2015 — Technique and Ball Speed in Female Fast Bowling

Source note: the full one-page abstract was recovered and read in full. It is an abstract, so it reports the regression outcome but not the individual coefficients, r values, or means. Where a number is absent below, it is absent from the source.

Research theme

Everything coaches were told about “what makes a fast bowler fast” came from male bowlers — principally Worthington et al. (2013), who found four variables explained 74% of ball-speed variance in 20 elite men. This study asked the same question of women for the first time: 18 elite female fast bowlers, 18-camera Vicon motion capture at an indoor facility, 47 retro-reflective markers plus a marker on the ball, six maximal-effort good-length deliveries each. Eleven technique parameters were computed and fed into a stepwise linear regression against ball release speed (entry criterion P < 0.10). This is a measured, correlational study — it identifies which technique variables travel with ball speed, not which ones cause it.

What they measured

• Ball release speed (from a reflective patch on the ball)
• How fast the bowler is moving as they arrive at the crease (run-up speed)
• How bent the front knee is at the instant the ball leaves the hand (knee angle at ball release)
• How far back the bowling arm is relative to the trunk when the front foot lands — i.e. how early or late the arm starts swinging over (shoulder angle at front foot contact / onset of upper arm circumduction)
• Seven further kinematic parameters, not individually named in the abstract
• Between-trial repeatability: within-subject sum of squares was 1–10% (mean 5%) of the between-subject sum of squares, so the best three trials were averaged per bowler

Findings

1. Three technique variables explained 74.1% of the variation in ball release speed among the 18 elite women: run-up speed, knee angle at ball release, and shoulder angle at front foot contact. (No individual coefficients or r values are given in the abstract.)
2. Faster run-up = faster ball. Direction agrees with every male study. This is the one point of full male/female agreement in the paper.
3. The faster women had MORE flexion of the front knee at ball release — i.e. a more bent, more collapsed front leg — not less.
4. The faster women had an EARLIER onset of upper arm circumduction — the bowling arm started coming over sooner, not later.
5. Findings 3 and 4 are the opposite direction to Worthington et al. (2013) in elite men, where a straighter front knee at release and a delayed arm were the faster patterns. The authors’ explanation: a strength and technique differential between the sexes, which “caused the front knee to collapse in the majority of the female fast bowlers used within this study.”
6. Note the logical trap the authors are pointing at: the female regression is describing what the faster women in this sample did, and most of them could not hold a straight front leg. A correlation inside a population where nobody braces well is not evidence that not bracing is good.

What a coach should look for on video

Cue 1 — Run-up speed into the crease

• Camera view + frame: Side-on, wide enough to see the last three or four strides. Scrub from the penultimate stride to back foot contact.
• What “good” looks like: Momentum still going forward and building through the bound; no visible check or brake in the last two strides. This paper supports “faster is associated with faster” in women as it does in men, but gives no target number.
• What the fault looks like: The bowler visibly slows or shortens the last two strides to “get set” before delivering.
• Why it matters: Run-up speed survived into the three-variable model that explained 74.1% of ball-speed variance. It is the single most transferable finding across sexes.

Cue 2 — Front knee at ball release

• Camera view + frame: Side-on, pause on the exact frame the ball leaves the hand.
• What this paper actually found: In these 18 women, the faster bowlers had a MORE bent front knee at release. Do not coach that as a target. Read it as a description of the sample: most of these bowlers’ front legs collapsed, and the ones with the most momentum arriving at the crease collapsed the most.
• What the fault looks like: Front knee that is bent at foot strike and continues to bend right through to release, with the hips sinking and the trunk having to reach forward to compensate.
• Why it matters: This is the most contested cue in the whole female-bowling literature. Lyons et al. (2023) found the opposite direction at front foot contact (straighter knee = faster, once height was controlled for). Treat front-leg collapse as a strength problem to fix off the field, not a technique target to imitate. See the CONTRADICTION note below.

Cue 3 — When the bowling arm starts to come over

• Camera view + frame: Side-on or behind-the-arm, pause on the frame the front foot lands.
• What this paper found: The faster women had the arm already further into its circle at front foot contact (earlier circumduction), the opposite of the male pattern.
• What the fault looks like: There is no clean “fault” this paper licenses. The safe coaching read is: in women, do not force a late, delayed arm on the assumption it works the way it does in men.
• Why it matters: Timing of the arm relative to trunk flexion is where the male and female evidence most obviously diverges, and this cue is the reason to be cautious about male-derived coaching in the female game.

Caveats and limits

• Conference abstract only — no coefficients, no r values, no means, no confidence intervals. Everything above is the whole of the reported result.
• n = 18. Elite female bowlers, indoor practice facility, artificial pitch. Cross-sectional and correlational — no intervention, no simulation, no causal claim.
• Stepwise regression with a lenient entry criterion (P < 0.10) on 11 candidate variables and 18 bowlers is prone to overfitting and to selecting variables by chance. The 74.1% figure should be read as an upper bound.
• The knee and arm findings are explicitly explained by the authors as artefacts of a strength constraint in the sample, not as an optimal technique.

Relationship to other Felton work

• Direct female counterpart to Worthington, King & Ranson (2013) in males (the 74% male model), which it deliberately mirrors in method.
• Superseded and expanded by Felton 2019 — Male vs Female Elite Fast Bowlers Compared, the male-vs-female journal comparison in this cluster, which uses the same 20-bowler female cohort logic and 55 parameters.
• Almost certainly the same body of work presented as Felton & King (2017), “Gender differences in cricket fast bowling,” XXVI ISB Congress.
• CONTRADICTION: Front knee at release. This paper says the faster elite women had a more flexed front knee at ball release; Worthington et al. (2013) in elite men, and Felton’s own simulation work, say a straighter braced front leg converts momentum better. Lyons et al. (2023), also on women, found straighter front knees at front foot contact associated with faster speeds (r = 0.68 with height controlled). Different instants are being measured (release vs foot contact), which softens but does not remove the conflict.
• CONTRADICTION: Arm timing. This paper says the faster women had an earlier onset of arm circumduction at front foot contact; Worthington et al. (2013) in men, and Lyons et al. (2023) in women at ball release (r = 0.95), say a more delayed arm goes with more speed.