Walk into almost any facility and you will eventually hear someone talk about a pitcher's layback as if it were a velocity dial. The reasoning feels intuitive. More external rotation means a longer runway to accelerate the ball, so more mobility must mean more speed, right? We stretch, we mobilize, we chase degrees of passive range of motion as though every extra one buys a tick on the gun. A 2025 study, still unpublished, took fifteen collegiate pitchers and put that assumption on trial, and the verdict was not kind to the mobility crowd. Shoulder mobility did not predict how hard these pitchers threw. The ability to produce force did.

What the Study Found

The researchers measured three things and then asked which one actually tracked with fastball velocity. The first was passive external rotation, the layback metric, pure range of motion. The second was internal rotation power, which is rotational force production measured from the cocked position of 90 degrees of abduction and 90 degrees of external rotation, the exact place the shoulder lives during the throw. The third was mean peak bar velocity during a submaximal bench press, a stand-in for general upper-body power. Then they recorded maximum and sitting fastball velocity with radar and ran the numbers.

Internal rotation power was the runaway winner. It correlated strongly with maximum fastball velocity, with an r of .726 and a p of .001, and on its own it accounted for 54.6% of the variance in how hard these pitchers threw. To put that in plain terms, more than half of the difference between the harder and softer throwers in this group could be explained by a single measure of rotational force output. When all three variables were thrown into a regression model together, internal rotation power was the only one that survived as a statistically significant predictor, carrying a beta of .675. The model itself reached significance, but only because that one variable was doing all the work.

Now look at what fell flat. Passive external rotation, the thing we spend so much energy chasing, showed a weak and non-significant relationship with velocity, an r of just .268 with a p of .167. The mobility was simply not where the velocity lived. And the submaximal bench press told the same disappointing story, with an r of .407 and a p of .066, close enough to flirt with significance but ultimately contributing nothing once rotational power was in the room. To be honest, this reminds me of judging a car by how wide you can open the doors. The doors opening wide tells you the cabin is accessible, but it tells you nothing about what is under the hood.

I want to be fair to the study's limits, because they matter. It measured only one slice of force production, internal rotation power, and there are other variables it left untouched. Total internal rotation force, time to peak force, external rotation force, scaption strength, all of these could inform velocity development, and none of them were on the table. The bench press inclusion was odd from the start, since testing bar velocity at a submaximal load in a study about rotational shoulder strength never had a clear rationale, and predictably it went nowhere. And fifteen pitchers is a small group, so I would hold the exact percentage loosely. But the direction of the finding is the part that holds up, and it does not stand alone.

Why This Information Is Important

The reason this matters is that it is not a one-off. The pattern shows up again and again once you go looking, and the convergence is what should change how we train.

Start with the cleanest echo. Cross and colleagues, in a 2023 study in the Journal of Strength and Conditioning Research, took the same population, collegiate pitchers, and found strong positive correlations between isometric rotational strength and pitch velocity, with external rotation strength showing the strongest relationship of all. And in the same breath, they found no significant correlation between shoulder range of motion and either velocity or pitching kinetics. That is the source study's conclusion, replicated by a different group: strength predicts velocity, mobility does not. When two independent studies on the same kind of athlete land on the same divide, you are no longer looking at a fluke. You are looking at a principle.

But here is where I want to slow the celebration down, because the force story is more textured than a slogan. Job and colleagues, in a 2024 paper in the Journal of Biomechanics, asked a sharper version of the question, separating peak force from the rate at which force is produced. They found that maximal force from the internal and external rotators was modestly associated with throwing velocity, on the order of one to one and a half miles per hour for every 30 newtons of strength, while the rate of force development showed no clear link at all. Sit with that for a moment, because it cuts two ways. It supports the central claim, that force matters more than mobility. But it also refines it, because the relationship was modest, and it was maximal force, not explosive rate, that carried it. So if we are going to retire the obsession with range of motion, we should not replace it with an equally naive obsession over a single force number. Strength is a meaningful piece of the velocity puzzle. It is not the whole puzzle, and the returns appear to taper.

Zoom out from the shoulder and the picture gets more convincing, not less. Yee and colleagues, in a 2025 study in Medicine and Science in Sports and Exercise, looked at major league pitchers and found that vertical jump acceleration predicted 45% of the variance in fastball velocity, climbing to roughly 60% once they controlled for age, while height and weight predicted nothing. Think about what that means. A measure of explosive force production from the lower body explained more of the velocity story than a lot of the upper-body metrics we fixate on. This is the kinetic chain talking. The pelvis is the platform, the trunk is the conduit, and the arm is the receiver of energy generated from the ground up. If you cannot produce force rapidly through the legs and hips, the arm has less to work with no matter how mobile the shoulder is.

The companion study from Oster and colleagues, also in 2024 and also in major leaguers, sharpens the point in a way that speaks directly to that strange bench press inclusion in our source study. They tested pressing movements and found that acceleration predicted fastball velocity while raw power output did not, with even the best pressing measure explaining only about 23% of the variance. So general upper-body pressing has a modest relationship with velocity at best, and what little relationship exists lives in acceleration, not in how much load you can move. That is exactly why a submaximal bench press at a fixed percentage of a one-rep max was always going to disappoint as a velocity predictor. The arm is the end of the whip. It receives energy, it does not generate it from a bench, and testing it as though it does was asking the wrong question.

If you want something close to causal evidence rather than correlation, Gdovin and colleagues provide it. In their 2024 study, collegiate pitchers who kept up a consistent throwing program but lost access to resistance training over an eight-week stretch dropped about 6.3 miles per hour, a significant decline, with no change in arm slot, arm velocity, or elbow torque to explain it away. Take the force-building work out of the equation and the velocity leaks out, even when the throwing volume stays. That is the strongest argument yet that the strength underpinning force production is not a nice-to-have, it is load-bearing for the whole performance.

And the chain logic holds one link further in. Eilen and colleagues, in 2024, found that oblique strength correlated with pelvis and torso rotation velocity in adolescent pitchers, with the glove-side obliques showing a particularly strong relationship to pelvis rotation speed. Rotational strength in the core feeds rotational velocity in the delivery, which feeds the arm. The theme repeats at every level: force production, not passive range, is what distinguishes the engine.

So what do we do with external rotation range of motion, the metric that keeps coming up empty for velocity? We give it the job it is actually good at. Shitara and colleagues, in a 2025 study in The Orthopaedic Journal of Sports Medicine, found that left-handed high school pitchers with passive external rotation at or above 109 degrees were significantly more likely to get hurt, with injured pitchers averaging nearly 115 degrees against about 107 in the uninjured group, and notably no meaningful strength differences between them. Read that alongside everything else and the role of range of motion comes into focus. It is not a velocity lever. It may be an injury flag. Excessive layback is information about risk, not about speed, and that is a completely different conversation than the one most facilities are having when they chase mobility for performance.

How Can This Information Be Applied

The first and most direct application is to change what you measure. If you want to understand a pitcher's velocity potential at the shoulder, stop logging how many degrees of layback they have and start testing how much rotational force they can produce from that cocked 90/90 position. Internal rotation power from the position the throw actually uses is a far more honest window into velocity than any goniometer reading. This is the diagnostic mindset over the pass-fail mindset. The number that matters is not how far the shoulder can go, it is what it can do once it gets there.

The second application is to train the force-production system, not just the shoulder in isolation. The jump acceleration and pressing studies are a reminder that velocity is a whole-body output, so the lower-half and trunk force production that shows up in a vertical jump or a rotational core test is part of the same story. Building explosive force from the ground up, through the legs, hips, and obliques, is building velocity capacity, and the detraining study tells us in fairly stark terms what happens when we let that strength erode. You do not have to choose between throwing and lifting. The pitchers who stopped lifting lost six miles per hour while still throwing.

The third application is to be appropriately skeptical of generic gym tests as velocity proxies. A submaximal bench press at a fixed percentage of a one-rep max did not predict velocity in our source study, and the broader pressing research suggests that is the rule, not the exception. If you are going to test upper-body output and have it mean something, the evidence points toward explosive, acceleration-based measures rather than how much someone can grind out at a given load. Match the test to the quality you actually care about.

And the fourth application is to stop pathologizing or worshipping range of motion and instead file it where it belongs. Passive external rotation has clinical value, but its value appears to be in flagging injury risk, especially at the extremes, rather than in forecasting how hard someone will throw. Screen with it, monitor it, use it to have honest conversations about durability. Just stop treating an extra few degrees of layback as a velocity intervention, because the data simply does not support that role for it.

I do want to hold the nuance from the Job study through all of this, because it keeps us honest. The relationship between rotational strength and velocity, while real and repeatable, was modest in that study, and it may level off. Force production clearly matters more than mobility, but it is not an infinite dial either. The point is not to swap one single-variable obsession for another. It is to redirect our attention from a quality that does not predict performance, range of motion, toward a family of qualities that does, force production across the kinetic chain, while keeping our expectations calibrated to what the evidence actually shows.

Conclusion

The simplest way I can say it is this. Range of motion without the strength to express force through that range is close to useless for velocity. A shoulder that can lay back into a beautiful position and cannot drive force out of it is a wide-open door on a car with no engine. Across an unpublished study and a stack of published ones, the message rhymes: rotational and explosive force production predicts how hard pitchers throw, while passive mobility predicts almost nothing about it, and probably belongs in the injury conversation instead. So if you are still spending your assessment time measuring layback and your training time chasing degrees, you are optimizing a number that the radar gun has quietly stopped caring about. Measure what the shoulder can do from the position that matters, build the force that feeds it from the ground up, and let range of motion go back to the job it was always better suited for. We have been measuring the wrong shoulder number. The right one was force the whole time.

References

Unpublished study. Shoulder mobility, internal rotation power, and bench press velocity as predictors of fastball velocity in collegiate pitchers. 2025.

Cross JA, Higgins AW, Dziuk CC, et al. Relationships Among Shoulder Rotational Strength, Range of Motion, Pitching Kinetics, and Pitch Velocity in Collegiate Baseball Pitchers. Journal of Strength and Conditioning Research. 2023.

Job TDW 3rd, Cross MR, Cronin JB, et al. Relationship of Shoulder Internal and External Rotation Peak Force and Rate of Force Development to Throwing Velocity in High School and Collegiate Pitchers. Journal of Biomechanics. 2024.

Yee SR, Jensen MR, Azevedo AS, et al. The Relationship Between Vertical Jump Acceleration and Fastball Velocity in Major League Baseball. Medicine & Science in Sports & Exercise. 2025.

Oster AJ, Jensen MR, Azevedo AS, et al. The Relationship Between Unilateral Press Performance and Fastball Velocity in Major League Baseball Pitchers. Medicine & Science in Sports & Exercise. 2024.

Gdovin JR, Hogan B, Williams CC, et al. Limiting Access to Resistance Training Equipment During the Off-Season: The Impact on Collegiate Pitching Metrics. Journal of Strength and Conditioning Research. 2024.

Eilen HT, Kokott W, Dziuk C, Cross JA. Relationship of Abdominal Oblique Strength on Biomechanics in Adolescent Baseball Pitchers. Journal of Athletic Training. 2024.

Shitara H, Tajika T, Ichinose T, et al. The Association between Excessive Glenohumeral External Rotation and Risk for Shoulder and Elbow Injury in Left-Handed High School Baseball Pitchers: A Prospective Cohort Study. The Orthopaedic Journal of Sports Medicine. 2025.