3.12 Medicine Ball and Ballistic Training

The short version

One thing separates ballistic training from ordinary lifting: you let go of the load instead of slowing it down. Bench press a barbell and you have to put the brakes on before your elbows lock, or the bar would leave your hands. Throw a medicine ball the same way and it simply goes — it leaves your hands at full speed and nothing gets braked at all. That single difference changes what your nervous system is being asked to do.

This article covers what ballistic training actually is, how to sort throws into useful categories, how to pick a ball weight, how to build from standing-still throws up to reactive ones, and where the claims about carry-over into sport are solid and where they are oversold.

‌

What makes a movement ballistic

Every normal lift ends with you stopping the weight. Pick up a barbell and press it. Somewhere in the last third of the push you have to quit pushing and start slowing the bar instead. On a moderately loaded bench press that braking can eat a surprising chunk of the range. During it, the muscles that were driving the bar switch off and their opposites switch on (Newton, 1996).

Which is awkward if you came for power. You have just spent the most important part of the movement rehearsing how to brake, at exactly the moment you wanted to rehearse how to accelerate. Throwing solves it by deleting the braking entirely. Release the ball, or jump with the bar, and you get to accelerate the whole way through.

It lands in a useful spot on the speed-versus-load scale. Heavy lifting sits at the slow, high-force end. Sprinting and unloaded jumping sit at the fast, low-force end. Medicine ball work sits in the middle, which happens to be roughly where the most power gets produced. That is the real argument for it, and it is a good one.

Sort throws by what the body does, not by what you are holding. A rotational scoop and an overhead slam are both "medicine ball exercises" and they have almost nothing in common. Filing them together because they use the same equipment is like filing a squat and a bench press together because they both use a barbell. Pattern and direction first, kit second.

Ball weight is the main dial you have. For most rotational and upper-body throws, three to four kilos is plenty. Go heavier and the ball stops flying, which means you have quietly turned a speed exercise back into a slow one. Genuinely heavy balls have exactly one job, and that job is slams.

Rotational throws are the most useful and the most butchered. They work when the sequence starts at the hips and travels outward, the way a good punch or a good golf swing does. They stop working the second the arms take over, which is what happens to almost everybody the first few sessions. If your arms are sore afterwards and your hips are not, the drill did not do what you wanted.

Ballistic, plyometric and explosive are three different words. Ballistic means you release the load. Plyometric means a fast stretch gets reversed quickly enough to reuse the springiness. Explosive just means you tried to move fast. A throw from a dead standstill is ballistic but not plyometric — you can make it plyometric by adding a quick countermovement or a catch, but it does not come that way by default.

And the honest limit. Carry-over into sport is real, but it is modest and it is specific to the pattern you trained. Throws sit alongside sprinting, striking and jumping practice rather than replacing any of them. Letting go of the weight helps the thing you are actually doing; it does not make the drill magically better than everything else. A light ball thrown by a weak athlete is still a light ball thrown by a weak athlete.

The rest of the article goes into the detail: full classification of throws by pattern and plane, ball mass selection, the four-stage progression from positional work to reactive throws, and an honest look at the transfer evidence. Those sections are more technical, so read them when you want the mechanism rather than the map.

‌

Where ballistic training sits on the force-velocity curveA downward sloping force-velocity curve with an overlaid inverted-U power curve. Shaded regions mark the zones occupied by maximal strength work, ballistic throws, and plyometrics.Where ballistic training sits on the force-velocity curveForce-velocity relationshipMechanical power outputSlowModerateMaximal0%25%50%75%100%Movement velocityRelative force or powerMaximal strength: Squats, deadlifts,pressesBallistic zone: Med ball throws, jumpsquats, Olympic derivativesPlyometrics: Bounds, depth jumps,pogos

Figure 1. Heavy strength work lives at the high-force, low-velocity end. Plyometrics live at the low-force, high-velocity end. Medicine ball throws occupy the middle, which is precisely where mechanical power output peaks and where the load can be released rather than decelerated.

‌

This is the whole argument for medicine ball work, and it is a good one. It is also more limited than it is often presented. Releasing the load helps with the pattern being trained; it does not make the drill universally superior, and a light ball thrown by a weak athlete is still a light ball thrown by a weak athlete.

‌

Ballistic, plyometric, and explosive are not the same word

These three terms get used interchangeably and should not be.

‌

• Ballistic The load is accelerated through the full range and released or projected. Medicine ball throws, jump squats, and kettlebell swings at the top of the arc all qualify. The defining feature is the absence of a deceleration phase.
• Plyometric The movement is preceded by a rapid eccentric stretch that is reversed quickly enough to reuse elastic energy. A depth jump is plyometric. A medicine ball throw from a static start is not, unless a countermovement is deliberately built in.
• Explosive A loose descriptive term meaning "performed with intent to move quickly". A heavy squat performed with maximal acceleration intent is explosive but neither ballistic nor plyometric (Behm & Sale, 1993).

‌

A throw can be made plyometric by adding a rapid countermovement or a catch, which is exactly what stage four in Figure 3 does. But it is not plyometric by default, and calling every throw a plyometric obscures what is actually being trained.

‌

‌

Why the medicine ball specifically

The medicine ball has three properties that are hard to replicate. It can be accelerated in any plane, including rotation, which barbells cannot do safely. It can be released without damaging equipment or the athlete. And its mass can be varied in small increments across a range that matters.

‌

That last point is underrated. The difference between a three-kilogram and a five-kilogram ball in a rotational throw is large in terms of velocity and small in terms of joint stress, which gives a coach fine control over the stimulus in a way that adding ten kilograms to a bar does not.

‌

• Rotation can be loaded and released, which is otherwise very difficult to train ballistically.
• The equipment is cheap, durable, and requires almost no setup.
• Distance thrown is a simple, valid, low-tech output measure.
• Throws can be performed at high velocity with far less eccentric cost than jumping.

‌

The last point matters most for athletes carrying lower-limb load. A rotational throw provides a genuine power stimulus with almost no landing exposure, which makes it a useful tool when jumping volume needs to be restricted.

‌

Selecting, sequencing, and loading throws

Classify before you programme

The single most common programming error with medicine balls is treating "med ball work" as one category. It is not. A kneeling chest pass loads horizontal upper-body push in the sagittal plane with the legs removed. A backward overhead throw loads maximal whole-body extension. A standing side throw loads trunk rotation at high velocity. These three drills share a ball and nothing else.

‌

Classifying medicine ball throwsA four column table classifying eight throws by movement pattern, plane of motion, typical ball mass and the dominant quality trained.Classifying medicine ball throwsMovement patternDominant planeTypical massQuality trainedChest passHorizontal pushSagittal3–5 kgUpper-body concentricspeedOverhead throw (forward)Whole-body extensionSagittal3–6 kgTriple extension andposterior chainOverhead slamWhole-body flexionSagittal6–10 kgTrunk flexion,deceleration toleranceRotational scoop throwRotation and extensionTransverse3–5 kgHip-to-shouldersequencingStanding side throwRotationTransverse2–4 kgTrunk rotationalvelocityKneeling chest throwHorizontal pushSagittal3–5 kgUpper body isolated fromleg driveBackward overhead throwFull extensionSagittal4–8 kgMaximal whole-bodyimpulseSquat to press throwVertical pushSagittal4–8 kgVertical power,leg-to-arm transfer

Figure 2. Choose the throw by the pattern and plane you want to load, not by what looks impressive. A rotational scoop and an overhead slam share a piece of equipment and almost nothing else.

‌

Read the table by asking two questions. What pattern do I want to load, and what plane do I want to load it in? If the answer is "rotation, because my athlete is a striker or a thrower", then chest passes and slams are irrelevant no matter how tiring they are. Fatigue is not the goal.

‌

A useful discipline is to write the intended quality next to every throw in the programme. If you cannot write one, the throw is filler.

‌

Rotational sequencing: The part that actually matters

Rotational throws are where medicine ball training earns its place, and where most athletes get it wrong. The fault is almost always the same: The athlete initiates the throw with the arms and shoulders, and the hips follow. That produces a throw powered by a small amount of muscle acting over a short lever.

‌

Correct sequencing runs from the ground upwards. The rear foot drives, the pelvis rotates towards the target, the ribcage follows the pelvis after a deliberate delay, and the arms arrive last (Putnam, 1993). This delay between pelvis and ribcage rotation is sometimes described as separation, and it is the mechanism by which the trunk musculature is placed on stretch before it contracts.

‌

1. Set up side-on with weight loaded onto the rear leg and the ball held at hip height.
2. Initiate by pushing the rear foot into the ground and turning the back hip through.
3. Allow the ribcage to lag behind the pelvis briefly, then let it follow.
4. Release the ball late, with the arms finishing the movement rather than starting it.
5. Finish with the weight transferred onto the front leg and the rear heel turned out.

‌

The coaching test is simple. Film from overhead or from directly behind. If the shoulders and the hips rotate at the same instant, the athlete is throwing with the upper body. If the hips clearly lead, the sequence is working. Distance thrown will confirm it: A properly sequenced throw travels substantially further at the same effort (Ikeda, 2007).

‌

A four-stage medicine ball progressionFour stacked stages moving from positional and pattern work at the base, through bilateral throws and rotational sequencing, to reactive and combination throws at the top.A four-stage medicine ball progressionStage 1 — Pattern and positionKneeling and half-kneeling throws with a light ball. Removing the legs forces the athlete to find trunk and shoulder positionshonestly, and keeps total velocity low while the pattern is learned.Stage 2 — Bilateral power expressionStanding chest pass, overhead forward throw, squat to press throw. Both sides of the body work together, the athlete learns tofinish the throw with the legs, and distance becomes measurable.Stage 3 — Rotational sequencingStanding side throw and rotational scoop throw. The demand shifts to producing velocity through the hips first and letting itarrive at the ball last, which is the sequencing that underpins throwing, striking and punching.Stage 4 — Reactive and combination throwsPartner or wall rebound throws, drop-catch-throw variations, and throws combined with a step or a jump. The ball arrives beforethe athlete is ready, which adds a reactive component to the ballistic one.

Figure 3. Ballistic training fails when athletes throw hard before they can sequence. Each stage below establishes the coordination the next stage accelerates.

‌

‌

Ball mass, velocity, and the point of diminishing return

Increasing ball mass increases the force required and decreases the velocity achieved. Somewhere in the middle, mechanical power output peaks (Cormie et al., 2011). That peak is what most ballistic training is aiming at, and it moves depending on the throw and the athlete (Kawamori & Haff, 2004).

‌

Choosing ball mass by intentFour stacked rows relating ball mass to the training intent, from very light high-velocity work at the top to heavy low-velocity work at the bottom.Choosing ball mass by intentVery light — 1 to 2 kgMaximal velocity work and rotational speed. Useful for striking and throwing athletes late in apreparation phase, but offers almost no resistance to organise against, so technique must alreadybe sound.Light — 3 to 4 kgThe default for most rotational and upper-body throws. Fast enough to remain genuinely ballistic,heavy enough that the athlete can feel the sequence.Moderate — 5 to 7 kgWhole-body throws such as backward overhead and squat to press. Velocity drops but total impulserises, which suits general power development.Heavy — 8 to 12 kgSlams and short-range throws only. At this mass the movement is no longer high velocity, and itsvalue is eccentric tolerance and gross force rather than power.Rising velocityRising mass

Figure 4. Ball mass is the main dial you have. Too light and the athlete cannot feel the resistance to sequence against; too heavy and the throw stops being ballistic and becomes a slow lift performed badly.

‌

For rotational and upper-body throws, the practical window for most trained adults is three to five kilograms. Below that, the ball offers so little resistance that the athlete cannot organise a sequence against it, and the drill becomes an arm flick. Above about six kilograms, rotational velocity collapses and the athlete begins to muscle the ball, which defeats the purpose.

‌

Whole-body throws such as the backward overhead tolerate more mass because the whole kinetic chain contributes. Slams tolerate the most because velocity is not the point; the slam is largely about producing and then absorbing force through the trunk.

‌

• Velocity-biased One to three kilograms. Late preparation phase, striking and throwing athletes, in-season maintenance where fatigue must stay low.
• Power-biased Three to five kilograms. The default working range for most athletes and most throws.
• Force-biased Six kilograms and above. Whole-body extension throws and slams. Expect distance and velocity to drop and treat it as a different exercise, not a heavier version of the same one.

‌

A simple field method for finding the right ball: Throw the same drill with three ball masses and measure distance. Multiply an estimate of ball velocity by mass to approximate momentum, or simply take the mass at which distance begins to fall off sharply and work just below it.

‌

‌

Where the transfer evidence is strong and where it is thin

Medicine ball throws are frequently sold as a direct route to a harder punch, a faster bat swing, or a longer drive. The honest position is more qualified than that.

‌

The strong claims are these. Throw distance is a reliable and valid measure of upper-body and rotational power, and it responds to training (Stockbrugger & Haennel, 2001). Ballistic training does increase rate of force development and peak power in the trained pattern (Seppänen, 2021). And throws provide a high-velocity stimulus at very low joint-loading cost, which is genuinely valuable.

‌

The weaker claims concern transfer to sport skill. A rotational throw shares a general sequencing pattern with a punch or a swing, but it does not share the timing, the constraints, the load path, or the perceptual demands. The evidence for improvement in skill outcomes from medicine ball work alone is modest, and where it exists it is usually in athletes who were doing very little power work to begin with (Cronin & Sleivert, 2005).

‌

The defensible position: Medicine ball throws are an excellent general power tool and a legitimate way to load rotation. They are a supplement to sport practice, not a shortcut around it. An athlete who wants to punch harder should punch, and should also throw.

‌

• Well supported: Throw distance as a power test, improvements in throw performance itself, low-cost power stimulus.
• Moderately supported: Contribution to general rotational power in athletes with a limited training history.
• Poorly supported: Direct, substantial improvements in complex sport skills from throwing alone.

‌

Building throws into the training week

Throws are neurally demanding and metabolically cheap, which dictates where they belong. Place them early in the session after the warm-up, before any fatiguing work, and treat the set as finished when output drops rather than when the rep count is met (Haff & Nimphius, 2012).

‌

Typical prescriptions run from three to six sets of three to six repetitions with generous rest, which is far fewer repetitions than most people perform. Twenty repetitions of a rotational throw is a conditioning circuit, not a power session, and it will produce a conditioning adaptation.

‌

1. Warm up thoroughly, including submaximal throws in the same pattern.
2. Perform the primary throw first, 3 to 6 sets of 3 to 6, with 90 to 180 seconds between sets.
3. Measure or mark at least one set so output is visible.
4. Stop the exercise when distance falls by more than roughly five per cent.
5. Follow with strength work, not before it.

‌

A note on complexes: Pairing a heavy strength lift with a throw in the same pattern is a legitimate approach, and the throw generally goes after the lift with several minutes between. The effect is variable between individuals, so measure rather than assume.

‌

An eight-week medicine ball block

This block assumes two throwing sessions a week alongside normal strength training. It is written for a general athlete; a striker or a throwing athlete would weight it further towards rotation.

‌

Weeks one and two: Positions

All throws performed from kneeling or half-kneeling with a three-kilogram ball. The legs are deliberately removed so the athlete cannot compensate for a poor trunk position with leg drive.

‌

• Kneeling chest pass: 4 sets of 5.
• Half-kneeling rotational throw: 3 sets of 5 per side.
• Kneeling overhead throw: 3 sets of 5.

‌

Weeks three to five: Standing power

Standing bilateral throws enter. Distance is measured and recorded from week three onwards so the block has an outcome.

‌

• Standing overhead forward throw: 4 sets of 4, measured.
• Backward overhead throw: 4 sets of 3 with a 5 kg ball.
• Standing rotational scoop throw: 4 sets of 4 per side with a 4 kg ball.
• Squat to press throw: 3 sets of 5.

‌

Weeks six to eight: Velocity and reactivity

Ball mass drops, velocity intent rises, and reactive elements are introduced. Volume falls accordingly.

‌

• Wall rebound rotational throw: 4 sets of 5 per side with a 3 kg ball.
• Partner drop-catch-throw: 3 sets of 4.
• Step-and-throw rotational: 4 sets of 3 per side, measured.
• Retest the standing overhead throw and the rotational throw at week eight.

‌

Across all eight weeks, keep the total number of high-intent throws per session between forty and eighty. The upper end is generous. If an athlete finishes a throwing session feeling metabolically wrecked, the session was programmed as conditioning by accident.

‌

‌

Sport applications

Striking athletes in boxing and mixed martial arts benefit most from rotational throws with light balls, performed at high velocity and low volume, because the sequencing rehearsed is the same hip-to-shoulder pattern that generates a punch (Earp & Kraemer, 2010).

‌

Throwing athletes in baseball, cricket and handball use overhead and rotational throws, but the shoulder loading needs care. Very light balls thrown at maximal velocity can add meaningful stress to an already heavily loaded shoulder, and volume must be coordinated with sport practice rather than added on top of it.

‌

Golf, tennis and racquet athletes gain from separation-focused rotational work, where the delay between pelvis and ribcage rotation is coached explicitly.

‌

Field and court athletes use whole-body throws such as the backward overhead as a general power stimulus and as a simple, repeatable test of total-body impulse.

‌

Athletes managing lower-limb load use throws as a substitute power stimulus when jumping and sprinting volume must be reduced, since the ground reaction forces involved are a fraction of those in plyometric work.

‌

Common mistakes

• Throwing for reps instead of for output. Fifteen or twenty repetitions turns a power exercise into a conditioning circuit. The adaptation follows the actual stimulus, not the label at the top of the programme.
• Using a ball that is too heavy. Above roughly six kilograms, rotational velocity collapses and the athlete muscles the ball. The movement stops being ballistic in any meaningful sense.
• Leading the rotational throw with the arms. The most common technical fault by a wide margin. The hips must lead the ribcage, which must lead the arms. Film from overhead to confirm.
• Placing throws at the end of the session. Power output is fatigue-sensitive. A throw performed tired trains a slower pattern and produces a worse measurement.
• Never measuring anything. Distance thrown is free to collect and is the simplest low-cost outcome, but it is not the only one. Radar or release velocity, a power estimate from a linear transducer, standardised video, and a sport-transfer test such as bat or club speed all answer the question, and each needs its own fixed conditions — same ball mass, same throw type, same run-in, same surface, same camera position — before the number means anything. Pick one and hold it steady. Without some measure, the exercise is a hopeful ritual.
• Expecting throws to replace sport practice. Transfer to complex skill is modest. Throws build the underlying power quality; the skill still has to be practised in its own context.
• Ignoring the deceleration cost of slams. Heavy slams involve a large eccentric trunk demand. They are a legitimate exercise but they are not a low-cost one, and they should not be piled on top of heavy trunk work.
• Only training one direction of rotation. Athletes with a dominant side will happily throw hundreds of repetitions to their preferred side. Balance the volume even where the sport is asymmetric.

‌

Coaching cues

• "Back foot first" to establish ground-up sequencing in rotational throws.
• "Let the hips beat the shoulders" to create separation.
• "Throw it through the wall, not at the wall" to encourage acceleration through release.
• "Arms last" to stop upper-body initiation.
• "Finish tall" on overhead and extension throws.
• "Every throw is a maximum" to keep intent high and volume honest.
• "If the distance drops, the set is over."
• "Catch soft, throw hard" for reactive and rebound variations.

‌

FAQs

How heavy should my medicine ball be?

For most rotational and upper-body throws, three to five kilograms. Lighter than three and there is too little resistance to sequence against; heavier than about six and velocity collapses. Whole-body throws such as the backward overhead tolerate more, and slams are the only category where eight kilograms and above is genuinely appropriate.

‌

Are medicine ball throws better than Olympic lifts for power?

They are different tools. Olympic lift derivatives load the whole body at high force and moderate velocity and require significant technical instruction. Throws load a wider range of planes, including rotation, at higher velocity and lower force, and take minutes to teach. Most programmes benefit from both, and an athlete who cannot access good coaching for the Olympic lifts will get more from throws.

‌

Can medicine ball training make my punch harder?

It can contribute, but it is not a substitute for punching. Rotational throws rehearse the ground-up sequencing that a punch relies on and develop rotational power, and there is reasonable support for improvements in that underlying quality. Direct evidence that throwing alone substantially increases punch force in trained boxers is much thinner. Do both.

‌

How many throws should I do in a session?

Between forty and eighty high-intent throws is a generous upper range for most athletes. Three to six sets of three to six repetitions per exercise, across two or three exercises, with real rest. If you are not resting, you are conditioning.

‌

Should I use a bouncing or non-bouncing ball?

Non-bouncing for slams and for any throw into the floor, for obvious safety reasons. Bouncing balls are useful for wall rebound work where you want the ball to return, and for partner reactive drills. Many athletes need one of each.

‌

Can I do medicine ball work every day?

Light rotational work at low volume is tolerable at high frequency, but there is little reason to do it. Power qualities respond to quality and recovery rather than accumulation. Two or three sessions a week is sufficient, and daily maximal throwing will simply degrade output.

‌

Is throwing safe if I have a history of low back pain?

Rotational throwing loads the trunk at speed, so it should be introduced carefully and progressed slowly in anyone with a history of back pain, ideally with guidance from a clinician familiar with the case. Start from a half-kneeling position with a light ball, which reduces both the range and the velocity, and build from there.

‌

What is a good throw distance?

There is no universal standard, because distance depends on ball mass, throw type, technique and body mass. The useful comparison is against yourself under identical conditions. Fix the ball, the throw and the surface, then track the number over months.

‌

Recommended videos

Each video below was chosen because it covers a specific part of this article in more depth than text alone allows.

‌

‌

‌

‌

‌

‌

Related reading on FitXplor

• 2.5 Power Development and Explosive Strength
• 3.1 Introduction to Plyometrics
• 3.2 The Stretch-Shortening Cycle
• 3.11 Hopping, Bounding, and Horizontal Plyometrics
• 3.8 Jump Training and Vertical Power
• 4.7 The Spine and Core: Bracing, Stiffness, and Force Transfer
• 4.5 The Hip: Mobility, Stability, and Power Transfer
• 2.1 Principles of Strength Training

‌

References

Newton, R.U. et al. (1996). Kinematics, kinetics and muscle activation during explosive upper body movements. Journal of Applied Biomechanics.

Cormie, P., McGuigan, M.R., Newton, R.U. (2011). Developing maximal neuromuscular power. Sports Medicine.

Kawamori, N., Haff, G.G. (2004). The optimal training load for the development of muscular power. Journal of Strength and Conditioning Research.

Stockbrugger, B.A., Haennel, R.G. (2001). Validity and reliability of a medicine ball explosive power test. Journal of Strength and Conditioning Research.

Ikeda, Y. et al. (2007). Comparison of the effect of different throwing techniques on medicine ball throw performance. International Journal of Sports Medicine.

Earp, J.E., Kraemer, W.J. (2010). Medicine ball training implications for rotational power sports. Strength and Conditioning Journal.

Putnam, C.A. (1993). Sequential motions of body segments in striking and throwing skills. Journal of Biomechanics.

Seppänen, S. et al. (2021). The effect of ballistic training on athletic performance: a systematic review. Journal of Human Kinetics.

Cronin, J., Sleivert, G. (2005). Challenges in understanding the influence of maximal power training on improving athletic performance. Sports Medicine.

Chaabene, H. et al. (2015). Physical and physiological attributes of elite boxers. Journal of Strength and Conditioning Research.

Behm, D.G., Sale, D.G. (1993). Intended rather than actual movement velocity determines velocity-specific training response. Journal of Applied Physiology.

Haff, G.G., Nimphius, S. (2012). Training principles for power. Strength and Conditioning Journal.