Hand a hitter a list of every pitch a starter will throw, in order, and you have ruined the starter — even if every pitch on the list is excellent. A 98-mph fastball the hitter knows is coming is far easier to handle than an 89-mph slider he never saw. This is the truth that raw pitch grades keep tripping over: a pitch’s value depends enormously on the pitch before it, and on whether the hitter could tell the difference in time to react.

Two ideas govern that in-between space. Sequencing is the order and pattern of pitch types and locations — what you throw, where, and in what succession. Tunneling is the geometry that makes sequencing deadly: throwing different pitches along the same initial path so they look identical out of the hand, until the hitter has already had to decide whether to swing. Get both right and two ordinary pitches become more than the sum of their parts, because deception multiplies what stuff alone can buy.

Sequencing: the order is the weapon

Sequencing is the strategic layer of pitching — the part a catcher and pitcher negotiate with every set of signs. It covers which pitch follows which, how locations move around the zone, how the pitcher changes a hitter’s eye level, and how he plays speeds off one another. A well-sequenced at-bat is built, not improvised: a pitch is thrown not only for what it does on its own but for what it sets up next.

The logic is the logic of expectation. A hitter has a fraction of a second to identify a pitch and commit, so to buy time he leans on probability — what is likely here, in this count, from this pitcher. Sequencing attacks that lean: establish a pattern, then break it. Show three straight fastballs to set the timing of his swing, then pull the string with an off-speed pitch out of the same arm action. The pitches need not be elite; the contrast between them, against the hitter’s expectation, is what produces the whiff or the weak contact.

Tunneling: same path, different destination

Tunneling is the physical mechanism that makes deceptive sequencing possible. A hitter cannot watch the ball all the way to the bat; the pitch arrives too fast. So he makes his swing-or-take decision early, while the ball is still a fair distance away — at what analysts call the commit point or decision point, the last moment he can still adjust. Before that point he is reading; after it, he is committed.

Two pitches tunnel well when they travel along nearly the same path up to that commit point — sharing what is loosely called a “tunnel,” a corridor in which the trajectories are still indistinguishable to the eye. Out of the hand they look like the same pitch. By the time they diverge — the fastball staying true, the breaking ball dropping away — the hitter has already decided, and the decision is now wrong: he has committed to a fastball and a slider is breaking under his bat. The break itself is old news; the deception happened earlier, in the shared corridor before the commit point, where the pitches refused to declare themselves.

Why tunneling multiplies stuff

This is the crucial idea, and it is multiplicative rather than additive. A nasty pitch in isolation is worth its nastiness. The same pitch thrown along the same tunnel as the pitcher’s other offerings is worth more, because it borrows the hitter’s wrong guess about everything in that tunnel. Deception is a multiplier on movement, not an addition to it.

Flip it around. A devastating breaking ball that announces itself out of the hand — a different arm slot, a visible hump, a release point that screams “curveball” — loses much of its bite, because the hitter identifies it early and adjusts inside his decision window. The break is the same; the value is not, because the hitter was never fooled. A fastball and slider sharing a path, by contrast, each play up beyond their individual grades, because each makes the hitter wrong about the other. The relationship between two pitches carries value neither pitch carries alone.

A clearly-illustrative example: the shared tunnel

Picture a right-handed pitcher with a four-seam fastball and a slider. Both leave his hand from the same release point, with the same arm speed, and for the first stretch of flight they ride the same line — toward, say, the inner third against a right-handed hitter. Early in the path the two are visually identical. The hitter, watching that shared line and respecting the fastball’s velocity, commits to swing at what he reads as a fastball boring in on him.

Then the paths split, after the commit point. The fastball holds its line; the slider, having masqueraded as that fastball, breaks down and away off the plate. The hitter has committed to the inside fastball, and his bat travels through a zone the slider has just vacated — swing and miss, or a roller off the end. Notice what did the work: not the slider’s break in a vacuum, but the fact that it lived in the fastball’s tunnel until the hitter could no longer do anything about it. (The pitcher and at-bat here are entirely hypothetical, chosen to isolate the one mechanism — the shared path before the commit point.)

How tracking data lets analysts study it

For most of baseball history, tunneling was a thing pitching coaches felt and described in metaphor. Pitch-tracking changed that by recording the full trajectory of every pitch — its release point to a fraction of an inch and its position throughout flight. With those trajectories in hand, analysts can ask quantitative questions that used to belong to the eye: How close are a pitcher’s fastball and slider release points? How tightly do their paths overlap up to the commit point, and how far do they separate after it? The distances between two pitches at release, at the decision point, and at the plate together describe how well they tunnel and how violently they break apart.

This is also where tunneling connects to perceived velocity — how fast a pitch appears, as distinct from its radar reading. A pitcher who releases the ball closer to the plate through longer extension gives the hitter less time to react, so a 95-mph fastball can “play” quicker; tunneling compounds it, because a fastball sharing a path with off-speed pitches makes the hitter wait and read. Studying release, extension, and trajectory together is the only honest way to study deception — a property of how pitches relate, not of any one alone.

The relationship to Stuff+ — and the caveats

It helps to set this beside its natural complement. Stuff+ grades a pitch’s physical quality — velocity, movement, spin — largely in isolation, asking “how filthy is that pitch on its own?” Stuff models partly account for how a pitch plays off an arsenal, but their core unit is the individual pitch. Sequencing and tunneling occupy the space those models see least clearly: the value created between the pitches. A modest-Stuff+ pitch thrown out of a great tunnel can outperform its grade; a filthy pitch telegraphed out of the hand can underperform. Stuff is the raw weapon; tunneling and sequencing are how it is aimed in time.

The honest caveats are large, and they are why tunneling resists clean measurement. First, it is hard to isolate from stuff and command — a pair of pitches may succeed because they tunnel, because the stuff is overpowering, or because the locations were perfect, and untangling those from outcomes is messy. Second, the catcher and game-calling matter: the same arsenal sequenced cleverly or predictably produces different results, so crediting the pitcher alone is a mistake. Third, hitters and counts vary. The mechanism is real and visible in the trajectories; the precise value it adds is one of the harder things in the sport to pin to a number, which is why analysts hedge when they try. It is a useful reminder, much like the context-stripping logic behind FIP and xFIP, that a pitcher’s true skill and his recorded outcomes can drift apart.

The bottom line

A pitch is never thrown into a vacuum. Sequencing sets a hitter’s expectation, and tunneling exploits it by sending different pitches down the same path until the hitter has already committed. Two pitches that share a tunnel are worth more than their individual stuff, because deception multiplies movement instead of adding to it: each pitch borrows the hitter’s wrong guess about the other. Tracking data turned all of this from coaching metaphor into something analysts can study through release points and trajectories, alongside Stuff+, which grades the pitch in isolation while sequencing and tunneling grade the spaces in between. Just hold the caveats: deception is hard to separate from raw stuff and command, the battery shares the credit, and the precise value resists a tidy number. The art between the pitches is real — it is simply the part of pitching that has always been easier to feel than to measure.

Sources & Further Reading

  • Baseball Savant — the Statcast pitch-tracking data on release points, movement, extension, and trajectories that makes tunneling analysis possible.
  • FanGraphs — for Stuff+, Pitching+, and broader writing on pitch quality and deception.
  • MLB.com — the official Statcast glossary, including perceived velocity and extension.