Strength Training: improving your trunk strength will improve your throwing and striking

Strengthening exercises to improve your throwing performance

Throwing and striking balls with an object all involve the use of the shoulder and the trunk. As James Marshall explains, exercise routines that strengthen these areas can not only increase the speed and power of a throw or strike, but also reduce the likelihood of injury

Coaches often look at a problem somewhere in the body and then concentrate on strengthening that area specifically – for example, the shoulder when hitting or striking. However, sportsmen and women use their whole body in order to generate power at the end of their action and if other parts of the body are not strong enough or have limited range of movement, then injury may manifest itself in the shoulder as they try to overcompensate.

At a glance

Basic biomechanics

Before knowing what to strengthen and how, it is important to understand the basic biomechanics behind throwing actions. Throwing or striking is usually a whole-body action. It’s true that there may be times when you have to throw or strike from a more awkward position, such as kneeling, sitting or lying down, but in this article we will look only at optimal delivery.

Competition is hardly an ideal time for the ‘perfect delivery’, but the Green Bay Packers quarterback Brett Favre was a prime example of this: he was able to provide excellent deliveries while throwing off-balance or across his body.

A study on throwing biomechanics by scientists at Bowling Green University in 2006 looked at 49 children ranging from three to 15 years old, and how well they threw a tennis ball at speed under various constraints. In particular, they looked at different variables in the throwing action and compared them with the skill level of the thrower (1,2).

These variables included:

  • stride length;
  • time from foot contact to ball release;
  • linear and rotational speeds of the trunk;
  • the degree of incline of the pelvis;
  • the position of the humerus and forearm at time of foot contact.

The study first looked at the trunk and lower limb, then at the forearm and humerus, in 11 different combinations, to assess how the ball was thrown farther and to establish an idea of progression for teaching the throw. A number of findings emerged from the study. For example, the progression of throwing technique gradually incorporated more body parts, each one leading to an increase in throwing velocity and distance. Trunk rotation, speed and linear speed of the pelvis and trunk were also key elements in increasing ball velocity, so it was important to optimise this.

It turns out that the throwing action can be performed with just the arm, the same side leg and arm going forward (described as ipsilateral); or with the opposite arm and leg (contralateral). Taking the example of boxing, a jab is ipsilateral and a right cross, contralateral.

The stepping action of either leg results in an increased linear speed of the trunk and pelvis;
and the faster the step action, the faster is the linear velocity. Throwing with the same side arm and leg forward helps to increase the linear velocity of the trunk, but reduces the ability of the trunk to rotate. By contrast, having the opposite leg going forward enables the trunk and pelvis to rotate (see figure 1). A longer-than-normal stride also helps with this rotation. In the contralateral throw, using the opposite arm and leg forward allows pelvis and trunk rotation, increasing throwing speed.

Figure 1

Tilting the pelvis forward is also important: this helps to increase movement of the upper torso forward after an initial trunk hyperextension caused by the step (see figure 2, overleaf). This delay causes the rectus abdominus and oblique muscles to be contracted eccentrically during the trunk hyperextension and then rapidly concentrically as the trunk moves forward. A delay between the trunk rotating and the upper torso rotating tends to occur in better throwers – this can be seen in the wind-up of major league baseball pitchers (3). This effect is due to the fact that the initial leg, pelvis and rotational trunk speeds are greater, creating a lag effect. The humerus then has to catch up with the trunk and does so at speed, resulting in an increase in ball velocity. Less able throwers do not generate such fast rotational and linear lower body speed and there is not the same lag effect. This lag is effectively an ‘upper body plyometric movement’. The internal rotator muscles of the upper arm are first stretched (because the lower trunk has accelerated away) and then rapidly contract to catch up with the trunk.

Figure 2

You can see this effect for yourself with an elastic band. Cut it in half and hold one end in your left hand. Just using your left hand, flick the band against a wall. Now try the movement again: but this time hold the other end with your right hand, then flick your left hand and, as the band stretches, let go with your right – you will see how much quicker the band moves.

A longer step results in an increase in forward lean of the trunk, contributing to the forward velocity of the throwing action; and the more joints moving at speed that are used in any action, the greater the overall speed of delivery. Increasing an individual joint speed helps the overall action, but just as important is the ability to coordinate all these actions smoothly and efficiently. The more skilled the thrower, the quicker is the chain of movements from initial footstep to release of the ball. Trying to falsely create a lag effect will result in less overall velocity, so there is no point having strength if you do not have the coordination to apply it. A long, quick step with forward pelvis tilt and quick trunk rotation will create a lag effect that causes the shoulder and arm to accelerate quickly before releasing the ball or striking the object.

Training exercises

Once you understand the basic biomechanics and what muscle groups are responsible for what actions, then you can start training them. The connection between different parts of the body has been known for some time; in their classic text Anatomic Kinesiology, Logan and McKinney describe the effect of the core musculature as being like that of a Mexican serape (4). A serape is a Mexican garment that wraps around the body in a criss-cross fashion. The abdominal muscles work together in the same way, and move and support the trunk in coordination with each other, not in isolation.

It therefore makes sense to strengthen the body as a whole, before looking at individual parts that need specific attention. Trying to isolate individual core muscles would not only take a lot of time (there are more than 20 different muscles attached to the pelvic girdle); it also doesn’t replicate how the body actually works.
Two good examples of whole-body strengthening exercises that work the trunk are the deadlift and the squat. The deadlift is an exercise that requires the athlete to pick up a weight from the floor; the squat requires the athlete to perform a sitting motion with weight on their shoulders. Sitting down and picking things up from the floor are both basic movement patterns that occur in everyday life, and the squat and deadlift are just more strenuous variations on this.

Ditch the stability ball

Trunk muscle activation in these two exercises is greater than, or equal to, that when performing stability ball ‘core’ exercises (5). The researchers in this study looked at muscle activation of the rectus abdominus, external oblique, longissimus and multifidus when performing deadlifts and squats at 50, 70, 90 and 100% of 1RM, and when doing three different stability ball exercises.

Even at loads of 50% of 1RM, the deadlift and squat had the same muscle activation as that of the stability ball exercises. When the loads increased, there was no difference in rectus abdominus and external oblique activation, but there was in the longissimus and multifidus, both of which are back extensors. The limitation of these exercises is that they are performed in a single plane of movement. Some resisted rotational work would be recommended to work in that plane of movement. For healthy athletes, there seems little point in performing stability ball exercises, when that time could be spent training with whole-body exercises that are just as effective, if not more so, and which also work the major muscle groups of the legs.

Kneeling down on a stability ball to practise your passing or throwing can actually disrupt your skill or potentially cause injury, because you are inhibiting the ability of the pelvis to rotate and tilt. Trying to throw or strike in this position as powerfully as when standing may increase the shear forces on your upper limbs, due to compensation for the lack of lower-limb involvement. The exception may be the likes of water-polo players, who are more susceptible to shoulder injuries than other throwing sports, due to the lack of ground reaction forces when throwing. In this case, training in positions without the use of the lower limbs would be more appropriate.

Training velocity

Some fundamental strengthening work needs to be done at slower speeds; but, unless you train at game-related speeds also, your muscles will not be placed under the correct degree of stress necessary for specific development. The velocities created at all the joints are high – perhaps as high as 7,000 degrees of rotation per second in elite throwers (6) – so your training should reflect this. In the deceleration phase of the throw, the force on the humerus can exceed 500N (about 135kg) (7).

Traditionally, shoulder-strengthening exercises have been performed concentrically as part of injury rehabilitation programmes. However, when throwing, the external rotators of the humerus and the rotator cuff muscles contract eccentrically during the deceleration phase of the arm, and an imbalance between concentric strength of the internal rotators and eccentric strength of the external rotators can lead to injury (8).

Two recent studies with female college tennis players and the baseball throw in male college players used eccentric rotator cuff and shoulder training protocols to address this imbalance (9,10). The female tennis players used elastic bands to create eccentric contractions from the external rotators of the humerus.

The limitation of the study in practical terms was that it did not assess the tennis serve itself before and after the interventions. Instead, the ratio of concentric to eccentric strength was assessed and an assumption made that this would increase shoulder stability and therefore help prevent injury when serving. The eccentric strength work took place only around one joint and in one plane of movement, so an assumption that this transfers to the multi-joint, multi-plane tennis serve needs to be made.

The study with the baseball players also used elastic bands for some of the exercises, but it also used a 2lb medicine ball to provide mass to throw forwards and backwards around a single-joint axis in standing and kneeling positions. Compared with a control group that used only traditional concentric contraction strength work of the shoulders, the plyometric group improved the velocity of their pitching from 83.15mph to 85.15mph after eight weeks of training. The key point here is that the isolation work on a single joint (with the potential to produce imbalances that could lead to injury) was augmented with strengthening exercises (using a medicine ball) to augment the overall throw.

Full-body exercises

The combination of full-body strength exercises and multi-joint, multi-planar ballistic throwing actions seems to be an effective way of developing the speed and power of throws, swings and strikes. For example, a programme in which high-school baseball players used medicine balls was found to increase their rotational torso strength and angular velocities of the hip and shoulder more than just full-body resistance exercises alone (11,12).

The control group just swung a baseball bat 100 times a day, three days a week, for 12 weeks. The full-body trial group used squats, stiff-legged deadlifts, bench and shoulder presses, dumbbell rows, triceps extensions and biceps curls, as well as the baseball bat swings. The medicine ball group protocol consisted of three days a week of doing the resistance exercises, followed by medicine ball exercises. The full-body medicine ball throws were performed on the middle day, when no leg weights were used, so as to avoid fatigue.

Most importantly, the combined medicine ball/strength group improved their bat swing velocity by 6.4%, compared to 3.6% in the strength-only group, compared to no change in the control group.

So why not just do medicine ball throws and technical work? Because there are also other beneficial side-effects of an overall strength training programme, such as reducing the likelihood of injury, or improving performance when running bases, and jumping or diving to catch the ball.

Medicine Ball


The problem with research is that it usually concentrates on just one aspect of the body, when we have seen that the whole body is used in an everyday action such as the throw. Working with one joint, such as the shoulder, enables specific measurements to be taken before and after a very specific exercise protocol is introduced. However, doing so may interfere with the overall movement mechanics. Also, the speed of the joint action during these specific exercises nowhere near matches that of the actual movement, because there is only one joint involved. The sequential effect of using multiple joints is what leads to the massive speed of the shoulder, elbow and wrist during the throw or strike.
Training should therefore work on strengthening the overall body with multi-joint lifts; isolating the muscles that are likely to get injured; working with the correct form of contractions in several planes
of movement; and using ballistic actions involving the whole body, with the aid of implements such as medicine balls. This must be coupled with skill practice, so that the underlying correct movement pattern is constantly reinforced. If the movement requires the athlete to not be in contact with the ground during the action (eg water polo, a volleyball spike, a tennis serve for some players) then the strengthening and throwing actions with medicine balls should not use the lower limbs.


1. Research Quarterly for Exercise and Sport 77(4), 417-427, 2006
2. Research Quarterly for Exercise and Sport 77(4), 428-436, 2006
3. Journal of Applied Biomechanics 17, 164-172, 2001
4. Anatomic Kinesiology. Logan, GA, McKinney,
WC (1982) WC Brown
5.JSCR 22(1) 95-102, 2008
6. Journal of Orthopaedic and Sports Physical Therapy 18, 402-408,1993
7. Clinical Sports Medicine, Bruker & Kahn, 2007 McGraw-Hill Professional.
8. Journal of Sports Medicine and Physical Fitness 41, 403-410, 2001
9. JSCR 21 (1) 208-215, 2007
10. JSCR 22 (1) 140-145, 2008
11. JSCR, 21(4), 1117- 1125, 2007
12. JSCR, 21(3), 894-901, 2007

Get on the road to gold-medal form and smash your competition.
Try Peak Performance today for just $1.97.

Privacy Policy [opens in new window]
Please Login or Register to post a reply here.