My 14-year-old son plays baseball, and his fastball is about 90% efficient. So naturally, I was wondering if the baseballs that they use, which have very large seams compared to an MLB ball, move similarly to MLB balls. In my previous post, a Wilson 1030 baseball was found to move less that an MLB ball when in a Looper orientation. Specifically, the Magnus movement seemed to be reduced. This was based only on 2 shots, but it left me wondering if Seam Shifted Wake (SSW) could kill Magnus on large-seamed baseballs.
Here’s the setup. We have a little less that 55′ to work with here, but I don’t have to bother anyone else to use this space. I used the High Frame Rate option on this Sony camera to acquire video at 480 fps (it does 960). The Washington State cannon is used to throw the 2 balls on 3 different axes in a few different orientations described below.
For each shot, I give my Matlab script a few pieces of information:
- The initial position of the ball at release
- Its speed
- The location of the ground at the net and the horizontal arrival location of a ball with no side spin (if I was smarter, I would have made a mark on the door)
- The height and distance to the right of release of the camera
- The number of pixels/foot at the full distance (based on the cal target)
The script uses a cross-correlation algorithm (XCORR2) to locate the ball in each frame. We assume the ball has a constant drag coefficient (0.33) and moves away from the cannon accordingly.
This produces a full x, y ,z ,t trajectory of the ball. In other words, this is a poor-man’s Hawkeye. Part of my motivation here is to convince smart, poor pitchers to use this setup to diagnose their SSW pitches. Although we don’t have it yet and my video would need work since you can’t see the seams at release, in principle, the orientation of the ball could also be obtained from a shot like this, and then you’d have something BETTER than Hawkeye.
Here are the two baseballs I used and I have marked the axis location for the looper pitches with a blue dot. The black dot on the Wilson ball (right) is the location I used for the looper in post 63.
All the Wilson pitches were repeated 5 times, and the MLB balls were repeated 3 times (I was more sure of what they would do and needed to keep the time involved here reasonable). I made 56 shots total. After each one, I had collect the ball, write the next shot number on the whiteboard, record notes about the shot and reload the cannon, all before the camera timed out. It was a bit of a workout.
Here’s some results: the MLB ball x, y trajectories. In all of these, I use the typical baseball coordinate system which has the distance from the camera as y. Actually, to be a right handed system, the numbers should go the other way, but sorry. Negative x values are heading into the right hand batters box and visa versa. For each pitch, the pitch number is listed first (ignore that) followed by the type of ball (MLB or Wilson) and the seam orientation (4S, or a Loop on top, bottom, left or right). This is followed by the spin axis. The 4S orientations are used to ensure seams have no effect. Generally, we expect the loopers to move away from the loop (so a right loop should move in the positive x direction).
Pitch 42MLB4S180 should have no left right movement, and although it is a bit buried on the plot, it does not and moves perfectly straight. Pitch 40 should and does move left due to Magnus effect. Pitch 38 should move similarly to the right, but it appears something is up. Let me tell you what happened.
The camera sits about 12″ to the right of the release point, and this is a critical number to know for this analysis. But, it turns out that the release point changes a bit when I switch from a 90º axis to 270º. Let me explain:
In addition to the tip causing an offset in the initial position, it was also hard to ensure the camera stayed fixed. Unfortunately, I have been unable to figure out how to remotely trigger HFR video on this camera, so I have to press a button on it each shot. Also, when the battery goes dead, I have to remove it from the mount to replace it. Bottom line, please excuse the small trajectory errors over the first few feet.
Going back to the plot, one can see that the loopers are disappointing. There is very little horizontal movement. Let’s look at the vertical.
Use the 42 pitch as a baseline. That is a 12:00 fastball. The two side-spin pitches (40 and 42) drop considerably compared to that. Not that the loopers are above and below the baseline pitch as reported in the original study. We still do not have an explanation for this.
Now here are similar pitches with the Wilson ball.
Overall, the result is pretty similar, although the looper pitches seem to work a bit better for this ball.
So far, I’ve only I have not show any top or bottom loop pitches. These are the ones I am here to check since the previous test seemed to show that Magnus movement was inhibited for these pitches on the Wilson ball. Here are the MLB ball results.
And the Wilson ball results.
These all match the original looper test, at least qualitatively. A top loop moves down compared to a 4-seam. It loses a small amount of Magnus movement (compare to the 4S270 case). A bottom loop moves up and increases its Magnus movement. And all of this seems equally true for both baseballs.
SO, kids, go ahead and try to throw seam shifted pitches with your high-seam baseballs.
Disclaimer: I don’t advocate trying to throw loopers. As the results above show, they are sensitive to orientation and don’t move as much as other SSW pitches. We test loopers in the lab because our equipment cannot throw more effective SSW pitches.
Lastly, don’t stand in front of the loaded cannon.