Biomechanics of Youth Windmill Softball Pitching

Sherry WernerDr. Sherry Werner Ph'D shared this with me, so I thought I would share it with you – Gary Leland.

From the Tulane Institute of Sports Medicine, New Orleans, Louisiana

Sherry L. Werner,* PhD, John A. Guido, MHS, PT, SCS, ATC, CSCS, Ryan P. McNeice, Jasper L. Richardson, MEd, ATC, CSCS, Neil A. Delude, MS, PA-C, ATC, and Gregory W. Stewart, MD

Background: Limited research attention has been paid to the potentially harmful windmill softball pitch. No information is avail- able regarding lower extremity kinetics in softball pitching.

Hypothesis: The stresses on the throwing arm of youth windmill pitchers are clinically significant and similar to those found for college softball pitchers.
Study Design: Descriptive laboratory study.

Methods: Three-dimensional, high-speed (240-Hz) video and stride foot force plate (1200 Hz) data were collected on fastballs from 53 youth softball pitchers. Kinematic parameters related to pitching mechanics and resultant kinetics on the throwing-arm elbow and shoulder joints were calculated. Kinetic parameters were compared to those reported for baseball pitchers.

Results: Elbow and shoulder joint loads were similar to those found for baseball pitchers and college softball pitchers. Shoulder distraction stress averaged 94% body weight for the youth pitchers. Stride foot ground reaction force patterns were not similar to those reported for baseball pitchers. Vertical and braking force components under the stride foot were in excess of body weight.

Conclusions: Excessive distraction stress and joint torques at the throwing-arm elbow and shoulder are similar to those found in baseball pitchers, which suggests that windmill softball pitchers are at risk for overuse injuries. Normative information regard- ing upper and lower extremity kinematics and kinetics for 12- to 19-year-old softball pitchers has been established.


The game of softball originated in 1887. In 1995, fast-pitch softball was the largest team sport in the United States, with more than 35 million participants, and an estimated 41 million people playing worldwide. Fast-pitch softball is one of the fastest growing sports for women at the high school and college levels. The Amateur Softball Association (ASA), the national governing body that selects athletes to compete on the US Olympic team, reports that 83 000 youth girls’ fast-pitch teams, composed of more than 1.2 million athletes, were registered with them in 2003. Currently, including the ASA, there are 5 major governing bodies of softball in the United States. A compilation of registration numbers from all 5 governing bodies indicates that more than 2 million female adolescents between the ages of 12 and 18 competed in fast-pitch softball in 2003. Despite its popularity, however, little bio- mechanics and/or sports medicine research has been carried out in this sport.

The fast-pitch game is essentially a scaled-down version of baseball. The playing field is smaller, with the bases 60 ft apart as compared to 90 ft in baseball. The fundamental difference between the 2 sports is pitching. Instead of an elevated mound, a softball pitcher throws from a flat pitch- ing circle with an 8-ft radius from the pitching rubber. The distance from the pitching rubber to home plate also differs; it is 40 ft for youth softball as compared to 60 ft and 6 in for baseball. In terms of the reaction time of a batter, the game of fast pitch is directly comparable to that of baseball. A regulation softball has a circumference of 31 cm (12 in) and a higher mass (200 g) than does a regulation baseball, which has a circumference of 23 cm (9 in) and a mass of 146 g. Figure 1 depicts the windmill delivery style used in fast-pitch softball.

Powell and Barber-Foss reported a significantly higher injury rate for girls involved in softball when compared to that of boys participating in baseball.

Figure 1. Side-view sequence of windmill softball-pitching delivery. TOB, top of the backswing; SFC, stride foot contact; REL, ball release.

Figure 1. Side-view sequence of windmill softball-pitching delivery. TOB, top of the backswing; SFC, stride foot contact; REL, ball release.

The only quantitative pitching injury prevalence study in the literature indicated that approximately 50% of the pitchers at the 1989 College World Series had a time-loss injury during that season. Results from this work suggested that more than 80% of the injuries occurred to the upper extremity. Loosli et al recommended that injury prevention strategies be investigated in softball pitching because of this high incidence of injury. Stress fractures of the ulna and fifth metacarpal have also been documented in softball pitching, as have cases of radial neuropathy.

In the only published study on softball-pitching biomechanics, Barrentine et al3 concluded that high forces and torques occurred at the elbow and shoulder joints of 8 collegiate pitchers. These authors reported distraction stresses of 70% to 98% body weight (%BW) at the shoulder and elbow joints and concluded that the biceps labrum complex is at risk for overuse injury in windmill softball pitching. Similar stresses have been observed in baseball pitching. It is the repeated application of these high loads that eventually results in the development of the classic overuse injuries commonly observed in baseball pitchers. Unlike in the sport of baseball, a softball pitcher may pitch as many as six 7-inning games during a weekend tournament. Traditionally, the best pitcher on a high school or college team pitches most, if not all, of the games each sea- son. As a result, approximately 1200 to 1500 pitches can be thrown in a 3-day period for a windmill pitcher, as com- pared to 100 to 150 for a baseball pitcher. Thus, it seems reasonable to assume that the chance of overuse injury from softball pitching is high.

To the authors’ knowledge, no published report exists regarding lower extremity kinetics in softball pitching. Even for baseball pitching, which has been well studied for the upper extremity, minimal data are available on ground reaction force characteristics. Ground reaction forces are important in pitching because the muscles of the lower extremity and trunk are larger than are those in the upper extremity, and the only external contact a pitcher has is between the feet and the ground. In softball, pitchers throw from a flat pitching surface, without the aid of gravity (ie, from a raised mound in baseball pitching). Therefore, knowledge of the magnitudes and timing sequences of ground reaction forces in softball pitching may be important for clinicians working with these athletes.

As a result of preliminary study and comparison to the literature available regarding baseball pitching, it appears that the loads on the shoulder and elbow joints are clinically significant in softball pitching and that the prevalence of upper extremity injury is high. Because of the lack of research in this area, additional data are necessary to draw conclusions about upper and lower extremity kinetic parameters (ie, ground reaction forces, joint loads, etc) and kinematic parameters (ie, range of motion, speeds of movement, etc), particularly at the youth level. Therefore, the purposes of this study were to provide a large biomechanical database on youth windmill pitchers, to compare youth pitchers to the data in the literature on collegiate pitchers, and to offer the first lower extremity kinetic data for these athletes.


Fifty-three healthy female windmill pitchers, ranging in age from 11 to 19 years, participated in the study. Fifty- one of the athletes were right-hand dominant and 2 were left-hand dominant. Mean age, height, and mass were 14 ± 2 years, 165 ± 8 cm, and 59 ± 9 kg, respectively. The pitchers were recruited from local recreation, travel, and high school softball teams. The mean years of pitching experience was 5 ± 2 years (range, 1-10 years). The average pitcher in the study started to pitch at age 9 ± 1 years.

Data Collection

After obtaining written consent (approved by the Institutional Review Board of the Tulane University Health Sciences Center) from each participant and her parent/guardian, height, weight, radius length, humerus length, and elbow and shoulder range of motion measurements were taken. Bilateral elbow extension and carrying angle were measured by a physical therapist with the athlete standing, and shoulder internal rotation was measured functionally (reaching as high as possible behind the back). The positions of the tips of the thumbs for the throwing and nonthrowing hands were noted as the athlete internally rotated the shoulders (one at a time) behind the back, and the distance between the thumbs was considered a measure of functional internal rotation. Negative functional internal rotation indicated less range of motion for the throwing side as compared to the non- throwing side. Passive shoulder external rotation was measured with the athlete lying supine on an examination table. Internal rotation was also measured in a supine position for 23 of the athletes.

Twenty-five spherical reflective markers (Motion Analysis Corp, Santa Rosa, Calif) were then placed bilaterally at the lateral tip of the acromion, lateral epicondyle of the humerus, anterior and posterior hip, medial and lat- eral epicondyle of the femur, and medial and lateral malleolus; between the second and third metatarsal heads; and on the heel. Markers were also placed on the throwing-arm radial and ulnar styloids, throwing hand at the third knuckle and on the right side, left side and top of the head (attached to a hat worn by the pitcher). The pitcher then per- formed her normal warm-up routine including throwing, pitching drills, and stretching of the upper and lower extremities.

Subsequently, each athlete was allowed to acclimate to an indoor pitching mound in the Human Performance Laboratory, Tulane Institute of Sports Medicine. The mound was positioned so that the stride foot of the pitcher would land on top of a 60 × 120-cm Bertec force plate (Bertec Corp, Columbus, Ohio), which was anchored and recessed in the ground underneath the pitching mound. The laboratory setup for data collection is depicted in Figure 2. The pitchers were able to throw the regulation 12.2-m (40-ft) distance to a catcher. Six electronically synchronized high-speed Falcon video cameras (Motion Analysis Corp) surrounded the pitching mound and were attached to the walls of the laboratory via a sliding track positioned approximately 2.5 m above the ground. The camera positions were chosen to maximize the number of cameras able to see each reflective marker in each frame of the video. The root mean square error in calculating the 3-dimensional (3D) location of a marker in the calibrated pitching area was 0.50 cm.

Once the athlete was adequately warmed up and acclimated to the indoor mound, fastball trials were recorded at 240 Hz by all cameras. Data collection continued until the athlete was satisfied that 10 representative fastball trials had been gathered. Wild pitches and those pitches excluded by the athlete were not part of the subsequent data reduction. Force plate data were sampled at 1200 Hz and synchronized in time with the video data using the Motion Analysis Expert Vision (EVa) HiRes integrated hardware/software system. All pitches were charted for location from behind the athlete, and a radar gun was used to assess ball velocity for each pitch.

Figure 2. Top view of data collection setup.

Figure 2. Top view of data collection setup.

Data Reduction

The 3 fastball trials, thrown for strikes, with the highest ball speeds were chosen for each athlete.3,10,11 The EVa 6.0 software was used to track the locations of the 26 body landmarks for these trials. The time interval from 50 milliseconds before the instant the ball left the glove until 50 milliseconds after ball release was studied. The direct linear transformation method was used to obtain 3D coordinate data for each landmark.1,22 Joint centers were calculated as virtual markers with the EVa 6.0 software for the throwing wrist and bilaterally for the ankles, knees, and hips. The ankle joint centers were defined as the midpoints between the medial and lateral malleoli, the knee joint centers were defined as the midpoints between the medial and lateral epicondyles of the femurs, the hip joint centers were defined as the midpoints between the anterior and posterior hip markers, and the throwing-wrist joint center was defined as the midpoint between the radial and ulnar styloids.

The locations of the dominant and nondominant shoulder and elbow joint centers were estimated as described by Fleisig et al.11 To define the long axis of the trunk, mid- shoulder and midhip points were calculated as the mid- point of the throwing- and nonthrowing-shoulder joint centers and the midpoint of the stride and nonstride hip joint centers, respectively. Coordinate data were then conditioned with a Butterworth fourth-order, zero-lag digital fil- ter (cutoff, 10 Hz).
Data Analysis

The duration of the windup phase of the windmill pitch (from the instant the ball left the glove until the top of the backswing) varied between pitchers. The interval from the top of the backswing until the instant of ball release was identified as the delivery phase. Motion of the throwing arm during the delivery phase was termed downswing. The delivery phase was further divided into 2 temporal phases, defined from the top of the backswing until the instant of stride foot contact and from stride foot contact until ball release (Figure 1). The follow-through phase was defined from ball release until 50 milliseconds after ball release.

Screen Shot 2015-04-04 at 3.20.36 PM

Linear velocity and acceleration for each landmark were calculated using EVa utilities. Stride length was calculated as the distance from the ankle of the nonstride foot to the ankle of the stride foot in the forward direction. To enable comparison between subjects, this distance was also computed as a percentage of body height (%HGT). Stride foot orientation was defined as the angle, in the horizontal plane, between a vector from the stride foot heel to the stride foot toe in the forward direction. A stride foot orientation angle of 0° indicated that the stride foot pointed toward home plate. Throwing-arm elbow, stride foot knee, and hip angles were calculated using standard 3D calculations. For both elbow angle and knee angle, 0° indicated full extension.

Shoulder flexion and abduction angles were calculated via methods described by Feltner.8 Shoulder (upper trunk) rotation was defined by a vector from the nonthrowing shoulder to the throwing shoulder. Hip (lower trunk) rotation was defined by a vector from the stride hip to the non- stride hip throughout the pitch. Hip rotation was considered fully open at 90° when both hips were in line with home plate and second base and was considered fully closed at 0° when the hips were square (facing) to home plate. Shoulder angle conventions are illustrated in Figure 3. All angular velocities were calculated as the first derivative of the time-dependent relative angular displacements.

The forces and torques at the shoulder and elbow joints of the throwing arm were calculated according to methods described by Feltner and Dapena. To normalize data between subjects, forces were expressed as %BW and torques as %BW.HGT.10

Anatomically relevant reference frames were defined at the throwing elbow and shoulder joints to make clinical application of the joint kinetics. The elbow-based reference frame, Re, is represented in Figure 4a. Resultant joint forces at the elbow were transformed to the elbow-based reference frame and defined in 3 directions: medial (+)/lateral (–), superior (+)/inferior (–), and distraction (+)/compression (–).

The transformed elbow joint torques were defined in 2 directions and indicated extension (+)/flexion (–) and varus (+)/valgus (–). The third torque direction, supination (+)/pronation (–), was not meaningful with the current reference frame.

The reference frame at the throwing shoulder, Rs, is depicted in Figure 4b. Resultant shoulder joint forces were transformed to this reference frame and defined in 3 directions: posterior (+)/anterior (–), lateral (+)/medial (–), and distraction (+)/compression (–). Resultant joint torques transformed to the shoulder-based reference frame were defined in 3 directions: abduction (+)/adduction (–), external (+)/internal (–) rotation, and flexion (+)/extension (–).

A trunk-based reference frame, Rt, was defined to calculate shoulder angular displacements. Figure 4c displays Rt. Peak vertical, braking, propulsive, and medially and laterally directed stride foot ground reaction forces were identified. Times from stride foot contact to peak ground reaction force components were calculated, as were areas under the force curves.

A standard statistical software package (SYSTAT, SYS- TAT Inc, Evanston, Ill) was used to further reduce the range-of-motion, kinematic, and kinetic data. Bonferroni- adjusted paired t tests were used to compare passive throwing- and nonthrowing-arm ranges of motion measured by the physical therapist (P ≤ .05). For each pitcher, data from the 3 fastball trials chosen for analysis were averaged and used for the subsequent evaluation. Descriptive statistics were obtained for the 53 pitchers and compared to those reported by Barrentine et al.

RESULTS Range of Motion

Throwing and nonthrowing shoulder and elbow ranges of motion are exhibited in Table 1. Shoulder external rotation range of motion was statistically significantly (P ≤ .05) greater for the dominant arm (129°) as compared to the nondominant arm (123°). Shoulder internal rotation, measured functionally, had a mean of –2.5 ± 2.5 cm for the 53 athletes, indicating less internal rotation range for the throwing shoulder as compared to the nonthrowing shoulder. For the 23 pitchers whose internal rotation was also assessed in a supine position, mean values of 57° ± 13° and 61° ± 11° were found for the throwing and nonthrowing shoulders, respectively. This side-to-side difference was statistically significant (P ≤ .05). Carrying angle had a mean of 12° ± 4° for the throwing arm and 11° ± 4° for the nonthrowing arm. Similar ranges of elbow hyperextension, 2° ± 7° and 3° ± 7°, were found for the throwing arm and nonthrowing arm, respectively. Side-to-side differences for carrying angle and elbow hyperextension were not statistically significant (Table 1).

Figure 4. Right-handed orthogonal reference frames: (a) Re, at the elbow joint; (b) Rs, at the shoulder joint; and (c) Rt, at the trunk.

Figure 4. Right-handed orthogonal reference frames: (a) Re, at the elbow joint; (b) Rs, at the shoulder joint; and (c) Rt, at the trunk.

Screen Shot 2015-04-04 at 3.21.04 PM

The mean ball velocity at release for the 159 fastballs (53 pitchers × 3 pitches) was 25 ± 1 m/s (55 ± 3 mph). The time interval from the top of the backswing to stride foot contact was 45 ± 19 milliseconds for the youth pitchers. The interval from stride foot contact to ball release was 117 ± 17 milliseconds.

As the stride foot contacted the ground, the knee angle was approximately 30° shy of full extension, with a mean value of 33° ± 8°. Stride length had a mean of 103 ± 10 cm for the 53 subjects and was 62%HGT ± 5%HGT. The stride foot tended to be oriented toward a closed position, with a mean stride foot orientation of 35° ± 20°. The position of the pitching shoulder at stride foot contact was relatively consistent for the youth pitchers. Mean shoulder abduction and shoulder flexion were 109° ± 19° and 217° ± 45°, respectively.

Angular velocities of trunk rotation were high during the delivery phase. Upper trunk rotation velocity reached a peak magnitude of 901°/s ± 162°/s, whereas the lower trunk reached a maximum rotational speed of 544°/s ± 139°/s. Rotational speeds of the throwing arm were also high. Peak elbow flexion angular velocity was 716°/s ± 201°/s, and the maximum speed of the arm as it circled through the windmill motion was 1250°/s ± 111°/s.

The throwing-arm elbow remained fairly straight throughout the pitching motion, but it began to bend dur- ing the late delivery phase. At ball release, the elbow was flexed 20° ± 9°. The throwing arm remained close to the body in the frontal plane throughout the pitch. At ball release, mean shoulder abduction was 3° ± 7°. The release point occurred close to the hip in the sagittal plane as well, with a mean shoulder flexion value of –5° ± 7° at ball release. From an open position of approximately 90° near the top of the backswing, the hips moved toward a more closed position of 43° ± 19° at ball release.


Upper Extremity. Joint distraction occurs at the elbow and shoulder joints in any throwing activity when the arm tends to follow the ball as it is released. A maximum elbow compression force, acting to resist elbow distraction, had a mean of 46%BW ± 7%BW (272 ± 60 N) for the 53 youth pitchers. This distraction load was aligned along the long axis of the forearm pointing toward the wrist. A similar distraction stress, aligned along the long axis of the upper arm pointing toward the elbow, had a mean of 94%BW ± 16%BW (555 ± 129 N) at the shoulder joint near the instant of ball release, as indicated by a compression force.

Mean peak elbow extension torque was 9 ± 5%BW.HGT (84 ± 45 N.m) for the 53 pitchers. A valgus load on the elbow, resisted by a varus torque, increased throughout the latter part of the delivery phase and peaked just before ball release. Maximum varus torque had a mean of 7 ± 4%BW.HGT (65 ± 42 N.m) for the youth windmill pitchers. Maximum internal rotation torque at the shoulder had a mean of 3 ± 2%BW.HGT (31 ± 15 N.m). Shoulder abduction torque increased throughout the delivery phase and reached a maximum with a mean of 6 ± 4%BW.HGT (60 ± 40 N.m) just before ball release. Shoulder extension was the highest torque at the shoulder joint and reached its peak just before ball release. This torque increased from zero at stride foot contact to a maximum magnitude of 13 ± 5%BW.HGT (125 ± 51 N.m).

Lower Extremity. Braking/propulsive, medial/lateral, and vertical ground reaction force components for the stride leg are shown in Figure 5 for a representative pitcher. The braking force peaked quickly after stride foot contact to a mean magnitude of 115%BW ± 46%BW and then reversed its direction. This anterior/posterior component of the ground reaction force reached zero at ball release and continued to move toward a peak propulsive force of 24%BW ± 11%BW just after ball release.

The medially directed component of the ground reaction force also increased rapidly after stride foot contact and reached a peak in the medial direction of 42%BW ± 27%BW. The maximum vertical ground reaction force had a mean of 139%BW ± 43%BW approximately 300 milliseconds after stride foot contact. The area under the vertical ground reaction force curve had a mean of 24 N.seconds and 411 N.seconds from stride foot contact to its peak magnitude and from stride foot contact to ball release, respectively. Mean times to peak force and magnitudes of the ground reaction force components are exhibited in Table 2.


Excessive external rotation and limited internal rotation ranges of motion have been well documented in base- ball.4,5,12,17 Similar results were found for the 53 youth softball pitchers in the current study. Throwing-shoulder external rotation range of motion was, on average, 6° greater than that for the nonthrowing shoulder, and internal rotation was 4° less for the throwing arm. Further research needs to be carried out to determine the extent to which these findings are due to the underarm pitching motion or to the many repetitions of overhand throwing also required of softball pitchers.

In 1994, the Council on Child and Adolescent Health’s Committee on Sports Medicine and Fitness investigated the risk of injury from baseball and softball in children 5

Figure 5. Representative (A) braking/propulsive, (B) medi- al/lateral, and (C) vertical ground reaction force curves. %BW, percentage body weight; SFC, stride foot contact; REL, ball release.

Figure 5. Representative (A) braking/propulsive, (B) medi- al/lateral, and (C) vertical ground reaction force curves. %BW, percentage body weight; SFC, stride foot contact; REL, ball release.

to 14 years of age.6 They concluded that injury risks were similar between the 2 sports but that softball players were less likely to sustain overuse injuries of the pitching arm. In 1992, Loosli et al14 reported that the “traditional view of softball” was that the underhand motion put minimal stress on the throwing arm and, thus, that softball pitchers were less likely to have overuse injuries from pitching. These authors’ own data countered this traditional view when they concluded that a significant number of time- loss injuries had occurred to collegiate windmill pitchers studied at the 1989 National Collegiate Athletic Association tournament championship. The results of the current study also contradict the “traditional view of soft- ball” and support previous work by Barrentine et al,3 who concluded that the windmill pitch creates significant stress on the elbow and shoulder.

Table 3 provides a comparison of the results of this study and those for the 8 collegiate pitchers from the study of Barrentine et al.3 Ball speed was almost identical for the 2 groups of pitchers, indicating that the youth pitchers were highly skilled given their level of play. Peak angular velocities and joint loads were similar between the 2 stud- ies. Peak elbow compression force, the only parameter that appears to be significantly different between the 2 populations, was lower for the youth pitchers as compared to the collegiate throwers. Thus, this study supports the contention of Barrentine et al that the stresses that occur during softball pitching are comparable to those reported for baseball pitching and that given the excessive number of games/pitches thrown in softball, the windmill pitch is a potentially harmful throwing motion. In particular, the high magnitude of shoulder distraction stress (94%BW) near ball release indicates the potential for posterior rota- tor cuff injury. The biceps labrum complex also appears to be at risk because of a combination of the high magnitudes of shoulder distraction stress and elbow extension torque just before ball release.

Screen Shot 2015-04-04 at 3.21.34 PM
The high magnitudes of shoulder distraction stress found in both the current study and that of Barrentine et al3 contradict one of the conclusions made by Maffet et al16 in their study of shoulder muscle activity during windmill pitching. These authors indicated that the stress on the posterior rotator cuff muscles was less in softball as com- pared to baseball pitching because of the dissipation of arm acceleration forces through contact of the throwing arm against the lateral thigh. In 2000, Meister17 further concluded that the results of Maffet et al pointed to an incidence of injury due to shoulder distraction during the deceleration phase of softball pitching which should be considerably less as compared to that due to overhand throwing.

The disagreement appears to be caused by the pitching technique used by the 10 college pitchers tested by Maffet et al.16 These authors reported that the throwing arms of their athletes, almost simultaneous to release of the ball, contacted their lateral thighs, thus stopping forward progress of the humerus. This release strategy would explain the lack of posterior rotator cuff activity found during the follow-through phase and would indicate little arm distraction during this phase. However, none of the 53 pitchers in the current study employed this strategy. On average, the youth pitchers’ shoulders were abducted 3° and flexed –5° at ball release. At the same time, the hips were moving toward a closed position (43°), and the elbow was flexed 20°. Figure 6 depicts 1 of the 53 pitchers in the current study at ball release. Therefore, the results of the study by Maffet et al16 should not be generalized to all windmill pitchers.

Screen Shot 2015-04-04 at 3.21.43 PM Peak vertical and braking components of the stride foot ground reaction force were greater than was body weight and were created quickly after stride foot touchdown for the 53 windmill pitchers. The medially directed force com- ponent had a low magnitude but also peaked quickly after stride foot contact. In general, these stride foot ground reaction force patterns appear to be different from those found in baseball pitching. A comparison of stride foot kinetics for the 2 types of pitching is presented in Table 4.

The magnitude of the peak vertical force found for the softball pitchers was similar to that reported for the stride foot in baseball pitching15; however, the rate of force development was different for the 2 pitching styles. MacWilliams et al15 reported a gradual increase in vertical ground reaction force for their high school and collegiate baseball pitchers, with a peak magnitude being reached just before ball release. The difference in the vertical force loading rate is most likely a result of the difference between the elevated pitching mound used in baseball and the flat pitching circle used in softball. The softball pitch- ers appeared to create vertical and braking force components quickly in an effort to “post” on the stride leg as the nonstride foot pushed the hips toward a closed position at ball release. Baseball pitchers, on the other hand, need to absorb the vertically directed force on the stride leg as they control the forward fall of their bodies from the raised mound. Differences in magnitudes and rates of force development for braking and medially directed force components are also evident between softball and baseball pitching for similar reasons.

Figure 6. Front view of ball release for representative subject.

Figure 6. Front view of ball release for representative subject.

Knowledge of the magnitudes and loading rates of the forces produced/absorbed by the stride foot provides a scientific basis for preventive and rehabilitative programs for clinicians working with windmill pitchers. Overuse injuries to the knee have been reported for windmill pitchers. Physical therapists and athletic trainers rehabilitating lower extremity injuries for a softball pitcher now have quantitative information regarding the demands placed on the stride leg. In addition, it seems prudent for strength and conditioning specialists to begin to incorporate exercises that replicate the impulsive loading conditions endured by the shoulder, elbow, and stride leg for these athletes. Continued study into the forces and torques on the individual joints of the stride leg need to be carried out, as does an investigation of the forces produced by the nonstride leg.


Normative ranges for kinematic and kinetic parameters have been established for a youth population of windmill pitchers. Passive ranges of motion for shoulder internal and external rotation were similar to those found for over- head throwers. High rates of angular velocity are generated by windmill softball pitchers as they move from stride foot contact to ball release in less than 0.120 seconds. The magnitudes of vertical and braking ground reaction forces under the stride foot are in excess of body weight and peak quickly after stride foot contact.

Information regarding passive and dynamic joint ranges of motion, speeds of movement, stride foot kinetics, and joint forces and torques on the throwing arm provides a scientific basis for improved preventive and rehabilitative protocols for youth softball pitchers. Joint loads at the shoulder are similar to those reported for baseball pitch- ing, which suggests that these athletes are at risk for over- use injuries. The increasing number of competitors and high stresses endured by the throwing arm suggest that limiting the number of pitches should be considered for windmill pitchers just as is done to protect baseball pitchers

Screen Shot 2015-04-04 at 3.22.08 PM ACKNOWLEDGMENT
The authors thank David Klein for his assistance with the artwork and Russell Giveans for assistance with data collection.


1. Abdel-Aziz YI, Karara HM. Direct linear transformation: from com- parator coordinates into object coordinates in close-range pho- togrammetry. In: American Society of Photogrammetry. Proceedings of the ASPUI Symposium on Close-Range Photogrammetry. Falls Church, Va: American Society of Photogrammetry; 1971:1-19.
2. Andrews JR, Wilk KE. Shoulder injuries in baseball. In: The Athlete’s Shoulder. New York, NY: Churchill Livingstone; 1994:69-389.
3. Barrentine S, Fleisig G, Whiteside J, Escamilla RF, Andrews JR. Biomechanics of windmill softball pitching with implications about injury mechanisms at the shoulder and elbow. J Orthop Sports Phys Ther. 1998;28:405-415.
4. Bigliani LU, Codd TP, Connor PM, Levine WN, Littlefield MA, Hershon SJ. Shoulder motion and laxity in the professional baseball player. Am J Sports Med. 1997;25:609-613.
5. Brown LP, Niehues SL, Harrah A, Yavorsky P, Hirshman HP. Upper extremity range of motion and isokinetic strength of the internal and external shoulder rotators in Major League Baseball players. Am J Sports Med. 1988;16:577-585.
6. Committee on Sports Medicine and Fitness. Risk of injury from baseball and softball in children 5 to 14 years of age. Pediatrics. 1994;93:690-692.
7. Elliott B, Grove R, Gibson B. Timing of the lower limb drive and throwing limb movement in baseball pitching. Int J Sport Biomech.1988;4:59-67.
8. Feltner ME. Kinematic and Kinetic Parameters of the Shoulder Joint During the Overarm Baseball Throw [thesis]. Bloomington, Ind: Indiana University; 1984.
9. Feltner ME, Dapena JJ. Dynamics of the shoulder and elbow joints of the throwing arm during the baseball pitch. Int J Sport Biomech. 1986;2:235-259.
10. Fleisig GS, Andrews JR, Dillman CJ. Kinetics of baseball pitching with implications about injury mechanisms. Am J Sports Med. 1995;23:233-239.
11. Fleisig GS, Escamilla RF, Andrews JR, et al. Kinematic and kinetic comparison between baseball pitching and football passing. J Appl Biomech. 1996;12:207-224.
12. Johnson L. Patterns of shoulder flexibility among college baseball players. J Athl Train. 1992;27:44-49.
13. Jowett AD, Brukner PD. Fifth metacarpal stress fracture in a female softball pitcher. Clin J Sport Med. 1997;7:220-221.
14.Loosli A, Requa R, Garrick J, Hanley E. Injuries to pitchers in women’s collegiate fast-pitch softball. Am J Sports Med. 1992;20:35-37.
15. MacWilliams BA, Choi T, Perezous MK, Chao EY, McFarland EG. Characteristic ground reaction forces in baseball pitching. Am J Sports Med. 1998;26:66-71.
16. Maffet MW, Jobe FW, Pink MM, Brault J, Mathiyakom W. Shoulder muscle firing patterns during the windmill softball pitch. Am J Sports Med. 1997;25:369-374.
17. Meister K. Injuries to the shoulder in the throwing athlete, part one: biomechanics/pathophysiology/classification of injury. Am J Sports Med. 2000;28:265-275.
18. Plummer B. Media Guide. Oklahoma City, Okla: International Softball Federation; 1996.
19.Powell JW, Barber-Foss KD. Sex-related injury patterns among selected high school sports. Am J Sports Med. 2000;28:385-391.
20. Sinson G, Zager EL, Kline DG. Windmill pitchers radial neuropathy.Neurosurgery. 1994;34:1087-1090.
21.Tanabe S, Nakahira J, Bando E, Yamaguchi H, Miyamoto H,Yamamoto A. Fatigue fractures of the ulna occurring in pitchers of fast-pitch softball. Am J Sports Med. 1991;19:317-321.
22.Walton JS. Close Range Cine Photogrammetry: A Generalized Technique for Quantifying Gross Human Motion [dissertation]. University Park: Pennsylvania State University; 1981.
23. Werner SL, Fleisig GS, Dillman CJ, Andrews JR. Biomechanics of the elbow during baseball pitching. J Orthop Sports Phys Ther. 1993;17:274-278.

Softball Junk

Sponsored by

Please become a BACKER of The Fastpitch TV Network (CLICK HERE)
Have A Question or Some Feedback? Click and send me a voice message.

softball facemasks Fastpitch Magazine Softball Meet Up Group Fastpitch Flowers Weighted Balls

Fastpitch TV Social Media Sites:

This content is provided with a Creative Commons Share-Alike License. Feel free to use this content, so long as you give credit to Gary Leland, of Fastpitch.TV and link to

Gary is a new media producer of fastpitch softball information. For advertising information send him an email to, or visit his personal website site at