# What the Heck Is a Flat Sinker, Anyway?

The week before Christmas, I conducted an investigation into the Giants’ strange sinker-ballers. Logan Webb and Alex Wood enjoyed spectacular years, and they both did it using sinkers they released from a low starting point, which created a unique look for batters.

You probably didn’t read that article, and that’s okay. It was the peak of the holiday season. You were likely out in the world like me – seeing family, drinking eggnog-flavored coffee beverages, and generally making up for last year. My consumption of baseball media went way down, and I do this for a living.

As a self-interested person, I suggest you go back and read that article. I thought it was pretty good! More importantly, though, I’ve been doing more research into what the heck a “flat sinker” even is. The concept just doesn’t fit into my brain, and when that happens, I like to hit myself over the head with data until something clicks. So today, please enjoy some random things I’ve researched while trying to understand why in the world “flat” (approach angle) and “sinker” (pitch type) coexist.

My first takeaway: the combination of sinkers and flat vertical approach angles isn’t as weird as you think. As it turns out, vertical movement isn’t as important for approach angle as the three more obvious factors: how high the ball is at release, how fast it is thrown, and how high it is as it crosses the plate. Other things matter too, but it just makes sense: if you want to know how quickly the ball is falling as it moves forward (that’s what vertical approach angle is in plain English), where it starts and where it ends up will tell you most of what you need to know.

Don’t believe me? Here’s an example. First, consider two pitches thrown at the same speed from the same release point that cross the plate at the same height. Let’s say, for the sake of argument, that one is a sinker and one is a four-seam fastball. Clearly, the four-seamer will be falling less steeply when it crosses home plate. Its movement fights against gravity – it has more “ride,” or positive vertical break after accounting for gravity.

But the vertical approach angle of those two pitches wouldn’t be all that different. That sounds wrong, because of how we conceive of each pitch. We call them “sinkers,” after all. But when you actually look into the formula (to handle large swaths of pitch data, I’m using regression weights), pitch movement simply doesn’t matter much for the angle the pitch describes when it crosses the plate.

Let’s use Zack Wheeler as an example, because I love watching him pitch. Here’s a pitch he threw on July 23, a 96.9 mph sinker that crossed the plate 1.29 feet off the ground after he released it 5.49 feet off the ground:

Here’s a pitch from June 16, a 96.7 mph four-seamer that crossed the plate 1.29 feet off the ground after he released it 5.5 feet off the ground:

Those were different fastballs, but they started and ended at roughly the same point. They just took different paths to get there. The sinker has a VAA of -7.18 degrees. The four-seamer? Negative 7.15 degrees. In other words, the angle at which both pitches crossed the plate is basically the same, despite the four-seamer having two extra inches of ride.

Only two inches? Maybe the minor difference in approach angle is due to a minor difference in movement. Only, it’s not. Try these two fastballs instead. First, a 98.7 mph sinker from May 1. This pitch was released 5.6 feet off the ground and crossed the plate 1.24 feet off the ground, with a VAA of -7.19 degrees:

Next, a 97.8 mph four-seamer from August 3. This one was released 5.62 feet off the ground and crossed the plate 1.25 feet off the ground, with a VAA of -7.22 degrees:

The four-seamer reached the plate with a more negative VAA than the sinker, despite 4.5 extra inches of upward vertical break. That extra velocity mattered more than the movement – the faster you throw a pitch, the less time it spends accelerating downward due to gravity. The one-sentence synthesis of all of this? Vertical approach angle doesn’t describe the same thing as vertical pitch movement.

Another way to think about it: if you want to release a sinker and a four-seamer at the same height, and have them cross the plate at the same height, you’ll need to aim lower with the four-seamer. That’s only logical – the four-seamer will fight gravity more, so if you threw them at the same initial trajectory, it would end up falling less. That more negative release angle more or less offsets the pitch movement, such that while vertical break matters, it’s a small component overall.

That’s really weird! It makes me wonder if the metrics we use to evaluate pitches reflect the reality that batters perceive on the ground. Do batters perceive “flat sinkers” – with a shallow VAA – similarly to four-seamers, and “steep four-seamers” similarly to sinkers? I’m planning on looking into that in the coming weeks.

In the meantime, here’s some other data I’ve collected on “flat sinkers,” which I keep saying because it’s fun to say. In the Giants article, I looked at how sinkers of various approach angles performed when thrown around the knees. Shallower approach angle pitches did better – that’s the effect I believe the Giants are looking for. The key reason for that is that shallow-angle pitches low in the zone are more likely to be taken for strikes.

An easy counterpoint to “batters are more likely to take these pitches for a strike” is “batters mash sinkers that don’t sink, and flat sure sounds like not sinking.” To test this theory, I looked at what happens when batters swing at either extremely flat, normal VAA, or extremely steep sinkers. First, here are the ball-in-play results for sinkers that cross the plate between 1.25 and 1.75 feet off the ground, roughly the bottom of the zone and the area just below that:

Batted Ball Results, Sinkers 1.25-1.75 Feet
Group VAA Velo V-Break (ft) wOBA xwOBA EV Barrel%
Flat -6.95 94.5 0.62 .339 .332 91.9 2.8%
Average -7.59 93.1 0.71 .332 .316 89.3 4.6%
Steep -8.21 91.4 0.75 .348 .324 88.5 5.1%

There’s not a lot to distinguish the batted-ball results of our three groups of sinkers (of note, I excluded pitches released less than four feet off the ground, because otherwise the flat population over-counts sidearm and underhand specialists). It’s strange to see flat/shallow sinkers allow the lowest barrel rate despite the highest exit velocity, but I think it’s mainly a sample size issue; the three groups look mostly identical to me aside from velocity. That velocity also tracks; as we covered above, the harder you throw a pitch, the shallower the approach angle, holding everything else constant.

What if you miss a bit higher in the zone? After all, plenty of pitchers would look good if they only counted the fastballs that dotted the bottom of the zone. How does it look between 1.75 and 2.25 feet off the ground?

Batted Ball Results, Sinkers 1.75-2.25 Feet
Group VAA Velo V-Break (ft) wOBA xwOBA EV Barrel%
Flat -6.44 94.4 0.66 0.377 0.361 90.6 6.0%
Average -7.09 93.1 0.75 0.383 0.366 91.4 6.2%
Steep -7.71 91.2 0.79 0.365 0.368 90.9 5.9%

Again, not much batted ball effect to be found. One amusing thing that carries through both samples? The flattest sinkers have the least positive vertical break. Most sinkers have positive vertical break, just like a four-seamer. They just have less of it, which means the pitch “sinks” relative to the path of a four-seam fastball. In this case, the flattest pitches actually sink the most on average, further underscoring the tenuous relationship between flat (a description of approach angle) and sink (a description of pitch movement).

What does this giant pile of technical terms and data mean? I’m inclined to say that approach angle just doesn’t mean much for contact quality, at least when it comes to sinkers. There are all kinds of confounding variables here – maybe flat sinkers do get shelled on average, but the pitchers who throw them are getting more movement, or more deception, or something else offsetting. I plan on digging in further to isolate these differences.

There’s also another confusing part. The reason that shallow-VAA sinkers seem to work so well at the bottom of the zone is that batters swing at them less frequently. But “swinging less frequently at borderline pitches” is mostly a good thing for batters. If you have a pitch in your arsenal whose standout quality is that it keeps bats on shoulders, you’ll need pinpoint command to take maximum advantage of it. If you miss low, you’re far less likely to get bailed out by a swing.

Still, despite all these qualifications, I’m convinced that there’s something here. The Giants have done a good job of developing and acquiring pitchers who release the ball low and command sinkers at the bottom of the zone, and it paid off for them in 2021. They brought Wood back in free agency, too – they clearly believe in his results. It makes me curious about Alex Cobb – he signed with the Giants and throws a lot of sinkers, but from a high release point. Will they ask him to tinker with his delivery? Was he just a good value regardless of the fact that he doesn’t throw this novel sub-category of sinker? I have no idea! But I’ll be watching to find out, and doing my own research in the meantime.

Ben is a writer at FanGraphs. He can be found on Twitter @_Ben_Clemens.