February 26, 2003
The Injury Nexus
A Look at Pitcher Injuries
It seems fair to say that the study of pitcher injuries is an important part of sabermetric analysis. The statistical evidence available to test theories about pitcher injuries, however, is often missing. While there are databases that contain every recorded statistic from the days of Al Spalding and beyond, and others that document every play of every game in the past 30 years, a comprehensive database of player injury history simply doesn't exist.
However, between a careful analysis of what data is available, the creative use of proxy variables in estimating injuries throughout time, and the application of some principles of sports medicine, we are at least in a position to make some educated guesses about the nature of pitcher injuries. Our particular focus in this article will be the progression of pitcher injury rates by age.
Methodology and Statistical Results
To create an actuarial backbone for our study, we applied the same approach that is used to calculate attrition rate in the PECOTA forecasts. Attrition rate describes the percentage of pitchers who experience a decline in their innings pitched of at least 50 percent. Such a dramatic decline will not always indicate that a serious injury has occurred--it can also reflect demotion, retirement, and so on. However, by placing a few restrictions on our dataset, we can serve to limit these cases, and use attrition rate as a reasonable proxy for catastrophic injury.
In order to be included in the study, a pitcher needed to have pitched at least 150 innings in the previous season, with a park-adjusted ERA no more than 10 percent worse than his league average. That is, our study was focused on pitchers who had already pitched at least one effective season in the major leagues, and who were likely to have every opportunity to do so again in the absence of significant injury. All pitchers from 1946-2002 were considered, with innings pitched totals prorated over a 162-game schedule. The chart below tracks attrition rate at different ages throughout a pitcher's career.
Even for a successful, established pitcher, the risk of catastrophic injury is meaningfully high throughout his career, almost certainly at least 10 percent in any given season. However, the risk does appear to be to some degree dependent on a pitcher's age. For the very young pitchers in our study--ages 21 and 22--the risk of injury is significantly higher, in excess of 20 percent. Injury rate then drops dramatically as a pitcher matures physically, reaching its lowest point at roughly age 24, while rising gradually throughout the remainder of his career. (Although pitchers aged 37 and up appear in the chart to be as vulnerable to injury as very young ones, that is also the age at which pitchers will begin to retire voluntarily. The uptick in injury risk at the tail end of a pitcher's career is probably not as substantial as what is implied here).
Discussion of Physiological Risk Factors
From the scientific data collected by Dr. Mike Marshall and Dr. James Andrews, to the years of wisdom accumulated by Dr. Frank Jobe, there is a general acceptance as to what factors lead to pitcher injuries. The three major factors are the underlying physical system, degree of use, and biomechanical efficiency.
The physical system includes the bones, muscles, ligaments, and tendons involved in the pitching process. These are centered in the shoulder and elbow of the pitching arm. Most significant are the muscles of the rotator cuff, the glenoid labrum, and the ulnar collateral ligament (UCL). Most pitchers will experience some degree of damage in their pitching arm. A 1999 study on members of the Toronto Blue Jays showed that 23 of 28 pitchers had tendonitis, while 22 had some extent of cartilage damage. Yet most of these pitchers were asymptomatic, and many were pitching effectively in the major leagues.
As an athlete matures, his bones calcify and harden, his growth plates close, and his ligaments reach full strength. Since no athlete matures on the same schedule as another, it is important to note that chronological age does not always directly correlate to physical age. However, as Dr. Jobe and others have noted, a pitcher is generally most vulnerable at a young age, before the bones and muscles of his upper body have fully developed.
The quantity and character of a pitcher's use--and the fatigue that they induce--have received more attention lately in sabermetric circles. As a pitcher fatigues, his biomechanics begin to break down. While the tipping point of fatigue can be difficult to pinpoint, it can be broadly measured by such approaches as pitch counts, velocity tracking, and even observed exertion. As Keith Woolner and Rany Jazayerli have suggested, the relationship between fatigue and injury risk is exponential rather than linear; an overworked pitcher is significantly more likely to experience a traumatic injury.
Finally, a pitcher with poor mechanics, fatigued or not, is at increased risk of injury. Although the lack of readily available data makes it difficult to discuss biomechanical efficiency with the same precision that we do pitch counts, there is no doubt that the makeup of a pitcher's delivery can separate those pitchers that can withstand high levels of use from those that cannot.
We have already discussed how the first of the three physiological elements of injury risk--the intrinsic strain that the pitching motion requires--is of greatest concern for very young pitchers. Indeed, based on a limited sample of MLB injury data reviewed by Under the Knife, pitchers under the age of 24 are especially likely to experience injuries to their elbows and shoulders, those body parts that are put under the greatest stress by the pitching motion. However, it may be more proper to associate the pitching motion itself with the underlying risk of injury observed among pitchers of all ages.
The relationship between age and fatigue is more ambiguous. Our attrition rate study focused only on performance in the most recent season, rather than fatigue accumulated over the course of the career. However, from what specific data we do have available, it appears that fatigue-based injuries are more likely to afflict older pitchers. According to MLB data, while the risk of tears and fractures decreases with age, the risk of strain and inflammation increases. So too does the risk of injury to body parts that are secondary to a pitcher's motion, such as his back, knee, and hamstrings. Fatigue-based injuries such as these may account for the gradual slope upward in injury risk after the age of 25.
It is the final factor--mechanics--that may be responsible for the high incidence of injuries among very young pitchers. It is likely that pitchers with inherently poor mechanics are weeded out very early in their careers. Our attrition rate data suggest that injury risk is very high even for 21- and 22- year-olds who have pitched successfully in the major leagues. One can imagine that it is higher still for pitchers who have not yet turned professional, and for pitchers whose mechanics are sufficiently poor that they do not develop the command necessary to reach the major leagues at all.
There is no ready statistical metric to evaluate a pitcher's mechanics, and even case-by-case observation can obscure the physiology unique to each pitcher. Thus, the most powerful measure of the efficiency of a pitcher's motion may simply be the passage of time without his encountering serious injury. The so-called injury nexus does appear to be a real phenomenon, but it occurs before the age of 23, a younger age than some previous studies have suggested.