[Back to Index] [Back to Personal]  [Back to Homepage]



Alfred Finch

Indiana State University, Terre Haute, Indiana, United States

Throwing may occur along an overhead, sidearm, or underarm pattern as utilized in baseball pitching, discus throwing, and softball pitching, respectively. Critical fundamental throwing characteristics are common across the cocking, accelerative, and decelerative phases. The cocking phase serves to stretch and load the elastic muscle/tendon of the throwing musculature prior to the accelerative phase. The acceleration phase translates the CM while transferring the velocity from the trunk, to the shoulder, elbow and wrist/ball. The deceleration phase blocks the horizontal momentum of the trunk and provides an accurate release in addition to protecting against injury. Commonalities between the throwing mechanics of baseball and discus throwing were examined.

KEYWORDS: throwing mechanics, speed, accuracy, baseball, discus


Throwing is a fundamental movement integral to numerous sporting activities. A throwing movement may occur along three planes of motion as illustrated in baseball pitching, discus throwing, and softball pitching., respectively(Figure 1).

Figure 1 - Overhand, sidearm, and underhand throwing patterns

The general motion of throwing may be delineated into the phase of cocking, acceleration, and deceleration.

Overhand Throwing – Baseball Pitching:

Cocking Phase: The cocking stage begins from the starting stance, followed by early cocking in which the shoulder joint is abducted, horizontally extended, and outwardly rotated, the trunk/torso is rotated in the same direction as the throwing arm (right rotation - right-handed throw), and the left hip is flexed, externally rotated, horizontally flexed, and the knee is flexed. This motion serves to preload and stretch the shoulder inward rotators (latissimus dorsi, anterior deltoid, subscapularis, and teres major), the horizontal shoulder flexors (pectoralis major, anterior deltoid), the preliminary spinal torsion stretches the left side rotators muscles (right external abdominus oblique, left internal abdominus oblique, and the left erector spinae) and the left scapular retractors (rhomboids and trapezius). The elastic loading of this serape musculature is necessary in the lengthening-shortening cycle used in effective throwing mechanics along an upper diagonal plane using the total body (Figure 2).

Figure 2 – Cocking phase and serape musculature that is pre-stretched

Acceleration phase: The accelerative throwing phase is initiated from the sit-out position by the hips outward rotation and leg extension. This motion initiates the uncoiling of the torso which is transferred and accelerated up through the vertebral column and shoulders. (Figures 3 & 4). As the torso rotator antagonists are stretched and co-contracted, in combination with the contraction of the pectoralis major and shoulder joint inward rotators (subscapularis, teres major, and latissiumus dorsi), the trunk velocity is transferred to the arm and later the elbow, wrist and fingers.

Figure 3 – Initial movements of throwing accelerative phase

Figure 4 – Final movements of throwing accelerative phase


Critical sequential timing between the successive accelerations and decelerations of the segments are necessary for effective kinetic link transfer of velocity (Figure 5). The sequential timing of the trunk rotation, horizontal shoulder flexion and elbow extension transfer the trunk’s torsional velocity out through the arm appendage while providing a shortened effective arm radius with a smaller rotational inertia in the beginning of the accelerative phase permitting high rotational velocities and then the extension of the arm leads to a flinging transfer motion which increases the arm radius and tangential velocity.

                          Figure 5 – Kinetic link sequencing of throwing segments


The stride during the accelerative phase provides a translated point of rotation of the shoulder which flattens the ball arc path, providing a longer time window in which the pitcher may provide an accurate release. Additionally, a strong push-off provides acceleration of the trunk and a significant contribution to the resultant ball velocity (Figure 6).

                         Figure 6 – Center of mass translation during stride


Atwater (1970) analyzed the segmental contributions to the ball velocity in overarm throwing performed by a male, major league baseball player and a highly skilled female. It was reported that the male thrower utilized more shoulder action and the female athlete exhibited greater trunk rotation to develop her throwing velocity and the ball velocities thrown were 39.9 and 29.2 m sec-1 for the male and female throwers, respectively (Table 1).



Table 1 – Segmental contribution to ball velocity in overarm throwing


The planting of the stride foot provides blocking action of the horizontal velocity of the stride and a transfer of velocity from the trunk to the arm and eventually the ball. Errors in the placement of the plant foot may include: 1) striding too wide, which prematurely opens the hips and tends to produce an arm lagged position, 2) closed stride would block the hip rotation, 3) overstriding may produce an extended leg position which brakes the throwing motion as the pitcher attempts to rotate over the planted foot or extreme overstriding will necessitate the pitcher to flex the knee in order to reduce the radius of rotation thus permitting sufficient angular momentum to rotate over the plant foot . This action has a tendency to elevate the shoulder/ball path and additionally places high stretching loads on the hamstrings / low back musculature leading to premature fatigue or muscle soreness (Figure 8).


Figure 7 – Overstride with extended knee and overstride with knee flexion


Deceleration phase: During this phase the posterior deltoid, rhomboids, and trapezius decelerate the limb just prior to release which increases the time period for an accurate release, while maintaining the shoulder and elbow joint integrity (Figure 8). High levels of activity of the biceps and brachialis are used in this phase to provide centripetal forces to counteract the centrifugal forces at the shoulder and elbow. The relative activity of the throwing muscles in the three throwing phases are presented in Table 2.

Figure 8 – Temporal sequencing between pectoralis major and

posterior deltoid prior to ball release

Strength training implications: Listed are some strength training exercises used to develop the following throwing muscles: pectoralis major (bench press, supine butterflies), serratus anterior (rope climbing, dips on parallel bars), latissiumus dorsi (latissimus pull-downs), trapezius and rhomboids (rowing), deltoids (shoulder abductions with varied arm positions, upright rowing, military press), biceps and brachialis (elbow curls ), triceps (pushups narrow hand position, dips), teres minor and infraspinatus (wall pulley - external rotations, pullovers), teres major and subscapulari wall pulley – internal rotations, pullovers), erector spinae (straight leg dead lifts, trunk extensions in a Roman chair, trunk rotations with a bar on the shoulders) and serape musculature (medicine ball throws, Figure 9).

Table 2 - Relative activity levels of musculature during throwing phases Hamill & Knutzen, 1995

Figure 9 – Serape musculature exercise, medicine ball throws

Sidearm Throwing – Discus Throwing

The discus throwing movement consists of 1 revolutions in which there is an initial double support phase (starting stance), the sprint stride, drive across to the power position with a plant block, throw position with trunk torsion and shoulder action through to disc release, and follow through. The sprint stride serves to provide the cocking action of the trunk and shoulder, the drive across to the power position and through the throw position represents the accelerative phase and the follow through produces the deceleration of the trunk and arm (Figure 9).

Figure 10 – Discus throwing sequence

Critical factors influencing the range of the thrown discus are: 1) disc projection velocity, 2) disc projection angle, 3) release height, and 4) disc attitude – angle of attack.

Three dimensional video analyses of the top four Men’s Discus performers’ best and worst attempts at the 1996 Atlanta Olympic Games were performed. For the purposes of kinematic comparison, the following throwing performance parameters were selected: disc release velocity, disc projection angle, release height, time of movement, peak trunk rotational velocity, and peak shoulder angular velocity.

Results and Discussion:

The top four Olympic discus throwers’ height and weights were: Riedel (199 cm,

110 kg), Dubrovshchik (193 cm, 115 kg), Kaptyukh (197 cm, 117 kg), and Washington (188 cm, 109 kg).

The best and worst throws recorded by the top four performers in the Discus event: 1) Riedel (Germany), 2) Dubrovschchik (Belarus), 3) Kaptyukh (Belarus), and 4) Washington (United States of America) were selected for kinematic analysis. The kinematic results were:

  1. The medalist throws were 69.4 m (OR) by Riedel, 66.6, 65.8, and 65.4 m for Dubrovshchik, Kaptyukh, and Washington. The performers’ poor throws were 6.3, 6.9, 2.0, and 4.1 m shorter, respectively
  2. The resultant release velocities calculated for the best (worst) throws were 3118 (3008), 2725 (3343), 2567 (2269), and 2500 (2440) cm/s for Riedel, Dubrovshchik, Kaptyukh, and Washington, respectively (Table 3 & Figure 11).
  3. The projection angles for best (worst) throws were 32.4 (30.2), 30.0 (36.4), 35.4 (30.8) and 29.9 (59.9) degrees for Riedel, Dubrovshchik, Kaptyukh, and Washington (Figure 11& 12).
  4. The horizontal velocity due to the body torsion, found that Riedel and Washington used less twisting action in their poor throws and Dubrovshchik used substantially more torsion.
  5. The changes between trials in the horizontal velocities due to the arm action were –4.5%, +.9%, -10%, and –43.3% for Riedel, Dubrovshchik, Kaptyukh, and Washington, respectively.
  6. The movement time across the circle during Washington’s best attempt was so much faster

than the other top competitor’s that it appeared that the movement was too fast to allow for

adequate time for the return of the elastic energy stored in the arm to be transferred to the

disc. His uncontrolled velocity is apparent in his inability to apply adequate centripetal force

to maintain a circular pathway during the final turn and this resulted in an errant disc path

Figure 11 – Discus projection velocities and projection angles

Figure 12 – Discus throwing kinematics



Table 3 Throwing Kinematics for Top Four Discus Performers at 1996 Atlanta Olympics ___________________________________________________________________________



Figure 13 – Washington’s errant disc path at release

To examine the contribution of the spinal rotation to the horizontal disc velocity, the differences between the right hip and shoulder horizontal velocities were calculated. The arm action contribution to the disc horizontal velocity was determined by the difference between the disc horizontal velocity and the right shoulder velocity. The effectiveness of the blocking action was determined by the differences between the right hip and CM horizontal velocities. The horizontal velocities for the trunk rotation, arm action and blocking action are presented in Tables 4a & b.

Table 4a Throwing kinematics for the block, trunk, and arm action

Competitor Variable cm/s Best Throw Worst Throw Difference
Riedel CM Hor. Vel. 29.8 173.0 143.2
  Block Vel. 99.9 242.4 -142.5
  R. Hip Vel. 129.7 415.4 285.7
  Trunk Vel. 233.9 -87.4 -321.3
  R. Shoulder Vel. 363.6 328.0 -35.6
  Arm Vel. 2192.5 2093.0 -99.5
  Disc Vel. 2556.1 2421.6 -134.5
Dubrovschck CM Hor. Vel. 166.6 -5.5 -172.1
  Block Vel. 230.7 146.7 84.0
  R. Hip Vel. 397.3 141.2 -256.1
  Trunk Vel. 120.8 354.5 233.7
  R. Shoulder Vel. 518.1 495.7 -22.4
  Arm Vel. 1803.5 1819.9 16.4
  Disc Vel. 2321.6 2315.6 -6.0




Table 4b Throwing kinematics for the block, trunk, and arm action



CM Hor. Vel. 118.6 87.2 -31.4
  Block Vel. 160.1 241.0 -80.9
  R. Hip Vel. 278.7 328.2 49.5
  Trunk vel. 116.3 100.6 -15.7
  R. Shoulder Vel. 395.0 428.8 33.8
  Arm Vel. 1686.0 1516.0 -170.0
  Disc Vel. 2081.0 1945.5 -135.5
Washington CM Hor. Vel. 72.6 48.8 -23.8
  Block Vel. -36.7 -134.4 -171.1
  R. Hip Vel. 35.9 -85.6 -121.5
  Trunk Vel. 334.6 303.7 -30.9
  R. Shoulder Vel. 370.5 218.1 -152.4
  Arm Vel. 1772.7 1003.6 -769.0
  Disc Vel. 2143.2 1221.7 -921.5


The horizontal velocity due to the body torsion, found that Riedel and Washington used less twisting action in their poor throws and Dubrovshchik used substantially more torsion. The changes between trials in the horizontal velocities due to the arm action were –4.5%, +.9%, -10%, and –43.3% for Riedel, Dubrovshchik, Kaptyukh, and Washington, respectively.

Additionally, Riedel’s gold medal attempt utilized an effective blocking action, facilitating

the transfer of velocity from the trunk, to the arm and disc (Figure 14).

Figure 14 – Riedel’s effective blocking action and rapid disc acceleration prior to

release during his gold medal attempt


An examination of the projection velocity, angle, and release height information found that Riedel, Dubrovschchik, and Washington increased their projection angles resulting in poor throws (Pfaff, 1994). Additionally, Riedel and Kaptyukh lowered their release height on their poor attempts. The influence of these throwing adjustments can not be fully determined because the aerodynamic effects of the disc attitude, angle of attack, and wind conditions were not analyzed (Atlmeyer, L., Bartonietz, K., & Krieger, D., 1994). The throwing velocities were similar to those reported by Ariel in 1976 on Silvester and Oerter (Ariel, G., Finch, A., & Penny, A., 1997) and the projection velocities decreased in 3 of the 4 performers’ poor attempt. Also, the time of movement decreased during the poor throws, and the gold medal throw took the longest movement time (Ariel, G., Finch, A., & Penny, A., 1997). The more successful throws typically had longer time durations which would provide more time for the storage of elastic energy in the arm during the turns and better energy return (Dapena, 1994).

Throwing technique information found that Riedel and Washington increased their arm action during their poor attempts rather than using a body torsion/flinging motion. The blocking action data showed that Riedel and Kaptyukh did not block their forward momentum and Washington was actually moving back during the blocking phase.

An examination of selected kinematic variables of discus throwing techniques found that poor throwing trials by the top performers at the 1996 Olympic Games were caused by improper projection angles, faulty plant foot blocking action, and poor transfer of velocities from the torso

to the arm and then the discus.


Atwater, A. (1970). Movement characteristics of the overarm throw. Unpublished doctoral dissertation. University of Wisconsin.

Finch, A., Ariel, G., & Penny, A. (1998). Kinematic comparison of the best and worst of the top men’s discus performers at the 1996 Atlanta Olympic Games. International Symposium on Biomechanics in Sports I, 93-96.

Ariel, G., Finch, A., & Penny, A. (1997). Biomechanical analysis of discus : 1996 Atlanta Olympic Games. In: G. Dales (Ed.), Proceedings of XIV Congress of International Track & Field Coaches Association. 131-135. Columbus, Wisconsin: Town & Country Printers.

Atlmeyer, L., Bartonietz, K., & Krieger, D. (1994). Techniques and training: The discus throw. Track & Field Quarterly Review. 94 (3), 33-34.

Dapena, J. (1994). New insights on discus throwing. Track & Field Quarterly Review.

94 (3), 37-42.

Hamill, J., & Knutzen, K. (1995).Biomechanical Basis of Human Movement. Baltimore,Maryland: Williams & Wilkins Publishing Company.

Pfaff, D. (1994). Discus Dynamics. Track & Field Quarterly Review. 94 (3), 32.



[Back to Home Page]




[Back to Personal]  [Back to Homepage]