The junkbox is full of mystery

Occasionally I take a look at the Kirby Index Leaderboards. In August, the expected names that populated the top of the rankings have: Kevin Gamman” Trevor RogersAnd Jacob Vegrom. Likewise, the bottom contained stereotypically wild Hurlers: Joe Boyle” Luis Gil” Freddy PeraltaAnd… Janson Junk???

If you are familiar with Junk, it is probably due to its excellent command: among all starters with at least 70 innings threw this season, 3.0% running speed are rank as the lowest. (In an excellent interview with David Laurila last month, Junk spoke about training his command during the winter.) Junk in last place on this classification was like spotting one polar bear in Arlington, Texas – In other words, a sign that something was seriously wrong.

Due to all four components of the Kirby index, Junk was poor. But his vertical release point was particularly inconsistent and was the last death between pitchers in the monster. In my article last year I introduced the Kirby index, to which I linked above, discovered that the ultimate location of the field is almost completely determined by release corners and release points. It follows that pitchers with inconsistent release pointed show a bad command. How did Junk varied so often and still threw so many strikes?

Paul Scenes Is not a similar pitcher, but he gave a potential answer. Skenes appeared at the bottom of the original Kirby -Index leaders because he changed his position on the rubber, depending on the skill of the batter. In the end that was an easy solution: by looking at fastballs thrown at right -handed batters and left -handed batters individually and on average the results, the rubber shifters could be compared with the rest of the pitching population.

However, Junk did not switch around the rubber, and even if he was, it would not change his vertical release point. But adjusting his arm angle would do that. Arm Angle was not publicly available when I first wrote about the Kirby index, but it appears that for Junk it explains the significant variation in his vertical delivery point and his vertical release corners.

From one pitch to the next, junk can switch from a release of a milquetoast three -quarters …

… to one of the more extreme exaggerated arm angles in the competition, somewhere in the Jona Tong/Trey YESAVAGE zone.

Here is still three -quarters:

And the exaggerated one:

Throwed among starters with at least 300 fastballs, only Antonio without And Logan Allen Contained more variation in Armhoek on their four-seater Fastballs.

Although I did not have access to the variation of the arm angle when I wrote that Kirby index article, this funky junkie -feit motivated me to see if this at all related to the possibility of recommending the ball. In short: apparently, not at all!

To test this, I used our two familiar models, location+ and pitchingbot’s botcmd, as proxies for command skills. I used the standard deviation of the four-seater corner of a pitcher as the explanatory variable; For each respective model I used both the overall figure (location+ and Botcmd) and their Fastball-specific models (Botcmd FA, LOC+ FA) as a response variables. I limited the monster to pitchers with at least 60 innings and thrown 300 fastballs. In all four combinations of these explanatory and response variables there was hardly any relationship.

Ok, maybe if you squeeze, something is going on with pitchingbot’s fastball command model. But in general there is really nothing.

If arm angle variation does not explain the assignment, it must explain something, right? After all, there is a strong and proven relationship between arm angle and induced vertical break:

Justin Willard, the Red SOX director of Pitching, resulted in a relevant quote AthleticsS Long article about induced vertical break Published last week.

“Arm -slot is the highest correlated thing to turn direction,” said Willard. “Spin direction is the highest correlated thing with Magnus Lift.”

But something else happens here that is a bit surprising. When Junk’s four-seater balls are released against their induced vertical break, it becomes clear that the relationship specifically disappears for him. Whether Junk is throwing exaggerated or from a standard three -quarter position, the amount of vertical movement he gets on his fastball hardly varies.

It appears that this phenomenon – although not so true for others as mess – applies in different cases. Although the R-Kwadraat between Armhoek and IVB at the pitch level is 0.34, the average R-squared within pitchers itself is 0.05.

This felt like another mystery that needed an explanation. How can the relationship at the wider level between IVB and Armhoek be so strong, but still explain almost nothing at the pitching level? I brought this question to my pitching friend Stephen Sutton-Brown Baseball Prospectuswho gave a compelling explanation. At individual pitching level, arm angle and speed remain (relatively) constant, and so the variation on a given fastball is explained more by other factors, such as hand positioning or grip. At the population level of the pitcher, the variation of the arm angle is of course much wider.

But Junk is a special case. He does have a considerable amount of arm angle variation, so I expected that part of his Fastball movement would be explained by the changes in his arm position. It wasn’t. Junk was one of the few pitchers with a zero relationship between his arm angle and fastball vertical movement.

What’s going on with this guy? I wanted to ask him. Unfortunately the Marlins did not have the same desire; They refused to make him available for an interview. For now, the mysteries of Junk remain out of reach.