Research theme
The same ten-bowler dataset as the 2023 Journal of Biomechanics paper, but pitched to a computer-simulation audience as a methodological contribution rather than a cricket one. The title says it: fast bowling is the “case study”; the question is whether running individual-specific optimisations on multiple individuals and looking for trends is a valid general method for finding commonalities in optimal technique for any maximal-effort sporting movement.
The framing is a four-way taxonomy of prior approaches, each with a stated weakness:
- Group-based (1): compare two heterogeneous groups
- Group-based (2): correlate technique with outcome in a homogeneous group
- Individual-based (1): analyse an individual’s natural variation experimentally
- Individual-based (2): optimise a forward-dynamics model of one individual
Group approaches yield only probabilistic “on average” statements (Glazier et al., 2019). Individual approaches cannot be translated to anyone else, because it is unknown whether the identified optimum is suitable or even attainable for another person. The proposed fix — do the individual approach many times and look for trends — has historically been blocked experimentally (individuals rarely display enough natural variation to analyse statistically) and computationally (simulation models are expensive, so almost all applications have been single-subject).
Method, plainly: the same 16-segment planar torque-driven forward-dynamics model used in Felton et al. (2020), customised and evaluated for ten elite male fast bowlers, with both the landing position at front foot contact and the subsequent movement pattern optimised to maximise ball release speed — 112 parameters per bowler, on Loughborough’s high-performance computing system. Paired t-tests on six discrete kinetic parameters (SPSS) and nine kinematic angle time histories (SPM1D).
What they measured
- How fast the ball leaves the hand (ball release speed, m/s).
- How hard the front foot hits the ground, forwards and downwards (peak horizontal braking and vertical force, BW).
- How fast that force arrives (horizontal braking and vertical loading rates, BW/s).
- Total forwards and downwards force over the phase (horizontal braking and vertical impulse, BW·s).
- Nine joint angle time histories through the phase, including front knee, front shoulder, bowling shoulder and bowling wrist.
Findings
Optimised techniques were 13.5% faster — current 35.9 ± 1.5 m/s vs optimised 40.7 ± 1.6 m/s.
Four common kinematic differences across all ten bowlers. In the landing position: more extended front knees, and more flexion of the front and bowling arm shoulders. In the subsequent movement: delays to the onset of front and bowling arm shoulder extension, and wrist flexion.
Kinetic results (Table 1), current vs optimised:
- Peak horizontal braking force: 4.1 ± 0.7 → 3.9 ± 0.4 BW, p = 0.39
- Peak vertical force: 6.0 ± 1.3 → 5.7 ± 0.8 BW, p = 0.09
- Horizontal braking loading rate: 154 ± 39 → 128 ± 45 BW/s, p = 0.01
- Vertical loading rate: 249 ± 118 → 172 ± 70 BW/s, p = 0.02
- Horizontal braking impulse: 0.2 ± 0.1 → 0.2 ± 0.1 BW·s, p = 0.88
- Vertical impulse: 0.3 ± 0.1 → 0.3 ± 0.1 BW·s, p = 0.38
The methodological conclusion: exploring the findings of multiple individual-based optimisations for common trends “can identify underlying commonalities of optimal technique in a maximal effort movement”, and could be used “to provide further evidence to support group-based research and increase confidence in the applicability of the findings within coach development, talent identification, and coaching practice”.
What a coach should look for on video
This abstract supports four cues — a subset of the five supported by the fuller Journal of Biomechanics paper. Because the journal version reports the same data with more detail and with the trunk flexion finding included, use Felton 2023 — optimal initial position and technique for the front foot contact phase as the primary coaching source and treat this file as the record of the methodological argument.
Cue 1 — Front knee more extended at landing
- The cue: Front knee angle in the frame the foot touches down.
- Camera view + frame: Side-on, at front foot contact.
- What “good” looks like: More extended than the bowler’s habit.
- What the fault looks like: Landing with a soft, already-flexing knee.
- Why it matters: Common to all ten bowlers’ optimisations.
Cue 2 — Both shoulders more flexed at landing
- The cue: Front arm higher, bowling arm further back, in the landing frame.
- Camera view + frame: Side-on and front-on, at front foot contact.
- What “good” looks like: Both arms visibly more “loaded” than the bowler’s habit.
- What the fault looks like: Arms already unwinding at landing.
- Why it matters: Common to all ten.
Cue 3 — Both arms start later
- The cue: The frame at which the front arm begins to pull down and the frame at which the bowling arm begins to come over.
- Camera view + frame: Side-on, from front foot contact.
- What “good” looks like: Both onsets later than the bowler’s habit.
- What the fault looks like: Either arm leading the action off the landing.
- Why it matters: Common to all ten. The paper notes this supports both group-based (Worthington et al., 2013) and individual-based (Salter et al., 2007; Felton et al., 2020) prior findings.
Cue 4 — Wrist flexion delayed
- The cue: When the bowling wrist starts to flex.
- Camera view + frame: Behind-the-arm or side-on, high frame rate, final frames before release.
- What “good” looks like: Wrist held back longer before snapping through.
- What the fault looks like: Early wrist flexion.
- Why it matters: Common to all ten. Weakest of the four — the wrist is the joint most constrained in this model.
A load cue with a correction attached: the optimised (faster) techniques had significantly lower loading rates (horizontal p = 0.01, vertical p = 0.02). That is a real and useful message: faster technique need not be higher-load technique. But do not repeat this abstract’s claim about peak forces — see the caveat below.
Caveats and limits
- n = 10 male elite fast bowlers (demographics not stated in this abstract; the companion journal paper gives age 20.7 ± 2.4 y, height 1.91 ± 0.08 m, mass 86.9 ± 8.5 kg).
- Two-page extended abstract. No model evaluation scores, no participant demographics, no confidence intervals, no effect sizes, no discussion of limitations. Everything here is reported more fully in the Journal of Biomechanics paper.
- Simulation, not intervention.
- Planar (2D) model, front foot contact phase only, initial centre-of-mass velocity not optimised.
- Reporting overclaim. The conclusion section states the kinetic differences “supports group-based evidence that the optimal technique lowers vertical peak forces and horizontal braking and vertical loading rates” — but the paper’s own Table 1 shows peak vertical force at p = 0.09 and peak horizontal force at p = 0.39, neither significant. Only the loading rates reached significance. The same overclaim appears in the Journal of Biomechanics abstract.
- Reference [X] is a broken placeholder in the Figure 1 caption (“simulation model of cricket fast bowling [X]”) — an unpolished abstract.
Relationship to other Felton work
- Same dataset and same optimisations as Felton 2023 — optimal initial position and technique for the front foot contact phase (J Biomechanics 158, 111765), presented to a simulation-methods audience. Use the journal paper for the cricket content.
- Uses the model from Felton 2020 — optimising the front foot contact phase.
- The 40.7 m/s optimised baseline is the reference condition for Felton 2024 — effect of increasing isometric strength on technique and Felton 2025 — effect of increased strength on ball release speed.
CONTRADICTION: (the individual-specific message is partially reversed) this abstract, like its journal companion, concludes that individual-specific optimisations across multiple bowlers do reveal common optimal characteristics — a substantial softening of the position taken in Felton 2017 — optimising individual performance, where the argument was that group-derived findings cannot tell an individual what to change and the deliverable was a bespoke prescription for one named bowler. What survives of the individual-specific claim is that attainability is individual, not that the target is.
TENSION: (numbers differ slightly from the journal version) this abstract reports peak horizontal braking force as 4.1 ± 0.7 → 3.9 ± 0.4 BW and horizontal braking impulse as 0.2 ± 0.1 → 0.2 ± 0.1 BW·s; the journal paper reports 4.13 ± 0.7 → 3.94 ± 0.4 BW and 0.15 ± 0.05 → 0.15 ± 0.05 BW·s. The force figures are just rounding, but the impulse figures (0.2 ± 0.1 vs 0.15 ± 0.05) are meaningfully different presentations of the same quantity. Use the journal values.
TENSION: (peak forces overclaimed) as above — the stated conclusion that optimal technique “lowers vertical peak forces” is contradicted by the abstract’s own p = 0.09.