February 24, 2008
Pitching is both an art and a science, and from youth leagues to the big leagues, so is the challenge of keeping pitchers healthy. The National Pitching Association (NPA) is on the cutting edge of research and instruction on all three fronts, and many of their concepts are shared in their forthcoming book, Arm Action, Arm Path, and the Perfect Pitch: a Science-Based Guide to Pitching Health and Performance. David talked to the NPA's motion analysis coordinator and coach, Doug Thorburn.
Baseball Prospectus: The title of the new book includes the term "arm action." How do you define arm action?
Doug Thorburn: That is a question that I've asked many coaches and scouts, only to get several different answers. Some define arm action as everything that the arm does from the time the pitching hand separates from the glove to ball-release and follow-through, while others use it to describe a more specific piece of that sequence. You'll hear other terms such as 'arm path' and 'arm circle,' and some use these terms interchangeably with arm action, while others consider them distinct. Personally, I like to use arm path to describe the route that the pitching hand takes from glove break to the start of upper-body rotation, and arm action to describe the overall arm speed as the pitcher incorporates the rotational elements of the delivery into ball release.
But these definitions are by no means standard across baseball, and I try to avoid these terms because they cause so much confusion. This is a big problem with much of the vocabulary for pitching mechanics, as there are so few definitions that are standard throughout the industry. Two coaches might agree that a pitcher has good arm action, but disagree on what that means. So it seemed an appropriate book title, given that we challenge conventional wisdom throughout the book, and much of that conventional wisdom is rooted in this misunderstood vocabulary. Meanwhile, terms like arm action have been passed down through generations of coaches, and like the conventional wisdom of pitching mechanics, they have survived largely because their interpretations were never challenged.
BP: The terms 'command' and 'control' are often used interchangeably. Should they be?
DT: Command and control are two more examples of pitching vocabulary that can cause confusion. At the NPA, we have unique definitions for each. Command is used to describe the ability to consistently execute a certain pitch type, or as a 'command pitch' that a pitcher can trust to locate whenever necessary. Most developing pitchers start without a command pitch, until they can consistently harness a fastball. But by the time a player reaches the majors, he usually needs to have command of three different pitches to keep hitters off balance. Control is what we use to describe a pitcher's ability to locate a pitch at any given time, as stuff will vary throughout the season or during a game. A pitcher might have exceptional control of his curveball today, but when he is faced with a jam and runners on the corners, he might go to his command-pitch fastball that he knows he can locate low in the zone. Kevin Goldstein has offered other definitions of command and control, with control representing the ability to avoid walks, and command defined as locating pitches within the zone, and hitting specific targets. I think that Kevin's definitions are outstanding, and like the NPA versions, they describe two unique aspects of pitch execution. Ideally, we can use Kevin's definitions in conjunction with the NPA's to better describe the ability to locate a baseball.
BP: At a recent SABR conference in Boston, Bill James questioned the impact longer games have on a starter's innings and health, theorizing that it is harder to keep your arm warm for three hours than it is for two. Assuming the same pitch count, is there anything to that?
DT: I don't know of any research in this particular area, but there could be something to it. We are currently studying several different elements involved with pitcher fatigue, including workloads, mechanical efficiency, functional strength, and stamina. The duration of games could contribute as well, but it's difficult to study the issue until we have an objective measurement of fatigue. PAP (Pitcher Abuse Points) is a great start, but unfortunately the 100-pitch threshold that is inherent in the system is a bit misleading when applied with individual players, unless an adjustment is made to account for the pitcher's particular fatigue rate. For example, the PAP barrier for Pedro Martinez might be 90 pitches, and Randy Johnson's might be closer to 110.
Understanding the factors that lead to fatigue is crucial, as is the ability for a coach to identify when fatigue sets in for an individual pitcher. Fatigue affects different pitchers in different ways, but many will sacrifice their mechanics and timing as they get tired, and the associated risk of pitching while fatigued is at least somewhat dependent on how a pitcher responds mechanically when it sets in. The challenge is different for starting pitchers than for relievers, and until we have a definitive measurement of in-game fatigue, we will have to use educated guesses and previous patterns to establish those thresholds for individual pitchers.
Tools such as motion analysis can help the process, to establish mechanical baselines so that coaches and players can more accurately assess fatigue during the course of a game. Other tools such as PAP and even time-of-game measurements can add to the information we can use to establish fatigue thresholds. A manager can go into a game knowing that tonight's starter typically goes 95-100 pitches before fatigue has an effect, that it sets in after an average of 140 minutes, and that this player tends to lose balance and posture as he gets tired. All of this information can go into the decision-making process of when to remove the starter for a reliever, and the costs and benefits associated with each side of the decision.
BP: It is often stated that throwing split-finger fastballs increases the likelihood of arm injury. Why?
DT: I haven't come across any research that found a convincing link between a split-finger fastball and a specific arm injury, and I would put this in the category of unproven conventional wisdom. If the pitch is thrown properly, and with the correct frequency of 15-20 percent, there shouldn't be an increased risk of injury beyond other types of pitches. Everything about a split-finger delivery is the same as a regular fastball, aside from the grip. The only difference is the physical split of the fingers, and it is true that players with small hands will feel pain in those fingers if they attempt to stretch too far, and get an extremely wide grip. A wide grip is not necessary to throw a split-finger, and what most kids try to find is actually a forkball grip. Forkballs are great if you're Bob Welch or Jose Contreras, but not so great if your hands are still growing and can't yet hold a baseball properly. During my playing days, I relied heavily on a split and eventually a forkball, as I couldn't grip a changeup properly, and had not yet learned how to throw a breaking ball that didn't hurt my elbow. I would spend hours with a forkball in my left hand, so that I could stretch out my fingers and grip the pitch easily. I never experienced any pain in my elbow from forkballs, and throwing the pitch felt the same as a fastball, but there were nights when my hand hurt from the stretching exercise.
To get around this, I have taught some young players to throw a 'pitchfork,' which is like a combination split-change, or a three-finger splitter. The name fits when the pitch is gripped correctly, with the index and ring fingers split to the sides of the baseball, and the middle finger gripped right over the top, bisecting the ball with the thumb. The pitchfork requires less stretching of the fingers, but allows the pitcher to grip the ball deep in his hand, and create a velocity drop with the same arm action and forearm angle as a fastball.
BP: So, you see no direct correlation between the pitch and increased injury risk?
DT: I am not an expert on how the finger spread impacts the connective tissue within the throwing arm, and would have to defer to our colleague Dr. Andrews for a proper opinion. Mechanically, we have not detected any injury risk factors that are associated with splitters, but the guys at the Andrews Institute or ASMI could definitely add to the discussion. The difficulty with injury prediction and prevention is similar to that for fatigue, as there are several confounding variables. Arm injuries are a product of mechanics, workloads, functional strength and flexibility, nutrition, genetics, and luck. You can have great mechanics and functional strength, like Mark Prior, and still get taken down by a Brad Hawpe line drive or a collision with Marcus Giles on the bases. Of course, Mark was also in the top four in baseball in PAP in 2003 and 2005, while still thick in the injury nexus, so he's dealt with a combination of bad luck and heavy workload.
BP: When Tom House talked to BP in July 2006, he mentioned the importance of timing in a pitcher's delivery. Can you elaborate on what he was referring to?
DT: Tom was referring to the amount of time that it takes for a pitcher to execute each individual phase in the kinetic chain of the pitching delivery, as well as the time for the entire pitch cycle from first movement to ball release. In my opinion, timing is the single most important aspect of the delivery, and teaching a pitcher to find his own ideal timing signature is the most critical phase in development. Each player has a unique personal timing, but all pitchers fall within a predictably narrow range, once they've achieved a strong level of mechanical efficiency.
Working with our colleagues at Titleist Golf, we have learned that the timing and sequencing for the kinetic chain of events is very similar between a golfer's swing, a batter's swing, and a pitcher's delivery. The best pitchers of all time have been able to consistently repeat their deliveries, including timing, positioning, mechanics, and sequencing. In the book, we challenge the established convention of using a slide step from the stretch, and study how a forced change of timing can impact the effectiveness of pitches thrown with a slide step. There are numerous mechanical flaws that arise from improper timing, and the key to correcting those flaws is the recognition that timing is important, and the identification of the timing that produces the best delivery for each pitcher. The key to mechanical consistency is repetition of proper timing, and it's one thing that pitchers like Maddux and Smoltz do better than anyone else in baseball.
DT: Well, I would love to get those guys under the high-speed cameras, pitching at 1,000 frames per second; then I could really do a breakdown for you. But going from what I have seen with my 32 fps eyeballs, here's a brief assessment of their deliveries:
On Hernandez, speaking of inconsistent timing, he is one of the first pitchers that came to mind when I answered the last question. In particular, he struggles with his timing from the stretch, though he is much better from the windup. He also has a pretty big posture change near release point on most of his pitches, particularly from the stretch. Those are the negatives, but there are pitches when he lines it up that look outstanding, with great balance, momentum, and stride, as well as excellent hip-shoulder separation (his greatest asset) and angular-trunk-rotation timing. He is explosive when he lines up all the pieces, but he needs to do that more often. In my opinion, he has a lot more in the tank, and with some refinement with his timing and his posture (possibly related to functional strength), he could really take a big step forward.
As for Willis, he provides one of my favorite examples of signature, and how genetics can play a role in mechanical efficiency. Dontrelle has had the extreme leg kick since he was in high school, and he depends on that leg kick to generate stride and momentum, and to create ideal timing for his personal signature. When the Marlins tried to change the leg kick, it had a disastrous effect on his timing, and he struggled to find rhythm for most of the season. The only issue I have is that he seems to lose balance momentarily as he brings the lift leg up to its maximum height, but he consistently regains that balance before he completes his stride. He could get going to the plate a bit sooner, but he efficiently directs his energy to the target once he gets moving, and finishes with a strong posture and glove position at release point, while releasing the ball close to home plate.
BP: How would you break down Tim Lincecum?
DT: Lincecum is a fascinating case, and he does a lot of things exceptionally well mechanically. He definitely has plenty of Koufax in him, both good and bad, but the pitcher I find myself comparing him to most often is Roy Oswalt. They are very similar with respect to mechanics, size, and stuff. I'm actually going to be breaking down individual players' mechanics in a series of articles on our website, within the vein of the great work being done by Carlos Gomez, and Lincecum versus Oswalt will likely be one of the first examples. I'm going to break down strategic pairings of pitchers, either based on age, handedness, stuff, hype, stats, or mechanical trends, and then compare and contrast the pitching motions of those players. The breakdown of similarities between Lincecum and Oswalt is also covered in the book, as we address the conventional wisdom of "don't rush."
BP: Michael Bowden, a top pitching prospect in the Red Sox organization, has a delivery that has been described as being long in back and short in front. What is your opinion of him, mechanically?
DT: Michael Bowden and I actually share a birthday, though my cake has seven extra candles every year. I have never seen Bowden pitch in person, and have only watched video from a single outing, so I can't make a full assessment, but my first impressions were that he is a "stay back" pitcher, who is a bit slow to the plate until after maximum leg lift. His balance is strong for the first part of the delivery, but his head tends to trail behind his center of mass as he gets into foot strike, which could hinder his consistency at release point. He has a good, high leg kick, with some funk as he brings his lift leg down near the ground, and then bursts toward the plate with the foot just off the ground. The lift sequence looks a bit funky, but it works well for him, as it helps him get a good stride despite relatively low momentum. He lands with a closed stride, which some scouts hate to see, but he is able to properly time his upper body rotation, and doesn't throw across his body. He has good delayed trunk rotation, including a bit of a hitch in his throwing shoulder that he uses for extra load, and to buy a split second of time. Like the leg kick, this is a good example of what I call 'functional funk,' which might look a bit weird to the eye, but actually helps him coordinate his delivery. This is an area where I often disagree with other scouts and coaches, because what they see as ugly, I see as something that can disrupt a hitter's ability to pick up the baseball. Funk can work as an edge for deception, as long as it's mechanically efficient. In the video I saw, Bowden had pretty good posture on his fastball at release point, but he got on top of some curveballs by sacrificing posture to get a higher release. It's a common trend that is correctable through mechanical consistency and proper timing, and is likely the result of the overall emphasis on downward plane, particularly with breaking balls. Incidentally, downward plane and its effect on mechanics and batted balls is another conventional wisdom that is covered in the book. Bowden also has a pretty solid glove side, keeping the glove in front of the torso through ball release. I like his delivery overall, as he does a lot of things well, and has some room for improvement. Of course, even Nolan Ryan had room for improvement, so that is no slight to Michael.
BP: You said Bowden has "a pretty solid glove side, keeping the glove in front of the torso through ball release." Why is that important?
DT: Conventional wisdom also dictates how a pitcher should position his glove, as many coaches instruct a player to 'pull the glove to the hip' before ball release. At the NPA, we teach players a strategy that we call 'swivel and stabilize' which involves keeping the glove out in front of the torso through release point. In the book, we break down the motion analysis pitchers into two groups, based on glove position at release point, and compare the data for the 'pull glove to the hip' pitchers to those in the 'swivel and stabilize' group. We also look at pictures of six elite major league pitchers throughout the study of conventional wisdom, including the evaluation of glove position.
BP: You also said that some scouts hate to see a pitcher land with a closed stride. Why?
DT: When evaluating a pitcher, many coaches and scouts follow the conventional wisdom of 'stride straight at the plate,' as pitchers that land open or closed are often assumed to have other mechanical flaws that are associated. So we did some research, using motion analysis numbers and 3-D video, to test the mechanical implications of an open or closed stride. We use a sample of 33 pitchers, aged high school to professional, and look for correlations or trends in the data. We also take a look at elite major league pitchers, using photos provided by Getty Images. The NPA Model is rooted in the motion analysis of elite pitchers, and we have found that mechanical efficiency is strongly related to performance, and that the best pitchers of all time have consistently displayed many of the same mechanical advantages.
DT: Curveballs are a hot topic of debate, given the perceived injury risk associated with Little League players throwing curveballs at an early age. But there is more than one way to throw a breaking pitch, and the NPA method is different from what some coaches teach. At the NPA, we teach a 'karate chop' curveball, where the pitcher throws the ball with the palm facing the body as the throwing arm goes through internal rotation and release point. The alternate method is what many Little League coaches teach, which is to snap or twist the wrist for a curveball, just before release point. Some coaches call it "pulling down the shade," while my Little League coach said it was like "throwing a Pringles can, end over end." Unfortunately, I found out pretty quickly that twisting my wrist near ball release was really painful, specifically to my elbow. But since I joined the NPA, I have developed a decent karate chop curve, which has never resulted in elbow pain.
Barry Zito actually has two curveballs, and takes advantage of both the karate chop curve and the twister. But he is most famous for his huge looping curveball, which I call the 'grandfather clock,' since it's so much deeper in shape than your standard 12-to-6. The grandfather clock is a twist curveball, and twist curveballs typically leave the pitcher's hand with a slight upward trajectory when compared to a fastball or karate chop curve, so many hitters can identify it right out of the pitcher's hand. So a pitcher needs to have exceptional spin and depth on his twister, like on Barry's grandfather clock, in order to be effective at the highest level. The karate chop curves typically look more like a fastball when they leave the pitcher's hand, and are more likely to generate a swing-and-miss when buried in the dirt. Gordon appears to be a karate chop guy, but I would have to take a closer look at high quality video to be sure.
BP: You said that Zito throws two distinct curveballs, a karate chop and a twister. Which is more common, and how many pitchers utilize both?
DT: Zito is the only example that stands out as a player that uses both, and most pitchers that make it to the major league level use a karate chop curveball. But I would need high-speed motion analysis video of every major league pitcher to make a true count, as the twist occurs in about 0.01 seconds, just prior to ball release, and is difficult to see with our eyes. Roy Oswalt is a great example of a major league pitcher with an outstanding karate chop curveball, as he gets a steep downward break on his curve without the use of a twist. [Editor's Note: Further information on the subject can be found on the NPA-produced DVD, Safe Curveballs.]
DT: Unfortunately, I cannot publicly go into the details of these three players' injury histories or their possible association with mechanical trends. But injury analysis is also covered in the book, as we have found some particular mechanical inefficiencies that are potential precursors to injury, and an astute reader that is familiar with these players' mechanics could put the pieces together. In the cases of Harden and Liriano, however, I will say that they have likely suffered from the injury cascade effect. As Will Carroll has noted, non-arm injuries can cause a pitcher to alter his mechanics in order to compensate and avoid pain, creating an immediate increase in further injury risk.
BP: If you did a motion analysis of every pitcher in a given organization, to what extent could you predict short- and long-term arm health?
DT: Again, pitching mechanics are just one element of the injury prediction equation, and the precursors that we have found are the result of some of the latest research. We continue to gather the data necessary to further verify the trends that we have come across, and to search for other potential precursors. From an organizational standpoint, motion analysis could be used to monitor the mechanics of every pitcher in the system as they develop, to help with the assessment of health and performance. This could be used to assess short-term injury risk by looking out for precursors, and perhaps setting up workload limits based on mechanical efficiency, so that pitchers with higher-risk deliveries are kept on a shorter leash until they improve. But the key would be to establish mechanical baselines and player signature, and then monitor that progress throughout the year, and from season to season. For pitchers in the low minors, that might mean two or three analyses per year, while the pitchers on the major league staff might benefit from an analysis per month.
I think the most practical and immediate application would be to find the baseline for individual major league pitchers, and anchor on the mechanics that produce the best results. Then, if a pitcher does go down with any kind of injury, the anchor would be treated as a goal line, and a pitcher wouldn't be put back on the field until he had reached his established level of efficiency. This would help to greatly reduce the occurrence of cascade injuries, and to ensure that pitchers have reached full strength in their recovery before they toe the rubber. Meanwhile, teams can test their own methods of coaching and development by making the appropriate measurements for the players in their system, and compare the motion-analysis results to performance. In this way, each organization has the ability to use motion-analysis data to create their own model for pitching or hitting mechanics.
BP: One last question. A BP reader recently asked if James Shields' changing mechanics may be due to his regaining velocity and control post-surgerically. Can you address that?
DT: Well, it is likely that the causation arrow is pointing in the other direction, in that his regaining velocity and control post-surgery may be due to changing mechanics, and some of the mechanical alterations may have been necessary due to the surgery. Unfortunately, I am not familiar with the specific mechanical adjustments that he has made since 2002, but I did notice an improvement with his balance and posture in 2007 compared to 2006, which I believe helped with his effectiveness. In general, the easiest way for a pitcher to improve his stuff is to work hard on mechanical efficiency, as well as functional strength and flexibility. Mechanical efficiency can improve real (radar gun) velocity, perceived velocity, command, deception, and the available options for playing the chess match between hitter and pitcher. This is why I have a hard time with scouts that are quick to hand out ceilings and projection, because in most cases they underestimate the ability for a player to hone his skills and improve his talent level beyond raw tools. By definition, player development requires athletes to learn, and the learning curve in baseball is steeper than any other major sport, as evidenced by the minor league system. Learning ability is one of the strongest tools a ballplayer can have, and the motivation to improve through hard work is what separates success from failure for countless athletes. But they can only get so far without quality information and communication, which is where we come in.