Survival of the Hittest, Part 1

An analytical hitter omnibus

Outside of the Pirates analytics department I may be the one person who has seen Esmerlyn Valdez swing a bat more than anyone. Really, we’re talking hundreds of swings in slow motion, looking at differences between fastball, breaking pitches, offspeed, inside, outside, whiffs, hard hits, home runs, trying to find minute patterns and differences. Eventually I noticed a pattern. As a pitcher winds up to throw the pitch Valdez gets into his load position. As the pitcher releases his pitch Valdez tips his bat forward just a slight bit, before cocking it back and letting it rip.

Most hitters rock the bat back and pauses until the pitcher throws the ball, and you can actually find the bat parked in the same spot every time. Valdez you would find somewhere different on every pitch, because the bat never stops movement. I call it the waggle, and it’s the most visible thing about one of the most interesting hitters in baseball, who also happens to swing through more hittable pitches than nearly anyone else in the game. So the natural question is whether the waggle means anything, and answering it turned into a much bigger question about what actually happens in the four-tenths of a second between a pitcher letting go and a hitter making contact (or not). This is the first of two parts. This one covers everything that happens before the swing really starts: what a miss looks like, what the waggle is doing, and when the bat finally goes. Part 2 covers the swing itself.

Take a Whiff of This

Lets start with the swing and miss, because it turns out to be the most predictable thing in baseball. Statcast's bat tracking records where the bat came closest to the ball on every swing, including the ones that miss, which means a whiff has a location just like a home run does. In the 3D Swing Explorer that location becomes a cloud of cubes around the hitter, and once you look at enough hitters' whiff cubes a pattern jumps out that holds for almost everybody. On fastballs, hitters miss behind: the spot where the bat passed the ball sits deeper, closer to the body, than where that hitter usually meets a fastball. On breaking balls and changeups they miss ahead, out in front of where they usually meet those pitches. Fastball whiffs are late and everything-else whiffs are early, for nearly every hitter in the league.

The numbers are about as lopsided as anything baseball gets. 86% of hitters have their fastball misses sitting deeper than their fastball swings overall, 96% miss those out front, and for changeups and splitters it is 180 of 181 we measured were out front. The miss is never far from the hitter's normal spot (it is a little behind or a little ahead, not a different zip code), but the direction almost never changes. A hitter who swings through a fastball was late, and a hitter who swings through a curveball was early, which sounds obvious until you realize how rarely anything in hitting is that clean.

Valdez is the extreme version of the first half of that sentence. Against right-handed pitchers he swings through 34% of the fastballs he offers at compared to the league average of 23%, and his fastball whiff cubes sit up and in close to his body, which is exactly where a late bat ends up. That has a lot to do with why his Contact+ (my hitting grade for a hitter’s ability to make contact, with 100 as average) is 67.9, the second lowest among qualified hitters, ahead of only Munetaka Murakami.

Video of his swings makes the same point one pitch at a time. Split every Valdez swing this season into thirds by how far along his bat was a tenth of a second before the ball arrived, and the misses sort themselves. On fastballs he whiffed 42% of the time when his bat was furthest behind and 23% when it was furthest along. On breaking balls it flips: 26% when furthest behind, 63% when furthest along. Changeups and splitters look like breaking balls, 32% and 67%. Those two patterns run in opposite directions, which is why lumping all his swings together shows nothing at all, and why a season whiff rate can hide the fact that a hitter is making two different mistakes. One detail to keep in your pocket for later: on his fastball whiffs against righties, Valdez started his swing at the same moment he did on the fastballs he hit. Being late on a miss is something that happens during the swing, not at the start of it.

Wiggle While You Waggle

So is the wag a problem? To find out, I tracked the bat in broadcast video of Valdez's swings from release to contact, frame by frame, marking the knob and the tip of the bat by hand on a set of swings and then following the bat through the rest of his season. What comes out is a routine. On every pitch the bat climbs to the top of the wag at the same moment after release, about fifteen hundredths of a second, whether the pitch is a fastball, a slider or a changeup. Then it tips over and the swing begins. The top of the wag arrives at the same moment on his whiffs as on his contact, and it does not care what the previous pitch was. The wag is a clock, not a guess.

Showing the technical details of how we used Computer Vision to track bat movement

It is not a perfectly rigid clock, and the way it flexes is the interesting part. The wag is biggest when Valdez is ahead in the count (2-0 and 3-1 are his loosest) and about 10% smaller with two strikes, the hitter's version of choking up without moving his hands. His bat speed follows the same pattern, a mile and a half per hour faster when he is ahead than with two strikes, which makes it tempting to think the wag is where the power comes from. It is not. On any single swing, a bigger wag does not come with a faster swing, and on balls in play a bigger wag does not come with harder contact. The count changes both things at once, the way a hitter sitting 3-1 does everything a little bigger and a hitter protecting with two strikes does everything a little smaller. The wag is a readout of intent. You can see roughly how hard he is about to go before the swing starts, but the wag is not the thing making it happen.

Valdez is also not alone. Plenty of hitters keep the bat moving after release, and some move it more than he does: Carlos Narváez and Taylor Trammell have the busiest bats of the hitters I measured. At the other end, Aaron Judge, Ian Happ and Kyle Stowers hold the bat almost perfectly still until they fire. If the wag were what made Valdez miss, the busy bats would be the whiffers and the still bats would be the contact hitters, and that is not how it shakes out. Narváez is busy and swings and misses a lot; Marcus Semien and Jackson Chourio are busy and make plenty of contact; Judge is as still as a statue and still swings through his share. Across hitters, how much the bat moves before the swing has nothing to do with Contact+. The wag is a style, not a flaw, which also means it is not where Valdez's problem lives.

T-Minus 0.4 Seconds to Swing Launch

If the wag is not the start of the swing, something else is, and Baseball Savant happens to show it for every hitter. The swing path page animates each hitter's average swing from Statcast's bat tracking, the same slice of tracked frames before contact for everybody, which makes it possible to find the frame where the bat actually starts rotating toward the ball. Call that the launch.

Seen this way, Valdez has two separate events where most hitters have one. First the wag finishes, with the bat tipping over and laying off to the side, which from the center field camera looks like the start of the swing. Then, about a tenth of a second before contact, the bat actually launches forward. Judge, by comparison, has one event: a still bat, then launch. And when it comes to the launch, Valdez goes later than almost anyone in baseball, tied for seventh latest of 402 qualified hitters.

Launching late is not a flaw in itself, and the league makes that clear. Split hitters into fifths by when the bat launches and a trade appears. The latest launchers make less contact (a Contact+ of 97.4, against 105.9 for the earliest group) and whiff more (22% of swings against 19%), but they also hit with more authority, with a Power+ (bat speed and attack angle adjusted for pitch location, 100 as average) of 104.1 against 94.9. Their xwOBA (expected weighted on-base average, a measure of overall production based on the quality of contact) comes out dead even: .305 for the latest fifth, .309 for the earliest. A late launch buys power and pays for it in contact, and the bill comes out about even. It is also its own thing: when a hitter launches has almost nothing to do with how long his swing is, and only a loose connection to how fast it is (later launchers swing a little faster, which is what lets them wait).

Which brings it back to Valdez, because the trade explains only part of him. The latest launchers as a group carry a Contact+ in the high 90s. He is at 68. Junior Caminero and Chase Meidroth launch even later than Valdez and make above-average contact, so a late start does not doom anybody. Late launch is part of the Valdez profile, but it does not cover the size of his contact problem on its own, and the rest of the story is in what happens after the bat starts moving.

So here is where the first four-tenths of a second leave us. The miss has a direction, and it almost never changes: late on fastballs, early on everything else. The wag is a personal clock that flexes with the count and does not decide whether a hitter makes contact. The real start of the swing comes later than it looks, and when it comes is a trade every hitter makes between contact and power. By the time the bat launches, the start is set. Everything after that is a budget, and Part 2 is about how hitters spend it.

The Short Version

  1. Whiffs have a direction that almost never changes: 86% of hitters miss fastballs late and 96% or more miss breaking balls and changeups early.

  2. Valdez's wag runs on a schedule: the top of the wag comes at the same moment after release on every pitch, on whiffs and contact alike, smaller with two strikes and bigger when he is ahead.

  3. The wag is a readout of intent, not a power source, and across hitters how much the bat moves before the swing has nothing to do with how often they miss.

  4. The real launch comes later than the wag suggests, and later launchers trade contact for power while producing about the same overall.

  5. Valdez launches later than all but a handful of hitters, but that only explains part of his whiffs. The rest happens during the swing.

Part 2 picks up the moment the bat launches: how much time a hitter can win back once he has started, and the hitters who guess, the hitters who react and the one who never changes his clock at all.

Get Balls & Sticks in your inbox

Subscribe for free to get new posts by email as soon as they publish.