SSW moves to MLB
March, 2020, I was invited to Blue Jays Spring training camp by Matt Buschman.
I met separately with the MLB side and the MiLB side. There was an MiLB coach who I didn’t get to talk to, but seemed especially interested based on the look on his face. I’ve since realized he looks at everything that way.
I thanked the Jays on twitter. Within hours, I was contacted by the Yankees, the Rays and the Pirates. I managed a trip to the Yankees camp and to the Trop to chat with two of their analysts.
On the way home from this trip, talk of Covid was escalating and the number of masks at the airport exceeded anything I’d ever seen. One week later, spring training was cancelled and we all hunkered down.
The pandemic was a terrible time, but ironically I think the SSW work benefitted from it.
March 9, 2020, I posted the scheme Trip Somers and I (mostly Trip) came up with to describe ball orientation with 2 angles. I wish I could remember when I first started exchanging ideas with Trip, but he’s been incredibly important to my progress on SSW. His spin visualizer (which I recommend) still uses top and front angles.
March 26, 2020, Andrew Smith launched UMBA, which was a baseball pitch trajectory simulator modeled after Alan Nathan’s famous spreadsheet.
MLB began using the HawkEye system in for the 2020 season. A front office guy shared a IA/OA plot with me that showed sinkers fell to the right of the curve. We both expected that. There was some evidence that 4S pitches were on the other side which made us both scratch our head.
In early April, I talked an MLB team into a zoom presentation of our stuff. They mostly yawned.
I received this DM from Matt Pilewski :
Matt put it together on April 19. I snuck a screen shot. Interesting cast of characters from PVS guys to Michael Augustine, some college coaches (Jimmy Jackson, Danny Borrell, Steve Nagy) and my future friends Harold Mozingo and Spencer Davis.
I put an offer on Twitter to do another zoom and Connor Hinchcliffe, then in the Phillie’s system, took me up on it. April 21, we had a public Zoom that had a major impact on my life. As luck would have it, the intense looking guy from the Jays played Little League ball with Jared Hughes and invited him that night. During the Q&A, Jared sent a video of one of his sinkers, and I had a new pitch to focus on.
I realize Hughe’s low-slot sinker is a much better example of SSW than the CH pitches I had been looking at. 50% of his sinkers fall more quickly than gravity. Skeptics were much less inclined to assume this was due to Statcast measurement issues that 10% of some changeups.
May 12, 2020, Andrew Smith releases UMBA 2.0, which includes SSW effects. Baseball Cloud expressed interest in the model and solicited a proposal from us, but they decided against funding that work.
Beth Woerner hosted my last public Zoom on June 18, 2020. This one had 100 participants. The video can still be viewed here and includes a lot of Q&A.
Jared and I stayed in touch. He bought a Rapsodo 2.0, built a mound in his back yard and started pitching in front of his Sony RX-10. We chatted frequently about his grips and the results. I totally miss in this process that he’s developed what has become termed a sweeper.
Very near the start of the shortened 2020 season, Hughes was awarded a contract with the Mets.
On September 28, 2020, I met Harry Pavlidis of Baseball Prospectus via Twitter. He had bounced around in my twitter feed for some time, but I was too ignorant about baseball intelligentsia to know who he was. It turned out we had many mutual friends/acquaintances. I was becoming more sure that SSW was an important effect in baseball and was making bolder claims than previously, which brought me to Harry’s attention. I was pointing out that on sinkers, there is a difference between the direction the ball spins and how it moves. Harry’s associates had been receiving HE data also and were not seeing this effect. Harry decided I was one of the myriad baseball crackpots out there (neither he nor I have much affinity for those people).
Harry called me out. I was taken back, not expecting this (I had not been met with much skepticism up to that point). Some of our mutual friends intervened and, eventually, Harry was able to determine that the information he’d been receiving was incorrect. We mended fences. I have an email that I never sent him that may have ended my progress on SSW. But this episode is a reminder of something I’ve encountered many times. If no one is giving you a hard time, it may be because what you’re doing is not that important or interesting.
The source of the confusion was due to attempts to calculate the spin axis of a pitch based on Trackman movement data. This calculation assumes that Magnus is the only force on the ball (other than gravity and drag). This same assumption, applied to Rapsodo spin data to find movement, had also been reinforcing the idea that Magnus was all that mattered. Many pitchers were also confused when Rapsodo would tell them that their 2S and 4S moved identically. Others were sure their 2S and 4S had unique spin axes, since Trackman said so. This incorrect assumption had the odd effect of reinforcing itself. It became self-fulfilling.
Harry invited Alan Nathan and myself to chew on HE data that his organization was in possession of. This effort lead to a publication on SSW in Baseball Prospectus on November 5, 2020. The article focused on the difference between the direction that Magnus would move the ball and the direction the ball actually went. It only dealt with 4S, Sinkers and Changeups. It showed that most CH and all Sinkers moved more glove side while 4S did the opposite. Driveline published an article on similar topics about one week earlier. The two works were independent and arrived at similar conclusions. Driveline went further to examine the outcomes of SSW pitches.
I concluded, incorrectly, that SSW had no role in breaking pitches. This was because what we now call sliders and sweepers were mixed in the data, completely washing out their effects. Harry taught me that the devil is in pitch categorization. There is a bit of a debate these days on the value of designating some sliders as “sweepers,” but without separating 4S and 2S sliders, which have seam effects in the opposite direction, one cannot see the effect of seams. We’ve done it all along on fastballs, so that required no new thinking. Now we know that the MOST SSW pitch is a sweeper.
It is very important to note that this was the first time that we were able to see the actual spin axis of pitches. One of the most striking features of the data was that for the majority of pitchers who threw both 2S and 4S fastballs, the two pitches spun essentially the same (spin axis, RPM, velocity, gyro angle). The only difference was seam orientation. This was counter to what nearly everyone seemed to believe at the time. Pitchers were often given cues to alter their 2S axis.
Meanwhile, I frequently discussed Jared’s pitching results with him along with HawkEye data he had access to from the Mets.
Andrew was wrapping up his Masters degree and John Garrett wanted to continue the work. But we needed funding. We developed a pitch to MLB teams and the league itself and delivered it (via Zoom) to 10 clubs and MLB. But we found no one interested paying for more work in this area. I realized that academic research and professional sports are not well aligned at all. John took a job driving the city bus and continue firing baseballs. Don’t worry, there’s a happy ending to that part of the story.