When I am not watching, thinking about, writing about, or attempting to analyze baseball, I am an applied cognitive psychologist. My academic pursuits keep me in tune with exciting findings from the scientific community. It is rare that my two interests directly interact, but it happened this week. A friend forwarded me an interesting article in which neuropsychologists examined the impact of a visual-brain training game on the vision and on-field performance of university baseball players.
Synopsis of the Brain Training Research
The article (Deveau, Ozer, & Seitz, 2014), published in Current Biology, assessed the effects of a perceptual-learning-brain-training program on the University of California Riverside (UCR) baseball team. A limitation of almost all brain-training games (i.e., Lumosity) is that improvements within the game do not transfer beyond the task being trained. You might get good at the memory games you are playing but this does not lead to gains in everyday use of memory and intelligence. The games are therefore limited in their utility. The authors of this recently published work attempted to overcome this issue by developing an integrated training program that combined multiple perceptual learning approaches that would be powerful enough to generalize to playing baseball (and other real world tasks).
UCR position players participated in the vision training, while the pitchers served as the untrained control group. This is not an ideal experimental design as it involves comparing something (i.e., the training intervention) to nothing. It would have been better to include an intermediate control group wherein some players engaged in a different visual-brain training task, or some other task (e.g., memory training). The employed design works as a first step, however, attributing any effects to the specific training program used in this study should be tempered.
Training consisted of 30, 25-minute sessions over an 8-week period with an average of 4 sessions per week. The training game required participants to locate Gabor targets on a computer screen as quickly as possible. Gabor targets look something like this, and can be adjusted across a number of variables to make detection more or less difficult. In the training game experienced by the UCR baseball players, the targets were presented amongst distractors (i.e., non-Gabor images) that became more and more similar to targets as training progressed.
Researchers recorded visual acuity (with typical Snellen eye-exam-charts) before and after training for all players. Some fascinating results were observed. Prior to any training all of the players had excellent vision (Position players [Training group] = 20/13; Pitchers [Control group] = 20/16). So, surprise! Having good vision might be critical for playing baseball at high levels of competition. While all the players had better than ‘normal' vision (i.e., 20/20) to begin with, only the training group exhibited significant improvement in their visual acuity after training (average acuity: 20/10); the control group did not show any change (average acuity still: 20/16). The trend of improved visual performance for the training group was also observed on two other measures. The take home message is that it seems as though the training program worked to improve the way the brain processes complex visual information.
Now, the critical question: did the training transfer to baseball performance? To examine this, the researchers looked at batting statistics for the players in the training group before (2012 season) and after (2013 season) the training. They focused on strikeout rate (K%) and Bill James' runs created (RC) metric. The logic for choosing these metrics was that seeing the ball and identifying good pitches to hit would require good vision, and one could expect that improved vision would lead to more positive outcomes. They found that K% for the trained UCR players decreased from 22.1% in 2012 to 17.7% in 2013. To ensure this improvement was not just a function of year-to-year improvement typical of college players, the researchers established a group of players from the conference who had played in both seasons as a control group. For this group K% decreased only 0.6% across seasons; not nearly the same decrease as the players at UCR.
To account for playing time in their examination of RC the researchers used RC/Out. They observed that the UCR players went from 0.140 RC/Out in 2012 to 0.188 RC/Out in 2013. This 0.048 difference was significantly greater than would be expected from year-to-year improvement, as the conference control group only improved 0.011 (0.169 to 0.180). Coinciding with this increase in RC the UCR players also showed greater than expected improvement in each component of their slash-line (AVG/OBP/SLG). It is interesting that they did not examine walk rate (BB%). Perhaps there were reasons for this but they are not stated within the article.
Taken together, the results of this study are quite compelling. The integrated visual training program developed by the researchers had clear, positive effects on players' visual acuity and transferred to their baseball performance. Follow-up research will need to establish how long the effects last, and the extent to which ongoing maintenance is required. Despite this, simple training programs like that used in this study (critically, those that show transfer to real world tasks) may be something that a major league team could implement for its players in order to exploit this potential advantage.
Are there certain players that should consider contacting these researchers?
If we accept that the training program used in the research described above is effective - which it seems to be, although replication is certainly required - are there certain players in major league baseball who could use some training? To attempt to answer this question, I looked at PITCHf/x plate discipline measures for the 2013 season for players with at least 350 plate appearances. Using these data I created a rudimentary plate discipline score by examining how far players fell from the mean on four measures of plate discipline: swing percentage on pitches outside the strike zone (O-Swing%), swing percentage on pitches inside the strike zone (Z-Swing%), swing rate (Swing%), and contact rate when swinging the bat (Contact%).
If a player's performance fell 1.5 standard deviations from the mean in the direction of negative performance, they were given a score of 1; otherwise a 0. For the measures used, this meant falling above the mean on O-Swing% and Swing%, or below the mean on Z-Swing% and Contact%, earned you a 1. Essentially, is the player swinging a lot more than average? Swinging at pitches that would be called balls? Not swinging at pitches that would be called strikes? Not making a lot of contact when swinging? If the answer is yes to a few of these, it can be reasonably argued that they are not demonstrating good plate discipline. Summing across the measures gives a player a score from 0-4 that provides a rough idea of their plate discipline package.
For 2013, 19 players had a score of 2 or greater. Here they are with the relevant statistics and their problem areas (i.e., statistics deviating from the mean by 1.5 standard deviations) identified:
|Josh Hamilton||0.397||0.773||0.552||0.706||O-Swing%, Swing%, Contact%|
|Yasiel Puig||0.370||0.756||0.532||0.677||Swing%, Contact%|
|Wilin Rosario||0.377||0.707||0.537||0.729||O-Swing%, Swing%|
|Dayan Viciedo||0.381||0.722||0.544||0.769||O-Swing%, Swing%|
|Chris Johnson||0.384||0.688||0.534||0.757||O-Swing%, Swing%|
|Oswaldo Arcia||0.386||0.646||0.507||0.686||O-Swing%, Contact%|
|Marlon Byrd||0.388||0.714||0.548||0.724||O-Swing%, Swing%|
|Brandon Phillips||0.388||0.698||0.534||0.794||O-Swing%, Swing%|
|Alexei Ramirez||0.391||0.693||0.544||0.882||O-Swing%, Swing%|
|J.P. Arencibia||0.395||0.678||0.529||0.715||O-Swing%, Swing%|
|Delmon Young||0.395||0.768||0.571||0.726||O-Swing%, Swing%|
|Torii Hunter||0.400||0.724||0.550||0.790||O-Swing%, Swing%|
|Howie Kendrick||0.403||0.681||0.534||0.801||O-Swing%, Swing%|
|Alfonso Soriano||0.409||0.693||0.533||0.718||O-Swing%, Swing%|
|Nolan Arenado||0.41||0.716||0.556||0.818||O-Swing%, Swing%|
|Evan Gattis||0.421||0.716||0.553||0.776||O-Swing%, Swing%|
|Pablo Sandoval||0.424||0.748||0.556||0.833||O-Swing%, Swing%|
|Adam Jones||0.448||0.751||0.581||0.741||O-Swing%, Swing%|
|A.J. Pierzynski||0.476||0.752||0.602||0.83||O-Swing%, Swing%|
Many of the names on this list are often thought of as free-swingers (with corresponding low BB%), so it seems this simple approach is picking up on something. Perhaps these players are having difficulty seeing the ball and could benefit from the visual training used with the UCR baseball players.
The results of the scientific study could have real impact on the game of baseball and how players train. Along with time in the gym, flexing the brain muscle responsible for visual processing may lead to significant on-field improvements. For now, it is a single experiment that must be replicated before a team should realistically consider implementing this for its players. Moreover, being the first team to experiment with such training is risky. Occupying players' time with this training (even the short time frames cited in the article) means less time on other tasks (e.g., weight training, taking ground balls). If the training does not have positive results, this is all time lost. On the other hand, if positive results are observed, the time frame for taking advantage of such results will be limited. It certainly would not be long before other teams simply copy the innovation and reap similar rewards. For those with a Baseball Prospectus subscription, I strongly recommend Russell Carleton's article that outlines a similar argument for why teams have not implemented nutrition programs. This risk-reward tradeoff of such programs (nutrition, brain-training or otherwise) is likely untenable for most teams. However, it might still be worth it for individual players - perhaps those in the table above - to incorporate the training into their personal training regimen.
All statistics courtesy of FanGraphs. Big thanks to my friend Adam Reddon for passing along the primary research article. A short video about the visual-brain training research at UCR can be seen here.
Chris Teeter is a Contributor to Beyond the Box Score. You can follow him on Twitter at @c_mcgeets.