2021 — Comparing power hitting kinematics between skilled male and female cricket batters

Research theme

Clearing the boundary is a major contributor to winning short-format cricket, but the biomechanics of power hitting had only ever been studied in male batters. Peploe et al. (2019) found that three variables explained 78% of the variance in maximum bat speed in 20 male batters: pelvis-thorax separation in the transverse plane (X-factor) at the start of the downswing, lead elbow extension during the downswing, and wrist uncocking during the downswing. This study asks whether skilled female batters show lesser magnitudes of those same three variables.

15 male and 15 female batters, university to international standard (males: 5 university, 7 professional county, 3 international; females: 5 university, 10 international; 10 of the males were in the Peploe et al. dataset). Each hit a series of shots (males 14 ± 3, females 18 ± 4) against a BOLA bowling machine, aiming straight back over the machine for maximum carry, in a match-representative manner. 18-camera Vicon at 250 Hz, 46 body markers + 5 bat markers + 5 reflective tape patches on the ball. Carry distance came from post-impact ball speed and launch angle fed into a validated ball-flight model with air resistance. The single best trial (furthest carry) per batter was analysed for 26 kinematic parameters.

Crucially, machine speeds were set differently by sex on an international coach’s advice as representative of each group’s normal training: males 32.4 m/s release / 25.2 ± 1.2 m/s incoming after bounce; females 25.7 m/s release / 20.1 ± 2.1 m/s incoming. Batters used their own bats. Analysis: general linear models for the effect of sex on each parameter, with height and body mass as covariates. Effect sizes (β) interpreted as trivial <0.2, small 0.2–0.6, moderate 0.6–1.2, large 1.2–2.0, very large ≥2.0.

Design note the authors are careful about: “the effect of gender” here is a container for all the differences between male and female cricket — anthropometry beyond height/mass, strength, bat inertia, boundary size, ball size and mass, incoming ball speed, coaching history and funding. It is not a biological claim.

What they measured

Performance outcomes:

• How fast the tip of the bat is travelling (maximum bat speed of the distal bat endpoint during the downswing)
• How fast the ball leaves the bat (post-impact ball launch speed)
• How far it carries (ball carry distance, m)

Technique (26 parameters; the full Table 1 list):

• Bat angle at the start of the downswing and at impact; total bat angular rotation; bat centre-of-mass height at start of downswing
• How cocked the wrists are at start of downswing and at impact; how much they uncock from their most-cocked point to impact
• Lead (front) elbow angle at start of downswing and impact; how much it extends through the downswing
• Rear (back) elbow angle at start of downswing and impact; how much it extends
• How far the hips and chest are turned at impact (pelvis, thorax transverse angles)
• The wind-up between hips and chest at start of downswing, in the transverse plane (X-factor) and the frontal plane (X’-factor); the maximum of each during the downswing; the extra stretch beyond the start value (X-factor stretch); and how much it unwinds from maximum to impact (X-factor reduction)
• How far the body moves forward into the shot (centre-of-mass anterior-posterior displacement from its minimum to impact)
• Front knee angle at impact, and whether it straightens from stride end to impact
• How wide the stance is at impact (base length, distance between the feet’s centres of mass)

Findings

Performance outcomes — all large, all in favour of males

1. Maximum bat speed: 28.4 ± 2.5 vs 22.6 ± 2.3 m/s (male vs female). β = 1.54 (large), p < 0.001.
2. Ball launch speed: 33.5 ± 2.6 vs 27.3 ± 2.8 m/s. β = 1.31 (large), p = 0.002.
3. Ball carry distance: 80.7 ± 10.0 vs 57.7 ± 8.8 m. β = 1.43 (large), p < 0.001. A 23 m gap — noting that incoming ball speed also differed by design (25.2 vs 20.1 m/s).

The three a priori hypotheses — two supported, one rejected

4. X-factor (transverse-plane pelvis-thorax separation) at the start of the downswing: 17.6 ± 8.3° vs 12.4 ± 10.1°. β = 1.14 (moderate), p = 0.030. Supported.
5. Lead elbow extension during the downswing: +29.7 ± 12.0° vs −3.0 ± 23.5°. β = 1.28 (large), p = 0.008. Supported — and this is a qualitative difference: males extend the lead elbow; females on average flex it.
6. Wrist uncocking from minimum angle to impact: 57.5 ± 14.7 vs 61.9 ± 14.4°. β = −0.14 (trivial), p = 0.819. Hypothesis rejected. Female batters uncocked the wrists slightly more, not less. One of the three pillars of the male bat-speed model simply does not separate the sexes.

Exploratory effects that reached significance

7. Bat angle at impact: 21.0 ± 7.0 vs 16.8 ± 8.2°. β = 1.31 (large), p = 0.022. Males rotate the bat further forward past vertical at contact — the authors read this as a consequence of the lead elbow difference.
8. Rear elbow angle at impact: 126.3 ± 12.5 vs 112.5 ± 10.6°. β = 1.03 (moderate), p = 0.044. Males’ back arm is more extended at contact.
9. X-factor reduction from maximum to impact: 23.3 ± 7.2 vs 19.6 ± 8.3°. β = 1.20 (moderate), p = 0.044. Males “recoil” the wind-up more through the downswing.

Everything else — not significant

10. No differences in the lower body at all. Lead knee angle at impact (141.6 ± 14.1 vs 146.1 ± 14.6°, p = 0.333); lead knee extension stride-end→impact (−4.1 ± 12.0 vs −0.1 ± 9.7°, p = 0.944); base length at impact (0.81 ± 0.10 vs 0.82 ± 0.10 m, p = 0.360); CoM forward displacement (0.37 ± 0.11 vs 0.46 ± 0.23 m, p = 0.113). Note both sexes’ lead knee extension means are negative — the front knee FLEXES into impact in both groups.
11. No difference in setup: bat angle at start of downswing (p = 0.690), bat CoM height (p = 0.568), wrist cocking angle at start of downswing (119.3 ± 11.8 vs 118.7 ± 12.2°, p = 0.968 — essentially identical).
12. Near-misses worth naming: lead elbow angle at start of downswing (121.2 ± 10.8 vs 133.7 ± 27.5°, p = 0.085 — note the female SD is 2.5× the male, i.e. far more varied); rear elbow extension DS→IMP (70.2 ± 13.4 vs 47.1 ± 17.8°, p = 0.086); maximum X-factor DS→IMP (p = 0.093); maximum X’-factor (p = 0.100); frontal-plane X’-factor at start of downswing (22.3 ± 7.0 vs 14.3 ± 8.5°, p = 0.110 — NOT significant).
13. X-factor stretch (the extra wind-up gained after the downswing begins) was identical: 7.2 ± 4.4 vs 7.0 ± 5.3°, p = 0.144. Both sexes gain the same extra stretch; males simply start with more.

The lead elbow finding in detail

14. Eight female batters (range −7° to −34°), and no male batters, flexed the lead elbow during the downswing.
15. The nine greatest lead-elbow-extension values were all male; the nine lowest were all female. Almost complete separation between the groups on one variable.
16. Direct parallel in golf: experienced male golfers extend the lead elbow ~10°, experienced female golfers flex theirs ~24° (Egret et al., 2006) — but this was not replicated in higher-skilled professional female golfers (Zheng et al., 2008). The authors take that as evidence the difference may be a skill/experience artefact rather than a fixed one.
17. Some female batters DID extend, by up to 30° — so the authors conclude “power hitting solutions involving elbow extension are possible for female batters”.

The authors’ explanations (all explicitly speculative)

18. Strength. Greater male muscle cross-sectional area (only partly controlled by the body-mass covariate) would ease segment acceleration. Absolute strength “may therefore contribute to the selection by skilled female batters of a movement solution involving less elbow extension”.
19. Bat inertia scaling. In baseball, increased bat moment of inertia produces exactly this pattern — a loss of velocity transfer from lead elbow to lead wrist, with the lead arm “controlling and stabilising the swing rather than increasing bat velocity”. The authors say this is “the same pattern observed on average in the present study’s female batters, suggesting that their bat moment of inertia may not be particularly well scaled to their absolute strength constraints.” This is the most immediately actionable hypothesis in the paper — it points at equipment, not technique.
20. Boundary size. Women’s international boundaries are smaller. “If a female batter is able to clear the smaller boundary whilst flexing the lead elbow and utilising relatively little pelvis-thorax separation then there may be little stimulus or benefit to exploring alternative techniques.” The checked-drive solution may be rational, trading bat speed for impact-location accuracy and timing margin.
21. Coaching history. “It is also possible that some female batters have not been coached to utilise a specific power hitting technique like that of the male batters.”
22. Anticipation. Greater anticipation of ball trajectory may facilitate greater torso rotation — but the authors check this against their own data and note no clear difference in torso rotation between university and international batters, which undercuts the skill explanation for that variable.

What a coach should look for on video

This is the most video-friendly paper in the whole cluster. Every significant variable is a big joint or segment angle at one of two easily-found frames — the moment the bat changes direction and the moment of contact. Two frames, one camera.

Cue 1 — Lead elbow extension through the downswing (the single highest-yield cue in this cluster)

• The cue: Does the front elbow straighten or fold between the top of the backswing and contact?
• Camera view + frame: Front-on (from the bowler’s end, offset slightly to the off side so the lead arm isn’t occluded). Highest frame rate the device offers. Two frames: (a) the frame the bat stops rising and starts down; (b) the frame of bat-ball contact. Measure the front elbow angle in each.
• What “good” looks like: The elbow opens by ~30° through the downswing (male mean +29.7 ± 12.0°). Any positive change is on the right side of the distribution; the top performers in this dataset were all in the +30° region.
• What the fault looks like: The elbow closes — the front arm folds in and the shot resolves into a controlled, checked drive. Eight of fifteen female batters did this, by up to 34°. A secondary tell: the bat is less rotated past vertical at contact (males 21.0° vs females 16.8°, p = 0.022) — the bat looks “behind” at impact.
• Why it matters: β = 1.28, large, p = 0.008 — the largest technique effect in the study, with near-complete separation between groups. Mechanism: a longer range over which to accelerate the forearm, and a longer bat-arm system at impact. It is also one of the three variables in the 78%-of-bat-speed model from the male-only parent study.
• Coach’s caution: the authors will not commit to this being coachable. It may be limited by strength, or by a bat too heavy for the batter. Before drilling the technique, check the bat. See Cue 4.

Cue 2 — X-factor: the wind-up between hips and chest at the top

• The cue: At the moment the bat starts down, how far is the chest turned relative to the hips, viewed from above (transverse plane)?
• Camera view + frame: High overhead is the correct view — this is a transverse-plane measure and a front-on camera will not give it to you honestly. Behind-the-batter from an elevated position is the practical substitute. Scrub to the frame the bat changes direction.
• What “good” looks like: Around 18° of separation (male mean 17.6 ± 8.3°) — the shoulder line noticeably more closed than the hip line.
• What the fault looks like: Hips and chest turned as a block; ~12° or less (female mean 12.4 ± 10.1°).
• Why it matters: β = 1.14, moderate, p = 0.030; and it is the first and most proximal difference in the chain — everything downstream of it in this study is arguably a consequence. Mechanism: stretch-shortening cycle, giving faster uncoiling.
• Related cue you can also see: males unwound more from maximum separation to impact (23.3° vs 19.6°, p = 0.044) — the recoil is bigger, not just the wind-up. But the extra stretch gained during the downswing was identical in both sexes (7.2° vs 7.0°). So the coaching target is the position at the top, not the stretch after it.

Cue 3 — Rear elbow at impact

• The cue: How extended is the back arm at contact?
• Camera view + frame: Front-on, frame of contact.
• What “good” looks like: ~126° (male mean 126.3 ± 12.5°) vs ~112° (female mean 112.5 ± 10.6°) — a noticeably straighter back arm driving through.
• What the fault looks like: Back elbow still tucked and flexed at contact.
• Why it matters: β = 1.03, moderate, p = 0.044. Note this survived while rear elbow extension through the downswing did not (p = 0.086) — so coach the position at contact, not the amount of movement.

Cue 4 — NOT a video cue: check the bat

• The paper’s baseball-derived argument is that a bat with too high a moment of inertia for the athlete’s strength produces exactly the lead-arm-stabilising, elbow-folding pattern seen in the female group. Before you spend six weeks coaching lead elbow extension, weigh the bat and check its balance point. Then re-film.
• This is a hypothesis the authors advance, not a result they tested. But it is cheap to check and would be a costly thing to get wrong in the other direction.

Cues this paper actively does NOT support — stop coaching these off this evidence:

• “Snap the wrists.” Wrist uncocking: p = 0.819, β = −0.14 (trivial). Female batters uncocked marginally more. Wrist cocking at the top was near-identical (119.3° vs 118.7°, p = 0.968). This was a pre-registered hypothesis that failed.
• Anything in the lower body. Lead knee angle, lead knee extension, base width, and forward CoM movement all non-significant. There is no evidence here that stance width or front-leg bracing separates big hitters from the rest — and the front knee flexes into impact in both sexes, so “brace the front leg” is not what these batters actually do.
• Setup position. Bat angle at the top, bat height at the top, wrist cock at the top — all non-significant. The difference emerges during the downswing, not before it.

Does the same model predict distance in men and women?

Be careful here — this is the question the study is most often misread on. This paper compared groups; it did not run a within-sex prediction model. So:

• The 78%-of-bat-speed model (X-factor + lead elbow extension + wrist uncocking) was established in male batters only (Peploe et al., 2019).
• This study shows two of those three variables differ between the sexes in the predicted direction, and the third does not.
• Whether X-factor, lead elbow extension and wrist uncocking predict carry distance within a female cohort is not reported in the source I could access. It has not been established by this paper or any other in this cluster.
• What you can say to a coach: the male-derived model’s two mechanical variables describe a real difference in how female batters solve this task, and one of them (lead elbow) is mechanically well-motivated and demonstrably achievable by some female batters. That is a reasonable basis for an individual intervention. It is not a validated performance model for women’s cricket.

Caveats and limits

• n = 15 per group, wide ability spread (university to international) inside each.
• One best trial per batter. By design this captures maximal individual performance, but says nothing about consistency; the authors flag intra-individual variability as future work.
• Incoming ball speed differed by sex by design (25.2 vs 20.1 m/s after bounce). Some of the 23 m carry gap is simply the ball arriving faster for the men. This is a deliberate ecological-validity choice, but it means the performance comparison is not like-for-like.
• Bowling machine, not a bowler. The authors state plainly that this removes pre-release visual cues and is itself a constraint that shapes the movement produced.
• Own bats — good for familiarity, but it means bat inertial properties are uncontrolled, which is awkward given that bat inertia is one of the authors’ leading explanations for the main finding.
• Height and body mass are covariates, but muscle cross-sectional area is not — body mass only partly proxies strength, as the authors acknowledge.
• 26 parameters, 3 pre-registered hypotheses, 23 exploratory tests, no multiple-comparison correction. The three exploratory “significant” results (p = 0.022, 0.044, 0.044) should be treated as leads, not conclusions.
• Cross-sectional and observational. No causal claim is made or supported. The paper’s own framing: “The purpose of the present study is not to fully explain the causal relationships underlying these differences but to identify the combined effects of organismic, environmental and task constraints for further exploration.”
• 10 of the 15 male batters also appeared in Peploe et al. (2019), so the male data are not fully independent of the model being tested.
• “Gender” here is a proxy for a bundle of anthropometric, strength, equipment, task and coaching-history differences. Reading any of these results as biologically fixed would be a misreading of the paper.

Relationship to other Felton work

• Journal version of 2019 McErlain-Naylor power hitting conference. Same participants, same lab, same task — different statistics and, in three places, different conclusions.
• Builds directly on Peploe, McErlain-Naylor, Harland & King (2019), Human Movement Science, the male-only study that produced the 78%-of-bat-speed model. Felton is not an author on that.
• Cites Felton, Lister, Worthington & King (2019), “Comparison of biomechanical characteristics between male and female elite fast bowlers” (J Sports Sci 37(6), 665–670) as its methodological precedent — and uses it to make a sharp point in the introduction: the parameters that differ between male and female fast bowlers are NOT the same parameters that predict ball speed in male fast bowlers. That paper lives in the fast-bowling folders, but the logic transfers: a sex difference in a variable is not evidence that the variable drives performance. This is the correct frame for reading both power hitting papers.
• Methodologically parallel to the spin work: 18-camera Vicon at the ECB National Cricket Performance Centre, best-trial selection, ECB/ICC funding.

CONTRADICTION (with the 2019 conference version, plane of separation): the conference abstract reports males having greater pelvis-thorax separation in the FRONTAL plane at the start of the downswing (moderate evidence). This journal version finds the transverse-plane X-factor significant (β = 1.14, p = 0.030) and the frontal-plane X’-factor non-significant (β = 0.879, p = 0.110). Directly opposite plane assignments for the same data. Prefer this version — it is peer-reviewed and it controls for height and body mass, which the Bayesian t-tests in the conference version did not.

CONTRADICTION (with the 2019 conference version, rear elbow extension): conference reports males extending the rear elbow more during the downswing with “very strong” evidence; this version finds it non-significant (β = 0.831, p = 0.086).

TENSION (with the 2019 conference version, downswing duration): conference reports females having “strong” evidence of longer downswing durations; downswing duration is absent from this version’s 26-parameter table entirely.

TENSION (with the 2019 conference version, methods and counts): 250 Hz here vs 400 Hz at conference; 46 body + 5 bat markers here vs 51 on participant plus bat; 26 parameters here vs 28; 8 females flexed the lead elbow (−7 to −34°) here vs 7 (−9 to −34°) at conference. Also GLM-with-covariates here vs Bayesian t-tests at conference. The covariate change is the most likely driver of the substantive disagreements: males in this sample were 12 cm taller and 12 kg heavier, and once that is partialled out, several conference-version “differences” fall away.

TENSION (with the spin-bowling papers, on front-leg bracing): neither the batting nor the finger-spin work found any relationship with front knee bracing, and in both the front knee actually flexes into the key instant. Bracing the front leg is a persistent coaching instruction that shows up as unsupported in every study in this cluster that measured it.