How to film your fast-bowling action

• Use slow motion at 120 frames per second or higher when the phone supports it.
• Place the phone around hip height, side-on to the crease and on the bowling-arm side.
• Keep the phone level and fixed on a tripod, fence mount or stable surface.
• Frame enough of the approach to see the run-up and keep both feet visible at landing.
• Leave space above the bowling hand and beyond the front foot.
• Record three to six deliveries at a representative intensity.

A two-dimensional image compresses depth. When the camera sits away from the plane of the delivery, a joint can appear more flexed or extended simply because part of the movement points toward the lens. Side-on footage is therefore the most useful view for sagittal observations such as front-knee position and forward trunk lean.

A front-on or rear view can answer different questions, but it should not be treated as an interchangeable duplicate. If you record more than one angle, capture the same delivery where possible and keep each camera fixed.

Use even daylight or a well-lit indoor net. Avoid placing the bowler in silhouette against a bright sky. Clothing should separate the arms and legs from the background; strong contrast is more useful than a particular colour. Keep other people out of the immediate delivery area so the system and the reviewer can follow one athlete consistently.

• Play the clip once and confirm the whole body remains in frame through follow-through.
• Check that front-foot contact and release are not hidden by another player or net post.
• Keep the original slow-motion file rather than exporting a screen recording.
• Do not add filters, zoom transitions or frame interpolation.

Reviews comparing two-dimensional video assessment with three-dimensional reference systems generally find better repeatability than validity, especially outside the sagittal plane. The camera setup above reduces avoidable error; it does not turn one camera into a motion-capture laboratory. See the systematic review on PubMed.

AI video analysis models estimate body positions from the footage, but those estimates alone do not establish cricket-specific accuracy. A cricket bowling comparison against marker-based motion capture and a separate standard-video validation study show why camera direction, calibration and out-of-plane movement remain part of every interpretation.

1. Abraham, Feros & Fox (2025). Exploring the accuracy of OpenCap for three-dimensional analysis of cricket bowling. 10.1177/17479541251348081
2. Shishov, Elabd, Komisar, Chong & Robinovitch (2021). Accuracy of Kinovea software in estimating body segment movements during falls captured on standard video: Effects of fall direction, camera perspective and video calibration technique. 10.1371/journal.pone.0258923