Sink or Spin: What Makes Sinkers Effective?

Sinkers have always intrigued me. In a modern setting, with baseball fully into a lift big/throw hard craze, the sinker has fallen out of favor. Is that shift justified? That of course depends on how you define effectiveness.

We as an industry generally tend to define the effectiveness of a certain pitch by its ability to generate whiffs, and sure, of course, this makes sense. Especially in the fantasy baseball community, where strikeouts are king, sinker-ballers tend to get overlooked, and once again that is justified. Granted this is a fantasy site, but hear me out; sinkers can be just as effective.

Sinkers don’t generate whiffs — in fact, they collectively have the lowest whiff% of any pitch.

But that’s not the point.

It is important to mention the difference between sinkers and two-seam fastballs, which have interchangeable definitions depending on who you ask. For that, look no further than this piece by David Laurila from FanGraphs.

Let’s dive into what makes for an effective sinker.

 

Turbo Sinker Central

 

The added emphasis on spin metrics, along with a continued league-wide increase in velocity, has led to a fair amount of what can best be described as “turbo sinkers.” No group of pitchers exemplified the turbo sinker quite like the world champion Dodgers’ bullpen, with a host of characters including Brusdar Graterol, Blake Treinen, and Victor Gonzalez all averaging 95 mph or higher on their SI in 2020. Add in the occasional appearance from Dustin May, who set pitching twitter ablaze with this pitch to Manny Machado, and you have a group that makes a hitter’s hands sting before they even step into the box.

 

 

To put it in perspective, Graterol was tied for the major league lead in average SI velocity at 99.3 mph (!), which is even a tick higher than his 4-seam FB. While Gonzalez only averaged 94.9 mph, that marginal velocity has an added impact for LHP, where he placed seventh respectively.

If you watched the Dodgers this season — and given the season’s outcome, chances are you did — I don’t have to tell you that those pitchers performed well in 2020. Treinen was resurgent after signing a one-year deal, Gonzalez emerged as a multi-inning weapon after a minor league career in the rotation, and Graterol was the Dodgers’ consolation prize for having to acquire Mookie Betts.

But how did they fare compared to more traditional sinkers? For that, we stay within the Dodger bullpen. Dylan Floro, Adam Kolarek, and Scott Alexander are also oft-used members of the Dodger staff and like their pen-mates above, all have sinker-heavy pitch mixes. Let’s compare them:

 

In this small glimpse into the world of sinkers, we don’t see many patterns. Kolarek has the best run value, but he throws it nearly 80% of the time. Alexander has similar usage, spin, and whiff rates as Kolarek, but with 4 mph added velocity causing the HH% to nearly double. All that before you even mention the anomaly that is Brusdar Graterol.

So at this point, the question remains: What makes a sinker effective?

 

Wake Up

 

To state the obvious, the simplest definition of pitcher effectiveness is run suppression. The general point of a sinker isn’t necessarily to miss bats, but to miss the barrel of the bat. Whether that is achieved by getting the hitter to hit the  ball weakly or by missing the bat entirely doesn’t really matter, so long as it effectively limits run production. Looking strictly at xwOBA, sinkers faired pretty similarly to 4-seam Fastballs in 2020:

 

How does this work? As David Price mentions in the FanGraphs article above, “You don’t want just lateral movement. You want vertical…That’s what misses [the barrel of] bats.”

But not all pitchers want batters to swing though. As Rick Porcello notes, “I just want it to have late run … I’m trying to freeze a hitter.” While this might not be the most common goal for sinkers, I want to focus on what Porcello calls “late run.”

We hear the term “late movement” used a lot in baseball, especially from players and scouts. But most physicists say that “late” movement is nothing more than an optical illusion, and can be predicted based on the ball’s spin, velocity, and release. Research on the difference between observed and expected movement has uncovered a phenomenon called Seam-Shifted Wake (SSW).

Now, as someone who nearly failed AP Physics 2, I am quite possibly the least qualified person to be discussing this. For much deeper explanations, you should absolutely read the work of Dr. Barton Smith and his team at USU ,whose research coined the term, Alan Nathan’s work on laminar pitches, and of course, the cutting-edge work of Driveline baseball.

The TL;DR version is this: A baseball’s combination of seams with an otherwise smooth surface uniquely effects the airflow around it, which in turn allows its flight path to be easily manipulated. This manipulation causes air behind the ball to become turbulent, breaking the symmetry of its wake. We can attribute a ball’s abnormal movement to the breaking of this wake — which we now know as Seam-Shifted Wake.

 

The Physics of Sinkers

 

There are many forces that contribute to pitch movement, and harnessing those forces is the overarching goal of modern pitch design. For the basics, I direct you to these these posts from Driveline Baseball.

With this, we know the two main contributors to the flight path of a baseball are spin rate and spin direction. A typical 4-seam FB is thrown with backspin, and due to the Magnus effect the more backspin it has, the less it will drop on it’s way to the plate. The faster the ball spins backwards, the higher the force pushes it upwards. In their research, Driveline found a correlation between velocity and spin rate on 4-Seam FBs to be between .83 and .96 (with variance depending on pitcher’s grip and ability). This is great for 4-seam FBs, in which pitchers strive for perceived “rise,” but what does it mean for sinkers?

Theoretically, high spin rates have a detrimental effect on the drop of a baseball. For sinkers, this means that higher spin rates will increase horizontal movement but also decrease vertical movement (drop) right? Not so fast.

For the sample of all pitchers who threw at least 100 sinkers in 2020, the correlation is just .11. Additionally, there was a correlation of .19 between spin rate and vertical movement and a correlation of just .01 between spin rate and horizontal movement. What this means is that a sinker’s raw spin rate can only explain around 11% of its velocity, 19% of its vertical movement, and 1% of its horizontal movement.

This is because we have ignored the second factor of a pitch’s movement: Spin direction.

As detailed in Driveline’s work, Sinkers are a logical starting point because their spin efficiency (amount of spin contributing to the ball’s movement) is typically closer to 100% compared to other non-fastballs. This makes it easier to compare the deviation between expected and observed movement (the deviation being attributed to SSW).

Because SSW is just a concept, this comparison takes some deduction, along with access to spin axis metrics. Spin axis metrics which, until recently, were not publicly available.

 

OK, Google

 

Enter Google Hawk-Eye.

With MLBAM and Statcast’s switch to Hawk-Eye, they are now able to directly measure a pitch’s spin axis (instead of reverse-engineering it like they had prior with Trackman). Lucky for us, this data is now public on BaseballSavant.

Using a pitch’s movement profile to determine an inferred spin axis, and Hawk-Eye data to determine a true observed spin axis, we can calculate the difference between them — a metric Driveline calls 2D Axis Deviation.

If there is one important piece to remember for spin axis analysis, it is this:

Pitches with a positive axis deviation have more movement down and to the arm side than expected based on spin. Pitches with a negative axis deviation have less movement down and to the arm side than expected based on spin.

So a pitch effected by SSW will have more sink than is typically expected. The physical reasoning is explained in research by Professor Glenn Healey at UC Irvine who attempts to isolate SSW from all forces on a ball, in which they find the leaders in “side force.” But does this phenomenon alter a sinker’s effectiveness? Driveline’s preliminary research thinks so, finding a pretty direct relationship between 2D axis deviation and their “arsenal score” (run value – expected run value). In essence, the more a pitch is effected by SSW, the more it will outperform expectations.

 

More Sink ≠ More Groundballs

 

Expectations aside, with the help of Jeremy Seigel we analyzed the correlations between spin axis deviation and general quality of contact metrics and found … nothing? Almost nothing. In a sample of all pitchers who threw at least 100 sinkers in 2020, there was no significant relationship between spin axis deviation and Whiff%, Hard-Hit%, exit velocity, or xwOBA. However, when changing to the absolute value of spin axis deviation, we found a relatively strong relationship with launch angle.

Now, there is a ton to unpack here, more than we have time for in this article. Granted, we are examining axis deviation’s effect on the general quality of contact metrics rather than its ability to outperform expectations, but the results were surprising nonetheless. The most surprising to me was finding a relationship with absolute value of axis deviation and launch angle, as we have already mentioned that having positive and negative deviations are opposite. This suggests that any large axis deviation is beneficial for generating grounders.

So while those affected by SSW (with more sink than expected) are in fact more effective at suppressing LA than average sinkers, so are those with less sink than expected.

 

Hitter’s Expectation

 

I have always hypothesized that sinker effectiveness was more dependent on a pitcher’s arsenal than a typical fastball. Anecdotally, you want either the low spin necessary to generate weak contact or enough velocity/movement to make up for it. Preferably both, but that’s impossible right? This was pure speculation based on the idea that sinkers are essentially tricking a hitter into making weak contact. To trick a hitter, you need an element of surprise, which can now be partially identified as SSW, but can also be aided by pitch mirroring.

So what pitch mirrors a sinker?

This reminded me of an article by The Athletic’s Eno Sarris that I keep coming back to. It might seem like the sinker by now, but if I had to pick a favorite pitch, it would be the changeup. We already have a pretty good understanding of what makes changeups effective; high velo changeups with lots of movement, and straight changeups with large velocity gaps to the FB. Changeups are notoriously ambiguous, but most tend to move down and to the arm side — similar to sinkers. Further, going back to Professor Healey’s work, we see that the two pitches with the most side force are sinkers … and changeups.

So, could a mirrored combination of sinker and changeup be as effective as the well established FF/CB combination? With the effects of SSW, pitchers could now have a potentially lethal two-pitch combination that not only looks the same to a hitter, but also has more movement than expected. No doubt, easier said than done. But in theory, this is maximum deception.

 

 Dodger Differences 

 

Now, back to the Dodgers, and specifically the effectiveness of turbo sinkers. For this, I’ve sorted all sinkers throw in 2020 by velocity.

This league-wide table would have you believe that turbo sinkers are more effective at both suppressing launch angle and getting more whiffs. Now let’s use this and what we know about SSW to put the Dodgers bullpen in context.

Here are the same tables from the beginning, but with quality of contact metrics and axis deviations:

 

 

The Dodgers’ turbo sinker group is interesting in that they generate less whiffs than their teammates with more traditional sinkers. This counters the tendencies of the league as a whole. But the Dodgers turbo sinkers are not like other turbo sinkers. As we can see, all eight of these pitchers have relatively large 2D axis deviations. This could be the main contributor to their overall effectiveness in limiting launch angle. It also seems to have an inverse effect on Whiff%, with the slower sinkers generating more whiffs than their turbo-throwing-teammates. This does not seem to be supported by league-wide data but throws yet another wrench into the Dodgers sinker mystery.

I set out to find what makes sinkers effective. I wish I could say that high-velocity sinkers were better, or that harnessing the power of SSW could help increase sinker whiff rates, but it is not that simple. We cannot for certain say any one factor, or even group of factors, are guaranteed to work — even if a particular combination works for a particular pitcher. What we can say is that that you can’t expect more whiffs simply based on velocity, and that judging a sinker based on whiff rate, even in a fantasy baseball landscape, misses the point entirely.

Seam-Shifted Wake has become a hot topic in some baseball circles. Rightfully so, it’s really cool! Even with these (non)findings, I’m inclined to believe that that harnessing the effects of SSW is a positive. But let’s not jump to any conclusions just yet. Does SSW actually make a sinker more effective? That is still unclear. My hope is that with further research, a better understanding of the forces at work, and public access to spin axis data we are merely scratching the surface.

Featured Image by Doug Carlin (@Bdougals on Twitter)