives untouched. The overwhelming majority of pitches are never challenged, because a club holds two challenges for an entire game and has to spend them where they matter. Every unchallenged borderline pitch is still adjudicated by a human being watching a glove. The mechanism that made framing valuable is intact for nearly all of the sample.

The top end of framing is gone. A strike stolen from well outside the zone is precisely the pitch a hitter will challenge, because it is the one he is most confident about and the one where the machine will most clearly agree with him. The catcher who was best at converting genuinely bad pitches into strikes has lost the part of his edge that was most visible and therefore most contestable. What remains is the marginal stuff: the pitch that clipped an edge, where the hitter is not sure enough to spend a challenge and would probably lose if he did.

The skill has grown a second half. The catcher is the best-placed person on the field to judge a pitch, because he sees it from directly behind, at rest, with the plate underneath him. Under the challenge system that judgement is worth something concrete, and it now sits alongside receiving as part of the same job. A catcher who frames a ball into a strike has no reason to challenge anything. A catcher whose good pitch was called a ball has to decide, in about two seconds and without help from the bench, whether to spend a resource he may need in the eighth.

The end state is not in doubt. If the sport ever moves from a challenge system to full automation, framing value goes to zero on the day it happens, because a measurement cannot be persuaded. The rectangle does not care where the glove finished. Everything described in this article would become a historical curiosity, along with the several hundred million dollars of contract value that the discipline quietly redirected over the last decade.

Whether that day comes is a question about appetite rather than technology. The challenge model exists because full automation was tried, worked exactly as designed, and produced a game some people did not enjoy watching. That is a preference, not a technical constraint, and preferences change.

What framing is worth in the meantime

The useful way to think about a receiver now is as a man doing three jobs whose relative weights have just been reset.

Receiving is still the largest of the three by volume, and it is now bounded rather than open-ended. A catcher can no longer be worth an enormous amount by stealing the indefensible, because the indefensible gets appealed. He can still be worth a real amount by never losing a strike his pitcher earned, which is the defensive half of the skill and the half that never got the attention.

That defensive half deserves a note of its own, because it is the part the public metric has always undersold. Some meaningful share of framing value is not strikes gained but strikes not lost: pitches that were plainly strikes and got called balls because the catcher stabbed at them, recoiled on impact, or blocked the umpire's view. Those are unforced errors, they are entirely avoidable, and a catcher who commits none of them is providing value that looks like nothing at all on television.

Blocking and the running game are the other two jobs, and both have quietly appreciated. Under a pitch timer with a disengagement limit, runners attempt more, and a catcher who cannot throw is a standing invitation. A pitching staff throwing more breaking balls below the zone needs a receiver who can keep them in front of him. Neither of those skills is contestable by a challenge, which means neither has a ceiling imposed on it, which means both are worth relatively more than they were.

The catcher's other job, the one nobody has ever successfully measured, is the sequencing conversation with the pitcher. It has resisted quantification for as long as people have tried, and it is the last part of the position that is still argued about in the language scouts used before anyone had a tracking system. Some clubs believe in it enormously. The honest position is that nobody has demonstrated it convincingly, which is not the same as demonstrating that it is not there.

What to watch, next time you look at a catcher

Four things, all visible without any data at all.

Watch when the glove arrives, not where it finishes. A receiver whose glove is set in the neighbourhood before the ball gets there is doing the hard part; the reception is the easy bit. A glove that is still travelling at contact has already given the pitch away.

Watch the low pitch specifically. That is where the difference between a good and an ordinary receiver is largest, because the bottom edge is the hardest boundary for an umpire to fix and the easiest for a catcher to present badly. Thumb under, glove finishing level, no downward drag.

Watch what happens after the ball is caught, for one beat. Elite receivers are dull to watch for exactly this reason. Nothing moves.

And watch which pitches get challenged. That is the new information the sport did not previously produce. A catcher whose stolen strikes are being appealed and overturned is doing something that used to work and no longer does, and the pattern of which pitches get taken back is the clearest picture anyone has ever had of where the human zone and the measured one actually differ.

The receiving position has always been the strangest job in the sport, and the comparison with the wicketkeeper's version of it is instructive: same crouch, same hands, entirely different relationship with the official standing behind. More on how the sport actually works is collected in the baseball archive.