Here’s this latest edition of Keeping up with Baseball Research (7/18/2018). I’ve been spending a great deal of time looking into command, motor learning, and trying to build a model for how it all fits together. This is obviously a long term project, but something that is of interest to me at the moment.

The following are just a few studies to get you thinking about some of the fascinating concepts relating to commanding a strike zone. Enjoy!

Relevant Baseball Research: July 18th, 2018

Throwing Speed and Accuracy in Baseball and Cricket Players – Freeston J, Rooney, K. 2014. 

Dr. Jonathan Freeston, who works with the Cleveland Indians, has done some interesting research looking at how throwing speed affects accuracy in baseball players – as highlighted twice in this post!

The first one here, investigates the speed-accuracy trade off (SATO) in two different studies.

SATO, essentially posits that accuracy in throwing is optimized at a certain percentage of maximum velocity. In other words, as you reach peak velocity, there is likely a trade-off between MPH and hitting a target. Prior to this study, the research on SATO had shown that percentage to be approximately 75% of maximal throwing speed (MTS).

These two studies by Freeston and Rooney were created to look at this SATO effect in baseball & cricket players.

Players were asked to throw at a specified intensity level towards a target (a cricket stump in study 1, a target just above the ground in study 2), in order to identify whether there was a SATO and at percentage of max velocity that occurred.

The authors found a few interesting findings:

• There was a significant trade-off between throwing speed and accuracy, such that total error was reduced at 80% of MTS compared with 100% of MTS in trying to hit a cricket stump. In the second study, accuracy was optimized at 70%. Fairly similar findings to previous research.
• Baseball players were more accurate at hitting the cricket stump than the cricket players – suggestive that there is a transferability of practicing hitting a target to less familiar tasks… and that these particular cricket players may not be a skilled equivalent population. 
• There was a significantly reduced error in the horizontal direction. We will see this corroborated in following studies. 
• “Important to note that the strength of the relationship between speed and accuracy was, for the most part, relatively low, suggesting that speed is just one factor that contributes to throwing accuracy.” Sounds similar to the effect that velocity has on the variance of successful performance measured by Whiteside. 

The authors provided two potential contributing factors to the fact that there is a certain speed where accuracy is optimized – higher throwing speeds creating increases in lateral trunk tilt and a reduced launch window.

What is a launch window? Glad you asked!

The Launch Window Hypothesis and the Speed-Accuracy Tradeoff in Baseball  – Freeston J, Ferdinands RE, Rooney K. Percept Mot Skills. 2015 Aug

“The Launch Window Hypothesis suggests that there is a finite time within which ball release must occur for an accurate throw to result (Calvin 83)”

The goal of this particular study was to quantify the SATO and launch window during a throwing task at two different speeds under non-constrained conditions. Most of the previous research was done with constrained conditions – seated or standing without moving feet. The errors in accuracy in these studies were attributed to two main factors:

• Hand-path errors – joint rotation of the hand
• Release errors – inappropriate timing of finger flexion leading to premature or late ball release

Of the two, release errors were shown to be more detrimental to accuracy.

The study took a look at accuracy and the launch window for two conditions, throwing 80% and 100% MTS. Here are some of the findings.

• Players exhibited higher vertical error than horizontal error. Again, more to come.
• The launch window, the range of time in which the ball release results in a successful hit of the target, was significantly smaller during maximal speed than 80% of maximal speed.
• “Every millisecond that elapsed between optimal release and actual release during the 80% speed condition created an accuracy of 0.61m” Think about that…for every millisecond that a player was off from ‘optimal release’, there was an accuracy error of almost 2 feet. Throwing strikes is incredibly hard. 
• Speed accuracy trade-off resulted from inappropriate timing of ball release rather than from errors in hand-paths.

Pitching Form Determines Probabilistic Structure of Errors in Pitch Location  – Masahiro Shinya, Shinji Tsuchiya, Yousuke Yamada, Kimitaka Nakazawa, Kazutoshi Kudo & Shingo Oda. Journal of Sports Sciences, DOI: 10.1080/02640414.2016.1258484

We know that pitchers are going to miss spots (especially when they’re trying to throw max effort!). With every motor action there will be motor error, even more so in a ridiculously difficult movement task like pitching. With each movement, there is what’s called a ‘probabilistic structure’ to these errors, in that there is a shape to it. Knowing this structure, and putting it into the context of pitching, is what the authors decided to look at.

Think about a batter’s hot zone. If a player is weak ‘down and away’, we’d probably want to attack it.

Consider two pitchers facing the same batter, both trying to execute a down and away fastball. Even if they had the same error size in execution, the shape of the error could be a significant factor in their outcome. If you’re ‘error shape’ includes missing over the heart of the plate, that’s a problem.

This begs the question – what determines probabilistic structure of throwing error?

The goal of the study was to investigate the differences in two-dimensional probabilistic structure of errors in pitching location for the different combinations of pitchers; overhand, submarine, side-arm, handedness.

18 healthy college pitchers – six RHP overhand,  one RHP side-arm, four RHP submarine, five LHP overhand, one LHP sidearm, and one LHP submarine – threw 100 pitches to a target over two days and data was collected.

Interestingly, the authors found an ellipsoidal pitch distribution as seen above. What this means is that the error from the target was not random (a circle) but followed a pattern with long and short axes. The angle of error, moreover, was identical to the angle of arm trajectory. Look at the side-arm versus the overarm ellipse shapes. 

This ellipsoidal pattern is interesting, as it has potential implications (if backed further by future investigations) for pitch selection or strategy.

Baseball Pitching Accuracy: an Examination of Various Parameters when Evaluating Pitch Locations  -Kawamura K, Shinya M, Kobayashi H, Obata H, Kuwata M, Nakazawa K. Sports Biomech. 2017 Sep

“Research has put more time into investigating velocity than control, and therefore a pitcher who can throw at high velocity is more easily identified than a pitcher with exceptional control.”

The authors decided to look into the ellipsoidal structure of accuracy error farther, this time investigating pitching accuracy using three quantitative measures applied to two groups of pitchers: pro and high school. Small sample size – 5 pros and 8 high schoolers. 

Each pitcher threw three 10 pitch sessions;  one session all fastballs, the other two sessions five fastballs and five curveballs prescribed randomly.

What did they find?

Not so shockingly, pitch location (PL) results demonstrated that “professional pitchers have better control in all three evaluation methods.”

Pitch location was distributed in roughly ellipse-like shapes as previously seen in the research (Shinya 2017). The lengths of major and minor axes are related to variability of release timing of the pitch and release point in space, as was shown in previous research by (Shinya 2017).

“Professional pitchers were shown to have remarkable abilities to control the PL, particularly in the lateral direction”. In other words, this study showed that professional pitchers out performed their high school counterparts in controlling for the horizontal plane. 

“Notably, the mean position of PL in breaking balls was approximately 30cm lower than in fastballs” – even though they were asked to throw the ball at the catcher’s mitt. This begs a few interesting questions to think about.

Are pitchers conditioned to throw curveballs down in the zone, meaning that even though they are ‘concentrating’ on hitting the mitt, there’s an automaticity of the task being ‘throw it down in the zone.’?

When a pitcher is asked to throw a curveball at the mitt, is the task ‘achieve an end-point of the ball’ with a specific pitch, or is it ‘throw the ball at the mitt, and let the break do it’s work?’

The utility of this information, to be honest, is something that I am still considering. At this time, I think that more work needs to be done before we can make definitive statements like right handed throwers are better off trying to throw down and in to a righty, because they are more likely to miss out of the zone. Though the idea seems nice! 

More likely, I need to do some more research on command / control patterns, break / movement per pitch, and how it can fit together with this information. If anyone has more information on this, or wants to guide me down a rabbit hole, I’m all ears!!!