Weighted Baseball Research and the Data Supporting Their Use

With the recent release of ASMI’s Biomechanical Analysis of Weighted-Ball Exercises for Baseball Pitchers we decided that it was appropriate to discuss its findings in a more in depth review.

The first part of this article is going to look at the details of the ASMI study itself. Our commentary on the findings in regards to training can be found below in the ‘Response’ section.

For the scope of this article we are only going to discuss the findings of pitching weighted baseballs off the mound and the crow hop throws, otherwise known as running throws or pulldowns.

Holds were included in this study and we did use them previously in our program but we have not used them since 2012. If you are interested why you can read more here, here, here, or here (video). The short answer: We found wrist weight exercises first used by Dr. Mike Marshall, and traditional small medicine ball rebounder exercises to be much more effective.


Review of ASMI’s Weighted Ball Study

What did we know about weighted balls before?

If you go back far enough we didn’t think weighted balls ‘worked’ either; through experimentation we found out that they were effective training tools at developing velocity.

The exact mechanism for how they increase velocity is unknown. But Soviet Sports Science research on over/underweight implements provides the hypothesis that they can be seen as a form of ‘speed-strength’ or ‘power’ training for the arm.

Besides this new ASMI study we have 6 weighted balls studies referenced in our pitching research review. We encourage everyone who is interested to find them and read them in full.

Coop DeRenne did much of the weighted baseball research previously and his paper Effects of Baseball Weighted Implement Training: A Brief Review provides us with this lovely summary chart.


So weighted baseballs, both under and overweight, have been shown to increase velocity. What wasn’t understood was how stressful weighted balls were in comparison to a regulation 5 oz baseball.

Examining the ASMI study, the gritty details

ASMI wanted to test out how stressful weighted balls were when throwing them off the mound and during crow hop throws. So they took 25 pitchers (18 high school and 7 collegiate pitchers) who had previous experience with weighted ball throwing program and had them throw weighted baseballs weighing between 4-7 oz. They didn’t describe what experience they had but they had some familiarity with weighted balls.

The authors hypothesis was that “ball and arm velocities would be greater with lighter balls and joint kinetics would be greater with heavier balls.”

In other words, the 4 oz ball would be thrown at a higher velocity and the arm would move faster, throwing the 6 and 7 oz would result in higher stresses on the arm.

We use to believe that underload balls were less stressful, after years of internal data we now believe they have a higher peak stress and are more stressful when compared to a 5 oz ball. We have internally believed that overload balls are less stressful but that belief has not been echoed in the baseball community.

The Methods

The researchers took 3 trials of 10 difference exercises (pitching 4-7 oz balls off a mound, crow hop throws with 4-7 oz balls and flat ground holds with 14 and 32 oz balls). Each player was allowed to use whichever crow hop technique they preferred.

NOTE: The researcher’s often referred to the flat-ground crow hop throws as “flat ground throws.” For the sake of this article I will be referring to those throws as crow hop throws to not get them confused with flat ground pitching, which was not included in this study. We use Pulldowns in our training and I will reference crow hop throws / pulldowns later interchangeably.

16 position, 5 velocity and 5 kinetics values were measured for a total of 26 biomechanical parameters computed per each trial. They were measured by using 38 reflexive markers which were tracked by 12 cameras sampling at 240 Hz.

Each pitcher was able to throw as many warm up pitchers as they felt were necessary. All participants were told to perform each exercise at maximal effort and the order of the exercises and ball weight was randomized for each participant.

Quick aside to take a moment to realize how much work this was. 3 trials x 10 exercises = 30 recordings per person x 25 people = 750 times those 26 biomechanical parameters needed to be computed. Plus those markers can sometimes fall off when collecting data! The point is: This was an impressive amount of data to collect and crunch.

The Results

In general as ball mass increased, elbow and shoulder joint torques and forces decreased for both mound and crow hop throws.

As ball mass increased  angular velocities of pelvis, upper trunk, shoulder and elbow decreased for both mound and crow hop throws. The exception was the 4 oz ball, which was not significantly different than a 5 oz ball.

The data contradicts the authors’ hypothesis that joint stresses would increase with ball weight.

What's a N-m (Newton-metre) again?

Newton metre (N-m) is a unit of torque. Torque is a moment of force that rotates an object, torque can also be thought of as a  measure of how much force is acting on an object that causes rotation. So the Newton-metre is a number that gives us the stress on a joint as it is rotating


Weighted balls off the Mound

The 6 and 7 oz balls were less stressful than a 5 oz ball for both elbow and shoulder stress, suggesting that as ball weight increases, arm stress decreases. The exception to this was the 4 oz ball, which was also found to be less stressful than a baseball in the elbow and shoulder off the mound.

The highest velocity was the 4 oz ball and with each increase in weight velocities decreased.

Looking at mechanical differences just between 4 – 7 oz balls off the mound. Few statistical differences were found between positions when comparing the 4 and 7 oz throws.

“Although there were some statistical differences in body positions when pitching balls of varying mass, the magnitudes of these differences were small (about 1°) and probably of little clinical relevance. Thus, it appears that pitchers can train with their normal mechanics when pitching 4- to 7-oz baseballs from a mound.”

The differences in mechanics when pitching off a mound with weighted balls were irrelevant.

“pitching baseballs that are slightly underweight or overweight (4-7 oz) produces variations in arm kinetics, variations in angular velocities, and relatively small changes in body positions; therefore, these exercises may be reasonable for training pitchers.”

In short, when comparing the different ball weights off the mound there were slight differences in kinematics but those DID NOT come with a noticeable difference in kinetics (joint forces/stress).

Weighted balls for crow hop throws

The elbow stress of crow hop throws with a 4 and 5 oz ball were identical (91.7 Nm). As ball weight increased to 6 and 7 oz, elbow stress decreased. Shoulder stress was highest in the 4 oz and decreased as ball weight increased.

The highest velocity crow hop throw was with the 5 oz ball. 4, 6, and 7 oz throws were all thrown at slower velocities than the 5 oz.

Comparing crow hop throws to mound

“As flat-ground throws produce increased shoulder internal rotation velocity and elbow varus torque, these exercises may be beneficial but may also be stressful and risky”

When compared to pitching off a mound, the crow hop throws produced greater elbow and shoulder stress and less elbow flexion torque, most likely because of the added running start.

When comparing balls of the same weight, crow hop throws were anywhere from 1.5-3 N-m more stressful on the elbow on average. The crow hop throws were 1.4-3.2 Nm more stressful on the shoulder on average. This supports our own hypothesis that crow hop throws are either equal to or more stressful than pitching a 5 oz off the mound.

Even though the crow hop throws were more stressful the authors also said:

“differences in elbow and shoulder kinetic values between flat-ground throws with a 5 oz ball and standard pitching were minimal and statistically nonsignificant”

This suggests that max effort throws off the mound with a 5 oz ball are similar in stress to crow hop throws – even when at different velocities.

There were other kinematic differences between pitching from the mound and the crow hop throws – seen specifically in arm, trunk and leg positions at front foot contact and trunk and leg positions at ball release.

A high-speed video clip of Trevor Bauer throwing vs. pitching can illustrate some of those differences:


Is this a big or small difference? You decide

Response To ASMI’s Weighted Ball Study

Averages, Peaks, And Individual Differences

The numbers from studies like this are fantastic but we can’t forget that we can sometime lose sight of the individual athlete when dealing with averages such as this.

It is likely that some athletes in this study experienced significantly higher stress during the crow hop throws while others might have experienced less stress during the crow hop throws. There also could have been bigger individual differences between balls themselves that we do not know of. At Driveline one factor we consider when doing weighted baseball programming for athletes is to examine weighted ball velocity spreads weekly (comparing velocities of each weighted ball thrown) and we’ll make training adjustments based off those variations. Clearly there are other factors (overall strength, ROM issues, mobility issues, fatigue, etc) as well.

Looking at velocity changes and monitoring athlete response is the key to using weighted balls.Click To Tweet

Coaches must be adaptable to how their individual athletes are responding to a weighted ball program.

Velocity And Stress

Shouldn’t baseballs that are thrown harder automatically cause more stress on the elbow?

The crow hop 5 oz throws were 7.8 MPH faster but only 1.5 Nm more stressful on the elbow and 1.4 Nm more stressful on the shoulder

Remember the authors found the differences in stress levels with the 5 oz throws as “minimal and statistically nonsignificant”, yet the velocity difference was quite large at 7.8 MPH.

This speaks more to the fact that we don’t yet fully understand the relationship between velocity and arm stress because there are so many variables to account for – also remember that there is no solved set of mechanics and training protocols that automatically cause velocity increases. The adaptation is different for each individual.

This would suggest that even though crow hop throws can be thrown at much higher velocities, the stress may not linearly increase as velocity increases.

Elbow stress in pitching may not linearly increase as pitching velocity increases.Click To Tweet

The difference between the 5 oz mound velocity and 5 oz crow hop velocity (7.8 MPH) is very large. This last summer we found the average difference to be 5.35, with a sample of 73 pitchers. Our closest comparison to this study is the 9 athletes we had throw 100+ MPH crow hop throws who had an average mound velocity of 92.5. Every other velocity bracket in our summer training group had a much smaller difference.

The spread (or difference between the 4 – 7 oz throws) was also much larger than what we normally see in gym for crow hop throws. The differences between ball weight in our gym for crow hop throws are much closer to the differences in velocity that were found off the mound – 3 MPH.

The velocities of both the mound and crow hop throws are also not what we would consider elite level velocities. This can partially be explained by the sample size with more high school athletes (18) than college (7). Therefore there may be differences in these findings if it was replicated with pitchers who throw 90+ off the mound.

Thoughts On Variation In Training

Introducing variation in training is a common theme in motor learning, talked about in Frans Bosch’s book Strength Training and Coordination: An Integrative Approach

“…learning a movement does not mean learning how to perform it in an ideal manner which is fragile and only usable in a single incidental environment, but how to apply numerous variations on a theme in order to create a movement plan that can withstand a variety of environmental perturbations.”

Remember the data this study shows suggests that “pitching slightly underweight and overweight baseballs produces variations in kinematics without increased arm kinetics (joint stress), these exercises seem reasonable for training pitchers.”

If you are looking to introduce variability in order to increase your kinesthetic awareness and proprioception pitching weighted balls off a mound may be a great training stimulus for that.

One of the more natural thoughts after this is we may be able to look to weighted balls off the mound not only as a trainer of velocity but also a trainer of command.

Changes Besides Arm Training

The researchers found statistically significant differences (p<0.05) among ball weights between both pelvis angular velocity and upper trunk angular velocity.

Lets look closer as some of the quotes earlier, edited for emphasis.

“Pitching baseballs that are slightly underweight or overwieght (4-7 oz) produces…variations in angular velocities, and relatively small changes in body positions…some statistical differences in body positions when pitching balls of varying mass, the magnitudes of these differences were small (about 1º) and probably of little clinical relevance. Thus, it appears that pitchers can train with their normal mechanics when pitching 4- to 7-oz baseballs from a mound.”

Suggesting the mechanics of the pitchers were nearly the same but the velocities that the body moved at to reach those positions were different with the overload balls.

These findings run counter to the claim that weighted balls only train the arm.

It give credence to the idea that even small changes in weight (such as 1 oz) make the whole body slightly change how it tries to complete the task of throwing a ball to a target.

We hope that this relationship is investigated further in the further to see if those angular velocities continue downward as ball weight increased above 7 oz.

Arm Stress And Underload Implements

There was no statistical difference in elbow stress between the 4 and 5 oz baseballs and the 4 oz crow hop throws averaged 4.3 MPH slower than 5 oz throws.

The most likely reasons for this were every participant threw the weighted balls in a random order and the athletes training history. Although these athletes did have previous experience with weighted balls we do not know how long they used them or what program. We’ve seen in our program that first-time users of underload implements have much smaller velocity differences between a 4 and 5 oz ball.

Below are the weighted ball pulldown spreads from 61 of our college summer athletes. You can see that the baseline spread between the 4 and 5 oz balls is 1.75 MPH. 14 of these 61 pitchers (23%) threw the 4 oz ball slower than the 5 oz ball in their baseline testing.

You can find the full numbers here under the ‘baseline’ columns.

Note: The peak velocities are not perfectly weighted. Some threw more pulldowns than others and the peak of each ball can happen at different weeks. But the data itself is interesting to look at.

This completely contradicts the claim that underload balls “artificially accelerate” the arm. The reality is the body takes time to learn how to move the kinetic chain in a more efficient way resulting in a faster arm and higher velocities. It often takes time for an athlete to learn to move more efficiently and throw underload balls faster.

It is generally understood that underloads balls are more stressful because they can be thrown at higher velocities making then more stressful, which is true. Often overlooked is the training value and stress that underload balls bring to a pitchers decelerators. The arm can be moving at faster speeds with a 4-oz ball but you have the same amount of time to decelerate as you would with a 5 oz ball.

So a benefit to including underload training to a throwing program would be training at increased arm speed and training arm deceleration, making your body more efficient at absorbing higher forces.

Arm Stress and Overload Implements

Here is what the authors of this paper had to say in regards to finding the stress of overload balls to be less than a baseball:

“The presumption that exercises with heavier balls demand greater torques and force about the elbow and shoulder was in general not supported by the current limited study.”

“Based on Newton’s second law, the force required to move a ball is equal to the mass of the ball multiplied by its acceleration. Although the heavier balls have more mass the decreased force and torques imply that these exercises had less arm acceleration.”

Force = Mass • Acceleration

So even though the mass of the ball increase, it reduced the arms acceleration in a way that resulted in lower stress numbers.

Force on the elbow = Mass (7oz) • Acceleration (?)

Our hypothesis on weighted ball stress has been that throwing overload balls is less stressful than throwing a regulation baseball. This is also why both our PlyoCare balls and weighted balls contain more overload balls than underload.

Although this study only went up to a 7 oz ball, we still believe that overload balls are less stressful than throwing a regular baseball. 7 oz is the heaviest weighted ball we use in pulldowns and it had the lowest elbow stress when compared to the 6-4 oz from both the mound and crow hop throws.

Dr. Fleisig talked further about these findings here.

Even though this study suggests that 6 and 7 oz balls are less stressful than a baseball this will most likely be the new critique of overload balls:

Even though overload balls have ‘less’ peak stress aren’t they actually more stressful because your arm moves slower so the stress is longer and different and therefore worse?

In other words, if it takes the arm longer to travel from max external rotation to ball release then wouldn’t that mean that total stress increased over that time?

It’s hard to say conclusively yes or no because it depends on how you want to measure ‘total stress’.


What we would need to know what the stress is at any given point throughout the pitching motion. During arm lay back into external rotation, at maximum external rotation (often but not always the peak stress), and then the stress as the ball is accelerating at ball release. Measuring elbow stress in milliseconds – which would be incredibly difficult.

An often overlooked point to this argument is that it isn’t the arm velocity but the arm acceleration that we would need to know. Once max external rotation is reached the arm is accelerating until it starts to decelerate; it doesn’t maintain a constant velocity.

Acceleration = Change in Velocity / Time

In order to know the acceleration we need to know the change in velocity during a specific period of time. We could calculate an average acceleration if we knew the time from max external rotation to ball release.

This would require extremely high-speed cameras perfectly synchronized. It’s possible that the cameras being used on a regular basis are insufficient for this purpose given the amount of variance between each frame of a video considering how fast the arm moves in the pitching delivery. 

This is a pretty hard claim to conclusively prove; it would involve a lot of math, agreement of when to start measuring stress, and some serious high frequency cameras to be able to calculate the load during the acceleration.

So you still think that PlyoCare balls are less stressful even though they are heavier than 7 oz balls?

Yes. Our plyo care work is primarily meant for submaximal throwing and for constraint training, as well as general arm care. They’re not meant for throwing as hard as possible with zero mechanical intent to change.

When the mass of the ball is increased, the acceleration is much lower, plus the majority of the PlyoCare throws are for mechanical work done at submaximal intent.

Secondly there is evidence that partial-effort pitching (60-80% of max effort) is significantly less stressful when compared to max intent pitching; there is no reason this can’t be applied to PlyoCare balls as well.

Crow hop / Pulldowns

Why still do running throws if they are more stressful?

  1.  To build intent, in order to throw hard you need to try to throw harder
  2. To train at stress levels near or above levels that pitchers would experience in game

The researchers had these thoughts on the crow hop throws:

“As flat-ground (crow hop) throws produced increased shoulder internal rotation velocity and elbow varus torque, these exercises may be beneficial but may also be stressful and risky.”

Again, our hypothesis before this study was that running throws equaled or surpassed the stress levels seen in a game, which is why they make a good training stimulus.

What we really want to do is stay away from long periods of submaximal stress and expect velocity improvements – which unfortunately is what a lot of the current physical therapist-led “throwing experts” are prescribing. Training pitchers for long periods of time at submaximal stress (offseason) and then putting them into situations where they may exceed previous maximal stress (in games) makes zero sense and is a good way to create serious injury potential.

Our pulldown (crow hop) throws are a programmed way – in the offseason – for athletes to push their stress levels at/above levels they would see in a game.

Crow hop throws with weighted balls are simply a tool to encourage velocity development through the stress response cycle, utilizing supercompensation and the body’s adaptation to training stimulus.


Crow hop/pulldowns are a programmed stress on the body in order to encourage positive adaptation (velocity development) over a long period of time.

Training stimulus: weighted ball pulldowns

Fatigue: from weighted ball training

Recovery: period between high-output training sessions

Supercompensation: faster velocities (Hopefully!)

That is why programming is so important. We want to stay away from high chronic training loads (overtraining) and low chronic training loads (undertraining) – otherwise we run the risk of what Tim Gabbett calls a training load error. A training load error can be seen as a mismanagement of stressors placed on the body, throwing off the Acute : Chronic load ratio.

If stress is chronically too low and then has a huge spike, we may be undertraining for what we want to do: Pitch at high intent levels to get professional hitters out! If the stress is chronically high and not properly managed, then we may be overtraining.

Unfortunately at this current time we lack a lot of ability to track training and pitching loads long term since real-time biomechanical feedback and wearable units are not yet that popular (or even allowed) in professional baseball. Pitch counts and pitch usage can bridge some of that gap for now, but technological embrace by MLB will be required to make meaningful progress in this area.

Pulldowns are ONLY meant for velocity development in the offseason and in order to do them you must be following an individualized and prescribed throwing schedule. We would never recommend doing pulldowns in season or doing them outside of a scheduled throwing program.

Things the Future Holds

We have jokingly pointed out on Twitter that in a very quick time we have gone from:

  • Weighted balls don’t work to develop velocity
  • Weighted balls do develop velocity BUT they’re very dangerous
  • Well it’s not the tool that’s bad it’s how you use them

Needless to say, this has been a surprising change in such a short period of time and that comes with both good and bad things to look forward to. There are going to be more studies on weighted balls with different methods and various conclusions in the future and that’s what is supposed to happen. Baseball training will change not only off of what future peer-reviewed research suggests but also off of experimentation by coaches and facilities, preferably with rigid data collection. Of those two the experimentation and tracking of results is probably the most important, since most people aren’t all that interested in reading published studies on baseball science.

The future of training is the ability to use peer-reviewed research to back your training, lead solid experimentation groups, and being data-driven. This is the difference between “this study said X and it applies to everyone” and the need for smart trial-and-error based experimentation using peer-reviewed research as inspiration!

In the end what does this new study prove? Nothing, no study proves anything.

A better question is: What does is suggest or what does it support?

Therefore these (weighted baseball) exercises may be reasonable for training pitchers. -ASMIClick To Tweet

ASMI’s study gives evidence that weighted balls aren’t inherently “significantly more dangerous” and that, if programmed correctly, they can be a benefit to a pitchers training. This study is particularly exciting to us because we get to replicate it, and indeed we are already on our way! We have been collecting Motus mTHROW data, high-speed video using our Edgertronic cameras, and kinematic data in our lab on pitchers when they pulldown so we can compare it to their mound work.

We look forward to continuing our own research on our program using our biomechanics lab and Motus sleeve among other technology available to track and measure athlete results.

This post was written by Michael O’Connell, Research Assistant in the R&D Department.

Felix Townsend

I’ve never been against weighted ball use. I feel like they are fantastic from a health standpoint as well. My thing is with the running throws and transfer of training. It’s pretty clear by the ASMI study and you’re summer program data collection that there is a pretty significant gap between velocities of the mound and the run. Seeing as how stress is nearly the same regardless of run or pitch, do you think more time can be spent doing max effort weighted ball throws off the mound instead of running? Maybe to get a little better carryover of training. Or will running throws always create significant increase in pitch velo. My question is in regard to the actual difference in running and pitching. Not weighted balls. Just the movement itself.

Driveline Baseball

Absolutely, it’s something we are working on right now. We’re testing out more weighted ball work on the mound this offseason as athletes move closer to needing to throw off the mound. We’ll make sure to let everyone know how it goes!

How To Start A Weighted Ball Program – Pitch Mechanics

[…] incoherent. To read more on Michael’s thoughts and reaction to the lengthy ASMI study, read his lengthy blog post. Seriously, read […]

Does Arm Speed in Pitchers Matter? - Driveline Baseball

[…] The first claim is not supported in any peer-reviewed (or non-peer-reviewed) research paper or case study. The American Sports Medicine Institute (ASMI) published an exhaustive paper on weighted ball training that did not support that weighted ball training increased elbow varus torque. […]

Comparing Overload vs. Underload Weighted Ball Stress on the Arm - Driveline Baseball

[…] In ASMI’s breakthrough paper, Biomechanical Analysis of Weighted-Ball Exercises for Baseball Players, Dr. Fleisig et al. published a huge trove of data on “flat ground” (high effort crow hop) throws using weighted baseballs to examine the kinematic and kinetic differences on the arm. The data chart below shows that data, which first appeared in our blog post: Weighted Baseball Research and the Data Supporting Their Use: […]

Leave a reply