Softball Windmill Pitch Velocity: A Field Test Protocol for Hip-Driven Mechanics

A 13U travel-ball coach borrows the baseball team's radar gun, sets it up behind the plate the way he always has for the boys' team, and clocks his best windmill pitcher at 43 mph. That number looks unimpressive next to a 65 mph high school baseball arm, and it doesn't match the video: a full hip turn, a hard leg drive off the rubber, a wrist snap through release that is anything but slow. The gun isn't lying. The setup is wrong for the pitch — a windmill delivery generates and transfers force through a completely different sequence than an overhand throw, and a radar protocol built for baseball simply wasn't built to read it.

This is a field protocol built around the windmill motion: where to stand the radar so it reads the release rather than a rising ball, how far the release point really sits from the plate once a legal leap is factored in, how to track the hip rotation that actually drives the pitch, and what a velocity number means once normalized for circle distance and level. Two biomechanics studies of elite windmill pitchers anchor the sequencing benchmarks here — Barrentine et al. (1998) and Werner et al. (2006) — so the hip-shoulder timing bands are grounded in measured kinematics, not a coach's gut feel.

Two mismatches cause most bad windmill readings, both from copying a baseball testing habit onto a pitch that doesn't move the same way. First, release height: an overhand pitcher releases near shoulder height, 5 to 6 feet up, on a flat trajectory toward a gun aimed at chest level. A windmill pitcher releases near hip height, 2 to 3 feet up, so the ball's early path sits closer to the turf before flattening out. A gun mounted at baseball height and angled down at the plate reads a steeper approach angle on a windmill pitch, introducing cosine error that quietly shaves several mph off the true number.

Second, distance. Fastpitch is thrown from 40 feet at most high school levels and 43 feet in NCAA and international play — nowhere close to baseball's 60 feet 6 inches. Most rulebooks also allow a leap, where the pivot foot leaves the ground before the front foot lands, carrying release several feet closer to the plate than the rubber distance suggests. A pitcher throwing from a 43-foot circle with a full leap may release closer to 37 or 38 feet out. Report a raw velocity number without noting the actual distance and you can't compare two pitchers, let alone compare one against a chart built at a different level.

Barrentine et al. (1998) filmed 24 elite windmill fastpitch pitchers in 3D to map the delivery's phases — wind-up, stride, the circular down-and-around cocking swing, acceleration, release, follow-through — tracking how segment velocities built through the sequence. The pattern was proximal-to-distal, the same principle behind an overhand throw, but reached through a different circular path: the pelvis rotates and peaks first, the upper trunk follows, and the arm whips through last, picking up energy the hips and trunk already generated rather than manufacturing velocity on its own. Barrentine's group also measured peak shoulder distraction force at roughly 80% of body weight, meaningfully lower than typical overhand baseball values, because more of the work is done by hip rotation and trunk sequencing than shoulder layback and late-arm effort.

The practical implication for testing is direct: a velocity test that only watches the arm and the ball measures the last link in the chain and misses where the pitch is actually built. A pitcher losing velocity or plateauing may have a perfectly fine arm and a hip turn that has quietly gotten lazier — a test that never looks past the wrist will never catch it.

Position the gun off to the pitcher's glove-side, 8 to 10 feet outside the pitching lane, at 2 to 3 feet high — level with the release point, not the plate — angled to catch the ball leaving the hand. This mirrors how radar clocks a low-release sidearm baseball pitcher: reading the ball at release avoids the cosine error above. If behind-the-plate is the only option, raise the gun no higher than release height and note that on the record sheet so sessions using different setups aren't compared directly.

Mount one wrist IMU on the throwing-side wrist, oriented with its long axis along the forearm, exactly as in an overhand throwing-velocity protocol. Add a second sensor at the belt or pelvis, centered over the sacrum, to capture pelvis angular velocity and timing through the circular swing — the piece that turns this from a velocity test into a loading test, showing the hip rotation Barrentine and Werner's data point to as the pitch's actual engine, not just its final output.

Run this as a standalone velocity day, not tacked onto a long bullpen where fatigue muddies the numbers.

Protocol Steps

1. Warm-up ramp: 12 to 15 pitches from roughly half circle distance, building through easy, moderate, and near-full effort before moving to the legal pitching distance for the athlete's level.
2. Record the circle setup: rubber-to-plate distance, whether a leap is used, and the pitcher's level, since all three change how the number should be read.
3. 10 pitches at full game intent from the legal pitching position, fastballs only — mixing in a rise ball or change-up defeats the point. Rest 30 to 45 seconds between pitches to match game pacing rather than a rushed bullpen pace.
4. Discard non-qualifying pitches: anything mis-thrown, a slip, or an obvious take-something-off pitch. Keep only pitches thrown with clear strike intent.
5. Record four numbers per session: peak velocity, the average of the top 3 pitches, peak pelvis angular velocity at release, and hip-to-shoulder separation timing from the IMU pair.

The top-3 average matters more than the single peak for tracking a pitcher over time — one clean pitch can be an outlier either way, while three consistently hard pitches reflect actual current output.

A batter's reaction window depends on time to the plate, not the mph number alone, and a windmill pitch travels a much shorter distance than an overhand one. A 60 mph delivery from roughly 37 feet of effective release distance demands a reaction time comparable to a considerably faster overhand pitch from 60.5 feet — which surprises observers who only compare raw radar numbers across sports. The bands below pair typical circle distances and raw velocity ranges at each level with a reaction-time-equivalent scaled to baseball's 60.5-foot distance. Treat that column as a coaching communication tool, not a validated conversion — a time-based estimate, not a peer-reviewed formula.

That top band isn't hypothetical — Team USA's Monica Abbott has been clocked in the high 70s mph, among the hardest-throwing windmill pitchers ever measured, and on the reaction-time scale above, in the same neighborhood as a mid-100s fastball. That's why raw mph alone undersells a good windmill pitcher.

Hip-shoulder separation here means the time gap between the pelvis reaching peak angular velocity and the upper trunk reaching its own peak — the window where the hips have already fired and the trunk is still catching up, stretching trunk musculature and loading the whip through the arm. Werner et al. (2006), studying 24 elite windmill pitchers with 3D motion capture, found ball velocity correlated significantly with pelvis angular velocity at release, upper trunk rotation velocity, and stride length normalized to height — moderate-to-strong relationships (roughly r = 0.4 to 0.6) pointing to the same conclusion as Barrentine's sequencing data: pitchers who rotate the hips harder and get real separation before the trunk catches up tend to throw harder, independent of arm speed alone.

Two limitations matter here. Both studies used small samples of already-elite pitchers (24 each), restricting the range of hip-timing variation captured and likely understating the relationship in a wider population with less efficient movers. And both are correlational single-session snapshots, not longitudinal tracking — better sequencing goes with harder throwing, but that doesn't prove training hip rotation alone will add velocity for every pitcher. The bands below are PoinT GO's field-practical translation of that pattern, not a reproduction of either paper's raw numbers.

A clean session: top-3 average velocity within the pitcher's usual band for her level, separation timing inside her personal 40-70 ms range, and no more than one discarded pitch out of ten. That combination means the hips are doing their job and the arm is simply finishing what the lower body started — the gun number can be trusted as representative.

A red-flag session looks different: velocity holding steady while separation timing has crept under 40 ms versus the last two sessions. That pattern means the pitcher is likely compensating with extra arm effort to hold the same ball speed while her hip contribution quietly drops off — the same masked compensation baseball pitchers show before a problem becomes visible on the gun alone. The most common false version of this flag comes from a rushed warm-up ramp; a pitcher tested cold shows reduced separation and lower velocity unrelated to her actual mechanics.

Run the full protocol three times before the season, roughly two weeks apart, to build a real baseline rather than trusting one number. In-season, retest monthly, and again whenever a pitcher returns from more than a week off — illness, a shoulder scare, an extended break — since separation timing is sensitive enough to catch a delivery that hasn't fully reset even when velocity looks fine early on.

Key References

• Barrentine, S. W., Fleisig, G. S., Whiteside, J. A., Escamilla, R. F., & Andrews, J. R. (1998). Biomechanics of windmill softball pitching with implications about injury mechanisms at the shoulder and elbow. Journal of Orthopaedic & Sports Physical Therapy, 28(6), 405–415.
• Werner, S. L., Jones, D. G., Guido, J. A., & Brunet, M. E. (2006). Kinematics and kinetics of elite windmill softball pitching. American Journal of Sports Medicine, 34(4), 597–603.