SPECIAL CONTENTS Training the“Athletic Brain Article 10 (Part 1 of 2) | Dec. 1, 2017

Zeroing in on the Secrets of Top Athletes (Part 1)

Do top athletes excel at their ability to predict?

The Sports Brain Science Project is currently carrying out a variety of experiments and measurements at its “Smart Bullpen” in Atsugi City, Kanagawa Prefecture. Research Associate Dr. Daiki Nasu is involved in measuring top athletes under conditions close to actual games. The experiment on “batters’ predictions,” featuring a player from the Japan Softball League, reveals some of the conditions of top athletes.

Measuring the movements of players in close-to-reality “Smart Bullpen”

—What is your role in the Sports Brain Science Project?

Nasu: I mainly measure the movements of batters and pitchers in the Smart Bullpen here in Atsugi and analyze the data. For example, I have players wear bodysuits with inertial sensors and hold bats with inertial sensors attached to the grip end of the bat. They then swing at pitches. I measure which parts of their bodies move and the speed at which they move, as well as how fast they swing the bat.

Kashino: One of the greatest advantages of the Smart Bullpen is that it is a controlled environment while at the same time providing conditions similar to an actual game. At present, the mainstream method of measuring body movements is to use optical systems. This means using cameras to sense and capture movements. As a result, multiple cameras must be installed around the batter and pitcher, and players can’t hit the balls because the cameras interfere. What we want to know, as thoroughly as possible, are the interactions between the pitcher and batter when they face off against each other. So we are using inertial sensors to capture movements as-is under actual play conditions.

Nasu: Meanwhile, there are five mixed cameras and two mobile cameras in the Smart Bullpen, so it is possible to watch the pitcher and batter from a variety of angles on a large screen. All the athletes who come here to be measured watch these images and video and use them for feedback.

Kashino: Because the timestamps of all sensing information are synchronized, it is possible to observe both the pitcher and batter in their face-off. This gives us valuable data we normally can’t obtain.

Exploring when batters figure out the type of pitches

—What kind of experiment are you specifically conducting?

Nasu: Whether it is baseball or softball, it takes a mere 0.5 seconds for the ball to reach the batter from the time it leaves the pitcher’s hand. The batter must determine the type of pitch and hit the ball in this short fraction of time. We are trying to capture the player’s judgment, body movements, and responses during this time. In short, we are focusing on unconscious movements and responses.

For example, in March 2017 we conducted an experiment to observe the response timing of batters belonging to Japanese professional leagues, including Eri Yamada (Hitachi Sun Divas), one of the best batters on the Japan national women’s softball team. For the experiment, pitchers randomly threw two types of pitches, a straight fastball and a changeup. If the pitch was a strike, the batter was to hit it. If it was a ball, the batter should ignore it. We measured the movements of body parts in the batters’ responses, and obtained extremely interesting results.

Kashino: For batters, the level of difficulty rises dramatically when changeups are mixed into the pitches. If they know in advance that only fastballs will be thrown, they can hit the ball by just timing their swings. But if they don’t know which pitch will come—a fastball or a breaking ball—then it is extremely difficult to the hit the ball just from timing.

—What is the time difference between a straight fastball and a changeup?

Nasu: About 0.15 seconds. The difference may seem slight, but it takes about 0.45 seconds for a straight fastball to reach the batter’s box, and 0.6 seconds for a changeup. If you think in terms of percent difference, then you will see that the difference is quite large. For batters, their temporal sense of these two types of pitches is quite different.

As Dr. Kashino said earlier, if players are told of the type of pitches beforehand, the top players can hit almost all of them. When we observe their body movements, we see that the way they hit is also different when it comes to pitches they know will come in advance. Randomizing the type of pitches is the heart of this experiment.

Batters who can hit act with timing in response to the type of pitches

Nasu: Figure 1 shows a comparison of data from Eri Yamada (left) and a young player (right).The images show the movements of the parts of body, such as hands, forearms, upper arms, trunk, hips, thighs, shins, and feet. Relative speed is indicated by the color gradient from blue (slowest), to green, yellow, orange, and red (fastest).

Figure 1. Distinguishing between skilled and unskilled batters

Nasu: The 0 in the graph indicates the moment the ball leaves the pitcher’s hand. When we see the response of Eri Yamada on the left, we see that she quickly moves her upper body at about the 0.4-second mark for a straight fastball. For a changeup, she quickly moves her entire body at just before the 0.6-second mark. Young batters, on the other hand, swing at both straight fastballs and changeups just slightly after the 0.5-second mark, and could not get a quick swing in.

—When we compare the results, we see that quality of the movements is also quite different.

Nasu: Right. The graph at the boom of Figure 1 shows the sum of the speed of each body part on a time series. From the results, we see that in the case of Eri Yamada, there is a time difference between the response peaks for straight fastballs and changeups. On the other hand, for the young player the graph shows a wide hill. This means that there is almost no difference between the timing of responses to straight fastballs and to changeups. In short, Eri Yamada has no wasted movements, and times her movements to effectively hit straight fastballs and changeups.

—What kind of players are the young player?

Nasu: She belongs to a team in Japan’s professional league, but she is a rookie who just graduated from high school. She could hit many home runs when she was in high school, and was a very good player. Incidentally, we received the cooperation of pro pitchers for this experiment. As expected, when facing off against top-league pitchers, inexperienced young player had a hard time adjusting their timing to their pitches.

Kashino: The level of the pitchers in this experiment is completely different from high school and grade school pitchers. Players who were successful at high school baseball cannot necessarily make it at the professional level. Eri Yamada was able to adjust her timing to their pitches and this showed us that there was a great difference between the performance of top athletes and rookies.

A batter has only 0.1 seconds to determine the pitch?!

Nasu: Figure 2 focuses only on the rotational speed of the hips. The graph shows overlapping data from each round of measurement.

Figure 2. Observing characteristics of batters from their hips’ rotational speed

Nasu: From the graphs on top we see that in the case of Eri Yamada, there is a clear divergence in the peaks of the graph lines for straight fastballs in blue and changeups in red. However, in the case of young player, the differences between the peaks are not distinct. What’s more, the graph shows that because Eri Yamada can’t predict the pitch at first, at about the 0.3-second mark, when she realizes the pitch is a changeup, she makes corrections every time so she can hit the pitch. The results show that top players excel in the ability to correct themselves.

The graphs on the bottom set time 0 as the moment the bat hits the ball (the moment of impact). With the impact as the baseline, we see here that the Eri Yamada’s timing of her swings is consistent.

Kashino: Eri Yamada quickly rotates her hips just before the impact. The power from her hips is transmitted to her arms, and this helps give her an ideal swing. And her timing is consistent each time. On the other hand, the young player is slow in rotating their hips when facing straight fastballs. This means she can’t hit the pitch with power.

Nasu: If you’re late in rotating your hips, you could swing and miss or foul the ball. If the pitch is a changeup, you could start your swing too early and mistime the pitch.

— In the case of Eri Yamada, she could clearly determine the type of pitches in about 0.3 seconds after the ball leaves the pitcher’s hand.

Nasu: The decision branching takes place at about the 0.3-second mark. It takes about 0.15 to 0.2 seconds for a person to obtain information from their eyes and reflect that information in their body movements. This means it takes at minimum 0.1 seconds after the ball is released to decide whether to swing the bat or not.

— It is surprising that she could tell the pitch in such a short time.

How do top players figure out the type of pitches?

Nasu: We next decided to study the timing of judging pitches. The experiment participants stood in the batter’s box, and instead of holding a bat, they held a button. We told them to press the button when they knew that the pitch thrown was a straight fastball. Eri Yamada had an accuracy score of about 80 percent. The young player had a score of about 50 percent. A score of 50 percent is no better than simply guessing. It means that, for the most part, the young player could not figure the type of pitches. However, there was no difference between the average time it look for each group to press the button. There was greater variability in the response times of the young player (Figure 3).

Figure 3. Discrimination of the type of pitches

— Is Eri Yamada’s score of 80 percent especially outstanding?

Nasu: Actually, we later measured a few dozen players. We found that Eri Yamada’s correct rate of 80 percent didn’t necessarily stand out. In the case of professional players, they got scores of about 80 to 90 percent in distinguishing between straight fastballs and changeups.

— I see. Is Eri Yamada predicting the pitch from the pitcher’s movement?

Nasu: When I asked Eri Yamada, she said, “I can’t predict the pitches from this particular pitcher’s movements.” It seems she was not conscious of the difference between straight fastballs and changeups just from looking at the pitcher’s form. Actually, when we measured the pitcher’s movements, we found almost the same movements for straight fastballs and changeups up to the -0.3 second mark (Figure 4). However, it may be possible that the batter unconsciously feels a difference in the pitcher's form (This is experimentally proved later).

Figure 4. Predicting from the pitcher’s form?

Outstanding pitchers don’t allow batters to read the type of pitches

Kashino: For most pitchers at the high school level, batters can tell the ball types?whether it is a straight fastball or a changeup?from the pitching form. Probably, the young player was able to hit well against that level of pitchers.

To put it another way, if the type of pitches a pitcher throws can be predicted from his form, it is highly likely that he will be hit, no matter how fast he throws the ball. How good or poor a pitcher is is largely determined by whether batters can predict his pitches, not simply by how fast or slow he can throw the ball.

In this sense, we can say that the pitcher Yu Darvish, who is currently playing in Major League Baseball in the U.S., is an extraordinary pitcher. Darvish can throw a mid-150 km/h straight fastball and about ten types of breaking balls. When we superimpose images of his pitching form, we see that his movements from the start until release are almost entirely the same. What’s more, the initial trajectories of his pitches are extremely similar. One of Darvish’s strength is that batters cannot read the course of his pitches until it is too late.

— Is it difficult to pitch so that the batter can’t read the type of pitches?

Nasu: It depends on the type of pitches, but it is quite difficult.

Kashino: Darvish and Koji Uehara can use different forms to throw the same pitch. They intentionally do things like changing when they lift their feet. This disrupts the batters’ expectations. For the rest of us, we just do our best to throw a pitch, so such tricks are utterly impossible for us (laughter).

Nasu: In our experiments, we learned that there is a big difference between regular and non-regular players in judging the course of the pitch. We are about to study further in detail what elements are useful in judging pitches.

(Text revised: April 17, 2018)

(Japanese text written by Madoka Tainaka)

Daiki Nasu, Ph.D.
Sports Brain Science Project, NTT Communication Science Laboratories

Next: Article 10 (Part 2 of 2) Zeroing in on the Secrets of Top Athletes
» Back to Articles


SPECIAL CONTENTS | Feb. 18, 2016

Article 1 (Part 1 of 2) Exploring how to control the mind and body

Research on ‘mind’ and ‘technique’ as a means to improvement in sports


SPECIAL CONTENTS | Feb. 18, 2016

Article 1 (Part 2 of 2)
Exploring how to control my mind and body

Brain activities we are unaware of control body movements and decision-making


SPECIAL CONTENTS | Apr. 26, 2017

Article 9 (Part 1 of 2) Understanding Athletes’ “Skill” and “State of Mind” with Virtual Reality Technology

Using highly-realistic VR for brain science measurements