2024 — PhD Thesis: Acceleration Transmission During the Fast Bowling Action in Cricket

Research theme

(Inferred from the conference outputs and the thesis keywords — the thesis text is embargoed.)

The thesis consolidates Matthew Lamb’s PhD: quantifying how impact shock from the fast bowling action travels up the body, whether it reaches the lumbar spine in a form capable of causing stress fracture, and what structures attenuate it. The three published components are:

1. Lamb et al. (2022, 40th ISBS) — an experimental study of 11 male fast bowlers with nine IMUs at 1600 Hz from ankle to C7, quantifying peak acceleration and its timing at back foot contact, front foot contact and the follow-through.
2. Lamb et al. (2023, ISB/JSB) — the same dataset with a frequency-domain analysis separating the 2–16 Hz joint-rotation component from the 16–58 Hz elastic wave.
3. Lamb et al. (2023, 19th ISCSB) — a 16-segment 2D whole-body simulation model of front foot contact with compressive springs at the front ankle, knee and hip, matched to one bowler, testing what happens when the lower limb is made 10× and 100× stiffer.

The ECB funding and the “Lumbar / Injury / Stress Fracture” keywords make the applied intent explicit: this work exists to inform how English cricket manages fast bowler loading.

What they measured

(From the published components only — the thesis may contain more, including work never published as a conference paper.)

• Impact shock size at each level of the body (peak resultant acceleration, g) at the ankle/ distal tibia, knee/distal thigh, hip/greater trochanter, L5, L1 and C7.
• How long after ground contact each peak arrived (time to peak).
• How much of the signal is high-frequency tissue “ringing” versus slow joint rotation (PSD integrals in the 16–58 Hz vs 2–16 Hz bands).
• All at back foot contact, front foot contact and the follow-through landing.
• In simulation: the same accelerations under manipulated lower-limb joint stiffness.

Known to be outstanding at the time of the last published component: optimisation of the simulation’s spring stiffness and damping parameters, which the ISCSB paper says would be “presented at the conference” but never appeared in print. That optimisation is the most likely novel content in the embargoed thesis, and I could not access it.

Findings

No findings from the thesis itself are available to me. What follows is the state of the published record that the thesis contains. Each is sourced to its conference paper, and each has its own file in this folder with the full detail.

1. Front foot contact is the largest impact. Peak ankle acceleration ~150 g at front foot contact, versus ~110 g at the follow-through and ~56 g at back foot contact. (Lamb et al. 2022, Table 1.)
2. The lower back sees a small fraction of it. L5 peaks: ~28.8 g at front foot contact, ~18.2 g at follow-through, ~5.8 g at back foot contact. (Lamb et al. 2022.)
3. Attenuation ankle→L5 is roughly 80–91%, and ankle→C7 is 93–95%. (Lamb et al. 2022; Lamb et al. 2023 ISB gives 91 ± 2% / 86 ± 5% / 88 ± 4% tibia→L5 for BFC/FFC/FT.)
4. Attenuation happens at every step of the chain, ankle→knee→hip→L5→L1→C7, each pair significantly different (p < 0.01), with a progressive time delay — the delay is the mechanism. (Lamb et al. 2022.)
5. The body attenuates proportionally harder when hit harder (significant position × phase interaction, F = 10.1, p < 0.001). (Lamb et al. 2023 ISB.)
6. The high-frequency elastic wave — the component most implicated in bone microdamage — is largely gone before L5, leading to the suggestion that it “may be unlikely to contribute to lumbar stress fracture risk.” (Lamb et al. 2023 ISB.)
7. Compliance in the lower limb causes the attenuation. Making the front ankle, knee and hip 100× stiffer in simulation raised shank acceleration from 76.1 to 196.7 g and L5 from 12.3 to 17.6 g; a more plausible 10× raised L5 by only ~5%. (Lamb et al. 2023 ISCSB, n = 1.)
8. Repeated methodological argument across all three: surface accelerations are not internal bone loads (muscle forces are excluded), and existing rigid-jointed cricket simulation models will mis-state spinal loading.

What a coach should look for on video

This thesis supports no coaching cue that I can verify, because I could not read it.

The published components it is built from support no direct video cue either — they are instrumentation and simulation studies that never relate an accelerometer measure to a joint angle, to ball speed, or to an injury outcome. The two practically useful statements available, both from the conference papers rather than the thesis:

• Film front foot contact side-on. It is the largest loading event of the action (~150 g at the ankle, ~29 g at L5) and it is where the front-leg trade-off is decided.
• Film the follow-through landing too. It is nearly as severe at the knee (~88 g vs ~91 g at front foot contact) and is attenuated the least by the knee there (18 ± 20%). It is routinely not filmed at all.

On the wearable/workload question a coach or S&C actually asks — “can I use an accelerometer to monitor a bowler’s load?” — the published record says:

• Yes, as a count of impact exposure. A tibia-mounted IMU sees the largest, cleanest signal and distinguishes the three impacts of the action reliably. Front foot contact and the follow-through are the two that matter; back foot contact is significantly milder.
• No, not as a lumbar load measure. 80–91% of the tibial signal is gone by L5, the attenuation fraction varies between individuals (±5–10% SD), and the group’s own conclusion is that the high-frequency component probably is not what damages the lumbar spine anyway. An ankle “bone load” score is a proxy for what the leg absorbed, not for what the back received.
• Everything above is from six maximal deliveries in a laboratory on 11 healthy young men. There is no fatigue data, no spell-length data, no season-long accumulation, and no injured bowler in any of it. The applied workload-monitoring question — the reason the ECB funded this — may well be addressed in the embargoed thesis, and I could not check.

Caveats and limits

• The primary caveat is access: I could not read this thesis. Embargoed until 2027-10-01; the repository description contains no abstract. Treat this file as a pointer, not a summary.
• Everything I can say rests on n = 11 healthy young men (experimental) and n = 1 (simulation), all male, all lab-based, all six-delivery maximal-effort protocols.
• Surface accelerometry ≠ bone load. Muscle forces excluded throughout.
• The simulation is 2D planar, which cannot represent lateral trunk flexion — the mechanism most implicated in lumbar stress fracture.
• No injury outcomes were measured in any published component. Nothing in this line has been prospectively validated against actual stress fractures.
• The simulation’s spring optimisation — the piece that would make the model quantitatively usable — is unpublished and presumably inside the embargoed thesis.

Relationship to other Felton work

• The parent document for the three Lamb conference papers in this folder: Lamb 2022 — surface measured accelerations, Lamb 2023 — acceleration transmission, Lamb 2023 — lower limb joint compliance and acceleration transmission.
• Supervised by the same Loughborough/NTU group as Manawadu’s 2023 thesis (King, Felton, McErlain-Naylor, Hiley), with Brooke-Wavell and ECB Chief Medical Officer Nick Peirce added — the injury/bone-health side. Sits alongside Alway, Brooke-Wavell, Langley, King & Peirce (2019) on lumbar stress fracture incidence and bowling workload in English county cricket.
• Explicitly positions itself against Felton’s own performance-optimisation line. All three components cite Felton, Yeadon & King (2020) as a whole-body cricket model that “focused solely on performance” and has “not yet explored the relationship between performance and injury”, and argue (via Allen, King & Yeadon 2012) that models with rigid pin joints are unsuitable for estimating internal loading. This thesis is the group’s attempt to build the missing half.

TENSION (the central one in this folder): the attenuation this thesis quantifies is produced by lower-limb compliance. Felton’s performance work (Felton, Yeadon & King 2020; Worthington, King & Ranson 2013) finds that a straighter, braced front leg produces more ball speed. Stiffer leg = faster ball = more shock transmitted upward. Lamb’s ISCSB simulation confirms the direction causally. No published Felton-group work I could access resolves the trade-off or says where the balance should sit — and the thesis that might is embargoed.

TENSION: the group’s own conclusion that the elastic wave “may be unlikely to contribute to lumbar stress fracture risk” undercuts the premise that impact shock is the right thing to measure for lumbar injury risk — the premise that opens all three papers and motivates the ECB funding.

CONTRADICTION (inherited, within the group): attenuation continuing beyond L5 up to C7 contradicts McErlain-Naylor, King & Allen (2021), who found none beyond L5 in drop landings. McErlain-Naylor co-authors both. Proposed resolution: fast bowling delivers a much larger input, so more acceleration remains to attenuate higher up.