News

The Rotational Medicine-Ball Throw: A Real Signal, Not a Velocity Shortcut

By Ground Up Baseball · Training · Published Oct 29, 2026 · Last updated Oct 29, 2026

Safety first: this is general education, not medical advice. It is not a plan to load a young body. Keep training supervised, age-appropriate and light, with good technique. Stop for pain and see a doctor. Pitch counts and rest rules still apply (ArmRules).

A screenshot of a journal abstract has been going around, with a striking claim attached: a rotational medicine-ball throw explained 38 percent of the variance in pitching velocity in 35 Division III college players, while the standing broad jump and the lateral-to-medial jump did not.

We tracked down the paper. The short version is that the screenshot is accurate. The longer version is more useful.

The study is real, and it is one of the cleaner pieces of baseball evidence that a rotational throw test tells you something a jump test does not. It is also a correlation in 35 college players, and it does not show that throwing a medicine ball makes anyone throw a baseball harder. Those two sentences have to live together. This article is about holding both.

The broader case for building athleticism and rotational strength before fine-tuning mechanics is a separate argument, made in Build the Athlete First. This piece is about one test, one study, and what a family should do with it.

What a rotational medicine-ball throw is, and how this study measured it

A rotational medicine-ball throw is a sideways turn and throw. The player stands with the lead foot pointing at the wall, winds the hips and trunk away from the target, then unwinds and throws forward as fast as possible. It looks a great deal like the trunk action of a swing or a throw, which is the whole reason anyone tests it.

In the study the protocol was specific. Players stood with feet about shoulder width, lead foot in the direction of the throw, holding a 2.72-kilogram (6-pound), 32-centimeter medicine ball with the arms at waist level. They rotated hips and trunk about 90 degrees away from the target as a countermovement, then reversed and threw at a concrete wall as fast as they could, reaching roughly 90 degrees of rotation at release.

After two or three practice throws, each player took three consecutive test throws with about ten seconds of rest between them. Velocity was measured with a Stalker Pro II radar gun positioned five meters away and 1.3 meters off the ground, and the best throw was recorded. That is the entire test. No lifting, no sprinting, no baseball.

Two details matter later. The load is light: six pounds is a speed implement, not a strength implement. And the measurement is a radar reading of a maximum-effort throw, exactly the kind of test that invites a kid to turn it into a contest.

What the study found

Thirty-five NCAA Division III players took part: 15 pitchers and 23 hitters, with three players counted in both groups. The authors correlated each power test against bat swing velocity, batted-ball velocity and pitching velocity.

The rotational throw was related to all three:

  • Bat swing velocity: r = 0.65 (p = 0.003), explaining 39 percent of the variance.
  • Pitching velocity: r = 0.62 (p = 0.02), explaining 38 percent of the variance.
  • Batted-ball velocity: r = 0.53 (p = 0.02), explaining 28 percent of the variance.

A correlation of 0.62 to 0.65 is what researchers call moderate. It is not a fingerprint. Two players with the same medicine-ball number can differ by several miles per hour on the mound, and two with very different numbers can land in the same place. What it says is that this expression of rotational speed lines up with the sport's own expressions of speed far better than chance would produce.

Remember what this sample is: college players throwing a 6-pound ball. A parent who reads a much smaller number off a radar gun at youth practice should not panic. The comparison that matters is a kid against himself, next season.

The part that is genuinely interesting: the jumps did not track

The study also tested two jumps everybody uses: a two-legged standing broad jump, and a lateral-to-medial jump, meaning a sideways hop from one leg to the other.

Neither jump showed a statistically significant relationship with bat swing velocity or pitching velocity. Only the rotational throw was related to all three outcomes. The authors note this is the first study to put all three tests in the same players at the same time, which is what makes the comparison worth anything. Plenty of earlier papers had found one jump or one throw correlating with velocity, but rarely in the same room.

Why does that matter for a kid? Because it argues that power is not one thing. A jump measures force pushed into the ground in a straight line. A rotational throw measures force organized around a turning trunk. Baseball's skills are rotary, and this study suggests the rotary test is the one that lines up with a rotary sport.

It is also the part of the paper we would hold most loosely.

Test results disagree between samples. In 2013, in 42 college players, a lateral-to-medial jump was consistently related to throwing velocity across throwing techniques; that study concluded the frontal-plane hop mimicked the throwing stride well enough to matter. The Division III study did not reproduce it. Both can be right. They used different players, different tests and different throwing conditions, and 15 pitchers in the Division III sample is small enough that a handful of athletes can move the answer. The honest reading is not "jumping is useless." It is "in this group, in this battery, the jump did not carry the signal."

In kids, a jump may do better. In 49 youth players averaging 11 years old, a standing broad jump was the best predictor of batted-ball velocity beyond height and body mass, ahead of the grip, hip strength and range-of-motion models. Different population, different outcome, different analysis, but a useful corrective. A null result in college players is not a verdict on whether a nine-year-old should jump.

Our read, and it is our read: keep the jump work, because jumping is good for young athletes and cheap to coach, and keep the rotational throw, because it is the closest simple test we have to what baseball asks the body to do. Do not throw out either one over a single study.

What this study cannot tell you

This is where most write-ups of this paper go wrong.

  • It is cross-sectional. Everything was measured once. A cross-sectional study can tell you that fast medicine-ball throwers tend to be fast bat swingers. It cannot tell you which caused which, or whether either caused anything. It is entirely possible that the players who throw the ball hardest are simply the better, stronger, more mature athletes, and that the number is a symptom rather than a cause.
  • Thirty-five players is small, and 15 pitchers is smaller. Correlations at this size wobble. A different 35 players could produce a noticeably different coefficient.
  • It is one division at one level. NCAA Division III baseball is college-age, largely recruited, already skilled enough to be on a roster. Nothing in the paper is about children.
  • The abstract does not report the players' ages or an analysis by sex, and we will not guess at either.
  • Nobody trained anything. This is a testing study, not a training trial.
  • Thirty-eight percent leaves most of the story unexplained. Something else accounts for the rest, and the study does not identify it.

The practical translation: a good medicine-ball number is a reasonable sign that a player moves athletically in the right pattern. It is not a diagnosis, a projection or a promise.

The distinction we will not blur

This is where families get hurt, so we will be clear. A medicine-ball throw and a weighted-baseball throwing program are not the same thing, and they are not close.

A medicine ball is a two-handed implement thrown at a wall. The arms stay roughly together, the throw is short, the ball is light relative to what the body can produce, and there is no baseball, no mound, no catcher and no game intent involved. The American Academy of Pediatrics lists medicine balls alongside free weights, kettlebells, elastic tubing and body weight as a mode of youth resistance training, and says children and adolescents can gain strength with low injury rates when activities are well supervised with an emphasis on proper technique. A 2008 program-design paper for coaches of school-age kids concluded that medicine-ball resistance training can be a safe, effective and enjoyable experience for that age group.

A weighted-ball throwing program is a baseball, heavier than regulation, thrown with one arm at high intent, often at maximum effort, often off a mound or into a net, sometimes on a radar gun, sometimes in volume. The arm is doing the work under load. That is a different mechanical and tissue demand. We do not recommend it for young players, and we do not recommend maximal loads on a bar or in a hand either.

If a coach's medicine-ball drill involves a ball thrown with one hand, a radar reading and a personal best, that is not the exercise this study measured. Do not let "medicine ball" and "weighted ball" get used interchangeably in front of your kid.

How to do a rotational medicine-ball throw with a kid

Everything below is our general coaching suggestion, not medical advice and not a prescription for any one child. It matches the study's protocol in spirit, not in load.

The setup

  • A light ball. Six pounds is a college implement. For a 7- to 10-year-old we would start with a 1- to 2-kilogram ball, roughly 2 to 4 pounds, and go lighter if the throw looks labored. A kid who has to grunt to move the ball is training something other than speed.
  • A clear wall and a clear lane. A solid wall you do not mind hitting, nothing breakable behind or beside the thrower, and no other kids in the flight path. This is the most common way these sessions go wrong.
  • A coach or a parent watching. The AAP is explicit that supervision and technique are what keep youth resistance training safe. A medicine-ball wall is not a place for a kid to work alone.

The movement

  • Stand sideways to the wall, feet about shoulder width, the foot nearest the wall pointing at it.
  • Hold the ball with both hands at waist level, arms fairly straight, not cradled against the chest.
  • Turn away from the wall to load, letting hips and trunk rotate together. Do not wind up by twisting the low back against still hips.
  • Turn back and throw into the wall as fast as the kid can control. Hips lead, trunk follows, hands come through, feet stay planted.
  • Catch it or pick it up, reset fully, and go again.

Teach the turn at half speed with no ball, then with a gentle throw, then for speed. A kid who cannot turn without losing balance is not ready to throw for speed.

The dose

Short and fast beats long and tired. We would suggest three to five throws per side, two or three sets, with real rest between sets, once or twice a week inside a normal practice. That is our coaching judgment, not a tested protocol.

The governing rule: the last throw of a set should look like the first one. When the throw gets slow, sloppy or arms-only, the set is over. That is the entire principle behind a speed implement, and it is the one kids and coaches abandon first.

What we would skip

  • Throws for time, or a personal best every session. That turns a speed drill into a max-effort contest.
  • Heavy medicine balls on the theory that more weight builds more power. Six pounds was the college load here, and a smaller kid does not need more.
  • Rotational throws with a kid who has back pain, or any throw that hurts. The AAP notes that explosive contractions where a tendon attaches to bone may raise the risk of an avulsion fracture until a child is closer to skeletal maturity. That is the accurate version of the growth-plate concern, and it is a reason to keep load light and volume low, not to avoid the movement entirely.
  • Any medicine-ball drill that replaces arm care or rest. A wall throw is not a substitute for pitch counts, a throwing progression or time off. See ArmRules for the rules that actually prevent injuries.

How to test it, if you test it at all

You do not have to, and most kids never should. But if you want a number, here is how to make it mean something.

Keep the protocol boring. Same ball, same distance, same wall, same warm-up, three throws, take the best. A retest only means something if you compare like with like, and the fastest way to lose that is to change the ball or the stance between sessions.

Know what the test is worth. A 2025 systematic review pooled nine studies and found the medicine-ball throw test has good test-retest reliability in children and adolescents, with an intraclass correlation of 0.80. A kid tends to score similarly on a second day. Two caveats, though. The review covered chest throws and overhead throws, not the rotational side throw this article is about, so that reliability figure is borrowed from a cousin of the test. And reliability is not validity: the same review found only three studies examining whether the test relates to other fitness measures at all. The number is repeatable. Whether it means what you hope is another matter.

Do not turn it into a radar contest for a pitcher. We would underline this. The American Sports Medicine Institute's position statement for adolescent pitchers advises avoiding radar guns, right alongside its pitch count, innings and time-off guidance. A young pitcher chasing a gun number is the opposite of the conservative approach. If you test a rotational throw, treat it as an off-season athleticism check, keep the reading to yourself, and do not let it become a throwing-velocity scoreboard.

Write it down and leave it alone. A number in a notebook once a season is a baseline. One checked every week is anxiety for a kid and temptation for an adult.

So what do you do with the number?

Almost nothing, and that is the point.

If a young player's throw is fast for his age, that is a nice confirmation he moves well. Keep doing what you are doing: other sports, jumping, sprinting, a light ball, rest days. If it is slow, that is not a diagnosis and not a reason to buy a program. It is a reason to spend more of practice on general athleticism, which is where we would put the time anyway.

What we would not do is treat it as a forecast. The study's correlation leaves most of pitching velocity unexplained, and the participants were college athletes. Nobody has shown that raising a child's medicine-ball throw velocity raises his pitching velocity. If someone tells you they have, ask for the trial.

The training literature points at strength and power, not at a number. In 49 high school players, twelve weeks of added rotational and full-body medicine-ball work produced larger gains in torso rotational strength and a sequential hip-trunk-arm throw than the same program without the throws. In 60 primary school children aged 10 to 12, twelve weeks of supervised medicine-ball training twice a week produced a significantly larger gain in a backward overhead medicine-ball throw than the comparison group, and both jump tests improved over the twelve weeks. Both are real results. Neither measured pitching velocity, and neither is a youth baseball velocity study.

Where we land, as our view rather than a finding: the rotational medicine-ball throw is a good exercise, a decent and repeatable test, and a useful piece of a young athlete's week. It is not a velocity program, and a family should be suspicious of anyone selling it as one.

For the drills and weekly structure that put rotational work in its place, see Rotational Strength Training and the wider training library. For the longer argument about order, read Build the Athlete First.

Where the evidence is limited

  • The headline study is cross-sectional and small: 35 players, one session, correlation only.
  • No youth data. No participant was a child. Our PubMed search turned up medicine-ball work in young athletes, but nothing measuring rotational throw velocity against pitching velocity in kids.
  • No training trial in this paper. Nothing here tests whether medicine-ball work raises velocity, and we did not find a youth trial that does.
  • The jump null is narrower than it sounds. The abstract reports no significant jump relationships for bat swing velocity or pitching velocity. We did not see the full table, so we are not extending that to batted-ball velocity or quoting the coefficients.
  • The jump findings conflict between studies, which is a good reminder that one null result is not a law.
  • Youth test validity is thin. Reliability in children looks good for chest and overhead throws; evidence on validity, and on the rotational version, is limited.
  • The coaching guidance is judgment, not a tested protocol. Our rep and set suggestions are a conservative starting point, not a dose anyone has tested in children.

What would change our mind? A controlled trial in players under 13 that adds supervised rotational medicine-ball work to a normal practice and measures throwing velocity and bat speed a season later. If one shows a real effect, we will write about it.

Sources

We could not open the full text of the Taniyama paper. The publisher page and PubMed serve only the abstract to non-subscribers, and every other route we tried, including the DOI, ResearchGate, Google Scholar, Europe PMC and the Internet Archive, either returned the abstract or was blocked. Every correlation and the sample breakdown above were read from that abstract, and the ball mass, throw protocol and radar setup were read from publisher full-text passages retrieved in search of the article page. We never saw the full results table, so the nonsignificant jump coefficients are described qualitatively, as the authors report them, and are not quoted. The AAP and ASMI documents were opened and read in full. The Kinesiology review was read in full from its open-access PDF. The 2007 medicine-ball trial, the 2013 throwing-velocity study and the 2025 youth field-testing study were read as PubMed abstracts. The Trajković primary-school study and the Faigenbaum and Mediate program-design paper were read as abstracts only, so no specific program recommendations are attributed to either. The movement cues and dosing in the training section are our own coaching suggestion and are not sourced to a study. Links checked October 29, 2026.

Last updated Oct 29, 2026. General education, not medical advice. Talk to your child's doctor before starting a strength program, and about any shoulder, elbow or back pain. More news

Ground Up Baseball newsletter

One email for GloveCode, ArmRules, BatCode and BatRules: new glove and bat data, plus pitch-count and bat-rule changes. (Placeholder: signup is not connected yet.)