Updated: Mar 7, 2021
Last week I was thinking what to write about and I was struggling to come up with ideas. I started combining through datasets, when I stumbled upon something interesting looking through pitch tracking data occasionally mentioned (including by myself in my previous blog on sliders), but not often in much depth. That thing being the change in spin efficiency during ball flight.
Today’s piece is going to be a little different and probably much more informal than most you’ve read on the channel so far. This topic is something I don’t really know a ton about right now and rather than claiming that I have all the answers, I’m going to take you guys through how my brain works when I go about discovering relationships and finding a better understanding of whatever it is I’m attempting to look into.
In my previous blog, I mentioned the way Gyroballs, or gyro-sliders, get “late break” is by an increase in spin efficiency as it approaches the zone. I did not really know how much spin efficiency was gained, when, or how, I could just see it was happening. Well, I was able to get my hands on a dataset that includes spin efficiency at about 40 feet and then again around 20 feet. So, we know Sliders have a change in spin efficiency as they approach the plate (or at least we think they do), but what about other pitches in this dataset? Here we go.
The first thing I wanted to do with the new dataset was compare the difference between the Spin Efficiency at 40ft to the Spin Efficiency at 20ft on a pitch by pitch level. I did this by simply subtracting the Spin Efficiency at 20ft from the Spin Efficiency at 40ft. If the number at 40 was larger, then I knew that there was a gain in efficiency. Super complex math huh.
Above is a density plot of the difference in efficiency for all 17,000 pitches in the dataset. The vertical lines represent the average change in efficiency for each pitch type and the taller the curve is the more frequently that value occurred. For example, on Changeups, we can see that on average there is roughly a 8% decrease in Spin Efficiency and that is fairly common. However, it is also somewhat common to have about a 15-20% decrease in efficiency as seen by the little hump down to the left.
Breaking down each pitch type:
Two-Seams almost always lose efficiency
Sliders sometimes gain efficiency, but the can also lose efficiency as well
Four-Seam fastballs have a wide range of possibilities ranging from losing 12% to gaining 12% efficiency. Weird.
Cutters are a bit unpredictable as well ranging from -5% up to a 16% increase.
Curveballs almost always gain efficiency and some of them a lot, up to 20%
Changeups almost always lose efficiency ranging from -20% to gaining 2%
Why is this happening? Why are some pitches gaining efficiency and some losing a ton as they approach the zone? How does this happen? As I previously explained, when the ball is released there are two vectors at play, the velocity vector and the spin vector. The velocity vector is initially pointed in a straight line right at your target. As the ball travels, gravity acts on the ball and that line actually starts to bend down a little bit rather than be perfectly straight. The spin vector is in a direction perpendicular to the spin direction of the baseball. In the example of a gyro ball, it is also pointed directly at your target. However, the spin vector, to my current understanding, doesn’t change it’s orientation.
So, as the ball starts to fall (gravity starts to affect it) the spin vector and the velocity vector are no longer perfectly aligned. Because of this, some of that gyrospin now becomes transverse spin (spin that affects ball flight).
This is how Spin Efficiency is gained throughout ball flight. What I haven’t quite wrapped my head around is how it is lost. I would imagine that it is through the same concept, but I’ll think about that and get back to you guys on it at another time.
So we saw above that the Spin Efficiency on Sliders usually increases, but occasionally decreases. Given that my level of understanding of changing spin efficiency is best geared towards Sliders at the moment, I wanted to dive into this pitch a little deeper to answer two main questions:
How do you gain vs lose spin efficiency?
Can you control how much spin efficiency you gain or lose?
To tackle the first question I stared at 50 column .csvs for hours trying to recognize any differences between those pitches that gained efficiency vs those that lost it. Then, I ran correlation matrices to see if anything had a strong linear relationship with the change in efficiency and that was sort of interesting. There were moderate to strong linear correlations between the change in efficiency and Spin Rate, Horizontal Acceleration, and Vertical Acceleration. I thought for awhile on why those specifically would have such a large impact on the change in efficiency and I haven’t been able to draw any solid conclusions yet.
Moving on, I bucketed the initial Spin Efficiency of each pitch into 10% intervals hoping to see any glaring differences in the changes occurring.
Unfortunately, I don’t see a ton of glaring differences in the bins other than the 50-60 range. Most are centered right around a 11% increase and have a few outliers that lose efficiency. There is an exception here with a weird double peak (or bimodal distribution for you statistics folk) in the 20-30% bin. I’m not sure if this is noteworthy or not, so at the moment we are going to ignore it but I’ll look into it later.
Like most, I’m drawn to outliers. I looked at all the pitches that lost efficiency even though the majority were gaining, hoping to have a better understanding of why this was happening. Thankfully, and shockingly, it was pretty clear. Those pitches that lose efficiency are the “backed up” reps, or pitches that break slightly arm side instead of glove side. This makes sense as to why the 50-60 bucket has so many as those are SL/CT that typically sit right around -1 to 1 Horizontal Break. Those reps that get gloveside movement gain efficiency and late movement while those reps that don’t get gloveside movement lose efficiency and fall down a little bit.
So, there’s our answer to question #1. How do you gain vs lose efficiency? Sliders with gloveside movement gain efficiency while pitches with arm side movement (the backed up reps) lose efficiency. This makes intuitive sense as Fastballs also move armside and typically lose efficiency quite often as we saw above.
Now question #2, can you control how much efficiency you gain/lose? At the moment, I have no idea. The more I looked into it, the more confused I got. My hypothesis would be that you can via location of the pitch. Throwing the pitch down and away gives it more time to break and more time to gain efficiency. This would make sense as to why horizontal acceleration is so strongly correlated with the change in efficiency. But, at the moment that’s just a guess. I plan to analyze the change in efficiency by location a bit more and also compare similar pitches to see why one may gain 5% and why one may gain 12%.
Food For Thought
Completely independently of the research for this blog, myself and the other SS Blog contributors were combing through video looking at Sliders just for fun. What we stumbled upon might just be in line with what we are talking about today.
While watching Edgertronic clips there was a pitcher throwing sliders with a large black sharpie dot drawn on the back of the ball. This is fairly common, I’m sure many of you have seen this before. However, what I hadn’t seen before was that the rotation pattern of the dot changed significantly throughout ball flight. It started off very loose and wobbly and then became tighter and more centered on the ball as it approached home plate. We were fully mind blown.
Is this caused by the increase in Spin Efficiency? Does the Spin Axis actually shift during ball flight? Maybe this explains why some have the “red dot” on their slider while others don’t.
If anyone has any thoughts on this or video of their own similar to what we saw I’d love to discuss. Shoot me a DM on twitter.
Today’s blog was a bit different in nature as it was much more exploratory with a lot of “I don’t knows” and unanswered questions. I wanted to take you guys through my process of discovery and how I figure things out. My next couple blogs will be similar to this one as we work through this discovery together to determine why/how Spin Efficiency changes.
Some things to chew on between now and next time:
Spin Efficiency is not constant throughout ball flight, it often changes which increases movement laterally or resists gravity less.
Is it possible to influence how much spin efficiency you lose or gain via location? Why would Horizontal and Vertical Acceleration (not velocity, acceleration) be so strongly correlated to the change in efficiency for Sliders.
Does the Spin Direction actually remain stable throughout ball flight? Or do the seams have the ability to alter it throughout?
Want to know more? David Kagan wrote a great piece on the physics of the gyro pitch for the Hardball times back in 2017 and I used some of his pictures for this piece. Check it out. The Physics of the Gyro Pitch | The Hardball Times (fangraphs.com)