How to check cricket bowling speed on your phone

Step by step: check your bowling speed on your phone

1. Prop the phone on a tripod or wedge it against something solid, side-on and square to the pitch, wide enough to see the bowler's hand at release and the batter's stumps.
2. Record in slow motion. iPhone Slo-mo records at 240 fps from the back camera by default, and recent Pixel phones offer up to 240 fps.
3. Open the clip in a player that steps one frame at a time.
4. Find two frames: release, when the ball has just left the fingers, and arrival, when it first reaches the line of the batter's stumps.
5. Time = (arrival frame − release frame) ÷ the frame rate the clip was captured at, not the rate it plays back at.
6. Speed = about 18.5 m ÷ time. Or enter the frames and frame rate in the bowling speed calculator, which also shows the ±1-frame error band.

Which distance to use: about 18.5 m, not 20.12 m

The stumps are 20.12 m apart, but the ball leaves the hand about 1.2–1.8 m in front of the bowler's stumps, so on a full pitch it travels about 18.5 m to the batter's stumps. Dividing by the full 20.12 m overstates every reading, so check which distance your app or calculator divides by; the bowling speed calculator shows by how much and lets you change the distance.

Counting frames, a speed meter app or automatic tracking

Counting frames yourself costs nothing and is as precise as your frame rate: one frame of doubt is worth about ±6% on a 120 km/h ball at 30 fps and under ±1% at 240 fps. Speed meter apps do the same sum from frames you mark. Some offer a tap timer instead, which is the least precise method: if your two taps are out by a combined 0.05 seconds, a 120 km/h ball reads anywhere from about 110 to 132 km/h. Automatic trackers find the ball themselves. Our guide to bowling speed meter apps explains how each type measures.

How automatic tracking uses the stumps

A camera records pixels, not metres. To turn one into the other you need something in the frame whose real size you already know. Cricket is unusually well suited to this because stump dimensions are regulated: the set is a fixed width and height at every ground in the world.

Detect the stumps, and you have a ruler standing in the scene. Detect the ball across consecutive frames, and you have distance travelled. Divide by the time between frames, which the frame rate tells you exactly, and you have speed.

This is why footage without stumps in view is a genuine problem rather than an inconvenience. Without the reference there is nothing to convert pixels against, and any speed figure would be a guess dressed up as a measurement.

Stance, our web app, works this way; it is in closed beta, and you can join the waitlist.

Line, length and movement come from the same trajectory

Once the pitch is calibrated from stump geometry, the ball's path can be reconstructed relative to the pitch rather than relative to the image. That immediately yields where it pitched — the length — and how wide of the stumps it passed — the line.

Movement in the air falls out of the same reconstruction. A ball that deviates from a straight line between release and pitching has moved, and the direction and magnitude of that deviation is what separates swing from drift.

What your footage needs

Keep the stumps in frame. This is the one non-negotiable — it is the reference the entire calculation depends on.

Film from side-on or an angle rather than directly behind. A ball travelling straight away from the lens barely changes position in the image, which is the worst case for measuring it.

Shoot in slow motion at 120 or 240 fps if your phone offers it, and never below 60 fps. At 30fps a fast delivery moves a long way between frames, which coarsens both the speed estimate and the pitching point.

Hold the camera still. A moving camera moves the reference along with the ball, and the calibration has to work harder to separate the two.

How accurate is this, honestly

Accuracy is dominated by your camera angle and frame rate, not by the software — and any app quoting you a single accuracy figure without asking how you filmed is quoting a number it cannot know.

Frame rate sets the floor. Speed is distance over time, and the time between samples is fixed by the frame rate. At 30fps a delivery at 120 km/h travels roughly a metre between frames, so the release point and the pitching point are each located only to about that distance. At 60fps you halve the interval and halve that uncertainty; at 120fps, which recent phones offer, you halve it again. This is the cheapest improvement available to you and it costs nothing but a settings change.

Camera angle sets the ceiling. A camera measures motion across its sensor plane well and motion along its viewing axis poorly. Film from directly behind the bowler and the ball travels almost straight away from the lens: it barely moves across the frame, so the pixel displacement the whole calculation rests on is tiny and noise dominates. Film square to the pitch and the same delivery sweeps across the full width of the frame, where it can be measured properly.

Treat any phone-derived speed as a measurement with an error band, not a certified figure. That is useful for tracking your own progress and not useful for quoting a personal best as fact, and the distinction is worth keeping: a measurement that is consistently a little off is still perfectly good for answering whether you are quicker than last month, which is the question most bowlers actually have.

Common mistakes that ruin the measurement

Filming from behind the stumps at the bowler's end. It is the most natural place to stand and the worst place to measure from, for the reason above.

Zooming in. Digital zoom discards sensor resolution, which is exactly the resource the ball-tracking needs. Move the camera closer instead, or stay wide and accept a smaller ball in frame.

Filming in low light. Short exposures are what freeze a fast-moving ball into a trackable shape; in poor light the camera lengthens the exposure and the ball smears into a streak with no definite position. Indoor nets under weak lighting are the usual culprit.

Letting the stumps drift out of frame. Without the known reference there is nothing to convert pixels into metres against, and the measurement stops being a measurement.

Handheld footage. Camera shake moves the reference and the ball together, which the calibration then has to separate. Any fixed surface beats a steady hand.