For pitchers searching for better pitching mechanics, video review is useful only when it leads to a clearer throwing program, better drill selection, and better workload decisions. This guide covers why video matters, how to film so your clips are actually useful, what to look for in each phase of the delivery, and where phone video stops being reliable.
Why feel is not enough
One reason video matters is that trunk timing and kinematic sequence are difficult to see accurately at full speed. In a study on kinematic sequence classification, pitches closer to a proximal-to-distal sequence showed less stress at the shoulder and elbow than some out-of-sequence patterns. That is exactly the kind of pattern video review can help flag.
Another example is early trunk rotation. In a study on early trunk rotation in baseball throwing, pitchers with early trunk rotation threw slower and showed higher shoulder joint loading. A systematic review of pitching biomechanics related to pain, injury, and surgery also identified early trunk rotation and higher elbow valgus torque as meaningful risk factors.
Fatigue changes mechanics too, and often in places you would not feel. In a simulated-game study of 28 adolescent pitchers filmed at 240 Hz from two views, velocity dropped, knee flexion at ball release increased, and hip-to-shoulder separation shrank as pitch count rose, while upper-arm kinematics stayed about the same. The lower half and trunk faded first. Without video from the same angles early and late in an outing, that kind of change is easy to miss.
That does not mean every pitcher needs a 3D lab session to improve. But it does mean mechanics work is stronger when it is built on something you can actually observe, compare, and revisit. That is why Driveline's Hacking the Kinetic Chain approach and modern biomechanics workflows both start with assessment before programming.
How to film your delivery
Consistency matters more than equipment. A phone on a tripod, filmed the same way every time, beats a better camera held by a different person from a different spot each session.
- Frame rate: use your phone's slow-motion mode at 120 fps or higher when you can. The arm moves extremely fast. Peak shoulder internal rotation velocity averaged more than 4,500 degrees per second among the pitchers in Driveline's motion capture study. At 30 fps each frame spans about 33 milliseconds, at 120 fps about 8, and at 240 fps about 4, so higher frame rates are what let you see arm timing instead of a blur.
- Lighting: bright, even light. Outdoors in daylight or a well lit indoor space. Dim light lowers shutter speed and smears the arm even at high frame rates.
- Framing: keep the full body visible from the set position through the follow-through, including the rubber and the stride landing area. Do not zoom so tight that the glove arm or front foot leaves the frame.
- Stability: use a tripod or fence mount at roughly hip height. Mark the camera spot so you can repeat it.
- Clip length: one clean throw per clip is easiest to review and compare.
- Pitch and effort: film the same pitch type at a stated effort level so comparisons are fair.
The three camera angles
| Angle | Camera placement | Best for |
|---|---|---|
| Open side | Perpendicular to the target line on the side the pitcher's chest faces at leg lift (third base side for a righty) | Stride length, lead leg block, trunk tilt, arm path, timing of trunk rotation, release |
| Rear (behind) | Directly behind the pitcher, centered on the target line | Direction to the target, stride foot placement, pelvis and shoulder alignment, glove side |
| Closed side | Perpendicular to the target line on the pitcher's back side (first base side for a righty) | Hip load, how long the hips stay closed, back leg drive, hip-shoulder separation at foot strike |
If you can only film one angle, use the open side. If you can film two, add the rear view, because direction and stride problems are hard to judge from the side.
What to look for, phase by phase
- Leg lift and load: balanced posture, weight loaded into the back hip rather than drifting forward early.
- Stride: the body moves down the mound with the hips leading. Check that the stride foot lands on a consistent line toward the target (rear view).
- Foot strike: are the shoulders still closed while the pelvis has started to open? That gap is hip-shoulder separation, which helped explain trunk rotation speed in a 2023 study.
- Lead leg block: after foot strike, does the front knee firm up or keep collapsing? A systematic review linked less knee flexion at ball release to higher velocity.
- Trunk rotation: does the trunk rotate after the pelvis, or open early with it?
- Arm action: is the arm up and in position when the trunk starts to rotate, or is it late and catching up?
- Release and finish: trunk tilt forward over the front leg and a controlled deceleration.
For a deeper breakdown of each phase, see the pitching mechanics guide.
What 2D video can and cannot tell you
Phone video is two dimensional, and pitching is not. Researchers have tested how far 2D measurements can be trusted. In a study of 30 high school pitchers filmed with two high-speed cameras alongside a 3D motion capture system, Oyama and colleagues found that 2D angles were highly reliable between and within raters, but only one variable (trunk lateral flexion at maximum external rotation) had high validity against 3D, and several others were only moderate. Their advice: 2D angles are not necessarily anatomically correct, so limit quantitative video analysis to angles with good validity.
Markerless motion capture, which estimates 3D body position from video using computer vision, is closing the gap. A validation of a single-camera markerless system (a commercial product called pitchAI, unrelated to the Pitch AI app) reported strong agreement with marker-based capture for trunk and pelvis rotation, with larger errors for some arm angles. A 2025 comparison of in-stadium and portable multi-camera markerless systems found the best agreement for stride length and more variability for shoulder rotation. The pattern is consistent: big-segment positions and timing are captured well, while fast shoulder rotation remains the hardest thing to measure without a lab.
The practical takeaway is to use phone video for what it does well: comparing your own clips over time, checking sequencing, direction, stride, and lead leg behavior, and spotting big changes. Treat precise joint angles and arm torques from a single phone clip with skepticism.
How often to review
Film a baseline before a training block, then recheck every few weeks under the same setup. Weekly quick clips during a mechanical change are useful, but avoid reviewing every throw, which tends to create more cues than you can use. Also film late in long outings or bullpens to see whether your lower half fades with fatigue.
Turn video into a plan
Video review is most useful when it changes the plan. If the clip shows timing issues, the training week should reflect that. If the clip suggests a stress pattern, workload and drill selection should change. Video without programming is just content. Video tied to a plan becomes coaching.
What good mechanics feedback should do
- Reduce the number of cues instead of adding more noise.
- Show you what changed, not just what looks wrong.
- Connect mechanics feedback to throwing drills, intensity, and recovery.
- Make reassessment easy so progress is visible over time.
How AI mechanics analysis fits
Pitch AI includes AI mechanics video analysis. The app asks for an open-side clip (required), strongly recommends a view from behind for direction, stride, and block feedback, and accepts an optional closed-side clip. It asks for the camera to be perpendicular to the target line (or centered behind it), the full body visible from set through follow-through, bright even light, and 120 fps or higher for detailed arm timing. When a clip's frame rate is lower, the app warns you and keeps fast arm-action feedback more general rather than overclaiming.
It is not a replacement for a motion capture lab, and it does not measure velocity. What it does is make regular, consistent review practical and tie the findings to your throwing program and drills.
Frequently asked questions
What frame rate should I use to film pitching mechanics?
Use 120 fps or higher when your phone allows it. The arm moves too fast for 30 fps to show detailed timing, so normal speed video is best limited to broad movement patterns like stride and direction.
Which camera angle is most important?
The open side, filmed perpendicular to the target line with the full body in frame. Add a rear view centered on the target line for direction and stride, and a closed-side view if you can.
Is phone video as accurate as a biomechanics lab?
No. Research comparing 2D video to 3D motion capture found high reliability but only moderate or limited validity for many joint angles. Phone video is best for comparing your own clips over time and checking sequencing and direction.
How often should pitchers review video?
Film a baseline, then recheck every few weeks with the same setup. During a deliberate mechanical change, quick weekly clips help. Reviewing every throw usually adds noise.
What camera angles does Pitch AI ask for?
An open-side clip is required, a view from behind is strongly recommended, and a closed-side clip is optional. The app recommends 120 fps or higher, bright even light, and the full delivery in frame.
Mechanics review
Turn your video into a training plan
Pitch AI reviews your delivery with AI mechanics analysis and connects what it finds to a personalized throwing program, drills, and workload management.
Sources and further reading
- Scarborough DM, et al. Kinematic Sequence Classification and the Relationship to Pitching Limb Torques. Medicine and Science in Sports and Exercise, 2021.
- Lin YC, et al. Is early trunk rotation really hazardous for shoulder biomechanics in baseball throwing? Journal of Shoulder and Elbow Surgery, 2022.
- Bullock GS, et al. Baseball pitching biomechanics in relation to pain, injury, and surgery: A systematic review. Journal of Science and Medicine in Sport, 2021.
- Erickson BJ, et al. The Impact of Fatigue on Baseball Pitching Mechanics in Adolescent Male Pitchers. Arthroscopy, 2016.
- Oyama S, et al. Reliability and Validity of Quantitative Video Analysis of Baseball Pitching Motion. Journal of Applied Biomechanics, 2017.
- Dobos TJ, et al. Validation of pitchAI markerless motion capture using marker-based 3D motion capture. Sports Biomechanics, 2022 (online).
- Aguinaldo AL, et al. Assessing the accuracy of in-stadium and portable multi-camera markerless motion capture for baseball pitching kinematics and kinetics. Journal of Sports Sciences, 2025 (online).
- Orishimo KF, et al. Role of Pelvis and Trunk Biomechanics in Generating Ball Velocity in Baseball Pitching. Journal of Strength and Conditioning Research, 2023.
- Kew ME, et al. Kinematic Parameters Predictive of Pitch Velocity in Youth to Professional Baseball Pitchers: A Qualitative Systematic Review. Orthopaedic Journal of Sports Medicine, 2023.
- Marsh JA, et al. (Driveline Baseball). Effects of a six-week weighted-implement throwing program on baseball pitching velocity, kinematics, arm stress, and arm range of motion. PeerJ, 2018.