Cricket batting technique analysis: what to measure in every shot

Why per-shot beats per-session

Batting analysis measures each shot separately rather than scoring a session as a whole. That matters because technique varies delivery to delivery, and an average hides the thing you most want to find: which shots broke down, and what they had in common.

A session average tells you roughly how you batted. It cannot tell you that your balance held for twelve shots and collapsed on the three that came at your pads, which is the actually coachable observation.

Measuring each shot independently makes patterns visible. When the same metric degrades on the same kind of delivery, that is a technical fault rather than a bad day — and it is specific enough to build a drill around.

What gets measured through a shot

Base and balance: how the feet are set and whether weight transfers through the shot or gets stuck.

Trunk rotation: how the upper body turns, and whether it turns in sequence with the hips or fights them.

Head stability: whether the head stays level and moves with the line of the ball through contact rather than falling away from it.

Kinetic sequencing: the order in which segments accelerate. Power comes from a chain firing in the right sequence, and a break in that order shows up as effort producing less than it should; cricket biomechanics from video explains why sequence beats strength.

The phases of a batting shot, and what to measure in each

Biomechanics research breaks a shot into events timed back from impact. In a study of the front-foot drive, Stretch and colleagues (1998) measured the height of the backlift and timed the start of the stride, the start of the downswing and the moment the front foot landed. Splitting a shot this way gives every measurement a moment, so “the head moved” becomes “the head moved during the downswing”. Back-foot shots such as the pull change the footwork: the back foot moves back instead of the front foot forward, as types of shots in cricket explains.

Most of the timing evidence comes from the front-foot drive. In 14 provincial batters filmed at 100 frames per second against medium-fast bowling, the drive’s stride began about 0.58 s before impact, the downswing about 0.36 s before, and the front foot landed only about 0.06 s before the ball was struck (Stretch and colleagues, 1998).

Bat sensors, ball tracking, labs and phone video: what each one measures

“Batting analysis” covers tools that measure different things. Decide first whether your question is about the bat, the ball or the body.

Research often combines methods. A 2025 study of 52 female batters paired high-speed video with bat sensors to compare backlift styles with bat speed and impact speed (Patel and Noorbhai, 2025). Pose estimation from ordinary video is also being used to recognise strokes automatically: one study used an open-source pose-estimation library to classify eight shots, from the pull and cut to the flick and sweep (Siddiqui and colleagues, 2023), and a 2026 study from VIT Chennai combined pose estimation with optical flow to classify shots and compare them with professionals’ shots (Rajarajeswari and colleagues, 2026).

Stance, our cricket video analysis app, analyses phone video with pose estimation: it measures the body from uploaded footage, shot by shot. What a single camera can and cannot see is covered in cricket biomechanics from video, how Stance compares with other cricket apps sets body analysis against ball tracking, and how to choose a cricket analysis app explains what each type of tool measures.

How to film your batting for analysis

1. Pick the angle for the question. Square of the pitch, side-on, is best for the stride, front knee and bat angle; from in front, through the net, is best for whether the head falls across the line. For Stance, film semi-side-on from a squarish angle with the stumps in frame. Our filming guide shows the set-up for each clip.
2. Film at 60 fps or faster. In the drive the front foot lands about 0.06 s before impact (Stretch and colleagues, 1998): under two frames at 30 fps, about four at 60 fps and about seven at 120 fps. Two drive studies analysed footage at 100 and 125 frames per second (Stretch and colleagues, 1998; Stuelcken, Portus and Mason, 2005).
3. Keep the camera still and the whole batter in frame. A fixed surface beats a steady hand. A front arm or back leg hidden behind the body has to be inferred rather than measured.
4. Face real bowling when you can. Young batters timed the drive differently against a bowling machine than against a bowler of the same pace (Pinder, Renshaw and Davids, 2009), so a machine-only session describes your machine technique.
5. Record several of the same shot and compare like with like. Mark each shot, put the camera in the same place every session, and judge drives against drives rather than against pulls.

Reading the output without over-reading it

A single value on a single shot is noise. Three shots showing the same deviation is a signal. Technique analysis is most useful read as a trend across sessions rather than a verdict on one afternoon.

It is also worth distinguishing metrics that describe a choice from metrics that describe a fault. A wider base is not automatically wrong; it is wrong if it stops you transferring weight. Numbers narrow where to look, they do not replace the judgement about what to change.

Head position: the first thing to check

If you check one thing first, check what the head does. Batting is a timing problem before it is a power problem, and the head carries the eyes that supply the timing. A head that moves late, drops, or falls away from the line of the ball degrades the information the whole shot is being built on.

The useful measurement is not only where the head is but how it moves between the bowler releasing and the ball being struck. Stability across that window is what lets a batter track the ball onto the bat. It is measurable frame by frame from ordinary footage, and it is one of the more reliable things a single camera can report, because head movement happens mostly across the camera plane rather than toward the lens.

It can also explain a fault attributed to something else. A batter told they are playing across the line frequently turns out to have a head that has already fallen leg-side before the front foot moves, in which case the bat path is a symptom rather than the cause.

Research supports watching the head, though not a simple keep-it-still rule. Eye and head tracking of two of the world’s best batters found that they turned the head in step with the ball, keeping the ball in a consistent direction relative to the head (Mann, Spratford and Abernethy, 2013). In a lab study of shadow drives, skilled batters kept the head further forward of the centre of their base than less-skilled batters did (Taliep and colleagues, 2007). Both studies are small; the cover drive guide covers them in more detail.

Why one shot tells you almost nothing

A single shot captures one interaction between a specific delivery, a specific decision and whatever state the body happened to be in. Any of those can be unrepresentative, and a good outcome can follow a poor movement just as easily as the reverse.

Repetition is what turns observation into information. Across a set of shots the consistent features separate from the incidental ones: a base that sets the same way every ball is technique, a base that sets differently each ball is noise, and knowing which you are looking at changes what you should work on.

This is why a session is defined as two or more marked shots analysed together rather than a single clip. A one-ball reading is a demonstration; a session is a sample. The same logic governs comparing yourself over time — a session average moves for real reasons, whereas a single-shot figure moves for any reason at all.