2023 — Range of Motion and Key Performance and Injury Technique Characteristics

Research theme

Everything else in this cluster tells a coach what technique looks like in bowlers who get hurt. This paper asks the next question: why do they bowl that way? Specifically — is a bowler’s technique constrained by how flexible they are? If it is, then a physio’s plinth-based flexibility screen becomes an injury-risk tool, and stretching becomes an intervention, without anyone ever needing a camera.

Method: 45 elite male fast bowlers (age 19.8 ± 2.4 years; height 1.87 ± 0.06 m; mass 83.7 ± 8.4 kg — National, A or U19 squad members, or professionals with international potential), tested at the ECB National Cricket Performance Centre. Before bowling and before warm-up, an experienced physiotherapist ran an 18-measure musculoskeletal screen (the same protocol used in Keylock et al. 2022, ICC 0.73–0.98 depending on test). Then 47 markers, six maximal-velocity good-length deliveries (landing 6–8 m in front of the batter’s stumps), 18-camera Vicon at 300 Hz. Fastest trial analysed. Pearson correlations between 18 ROM measures and 13 technique parameters — six previously linked to ball speed (Worthington et al. 2013), seven previously linked to lumbar stress injury (Alway et al. 2021).

This paper does not measure injury. It measures flexibility against technique, and relies on the 2021 Alway paper to say which technique is risky.

What they measured

Off-field flexibility (18 measures, both sides where applicable):

• How far the shoulder rotates in and out (passive shoulder internal and external rotation, in 90° abduction; and their sum, “total arc of rotation”) — bowling arm and front arm separately.
• How high both arms lift behind the back when lying face down (combined shoulder elevation, cm).
• How far the hip rotates in and out (passive hip internal and external rotation; total arc) — rear leg and front leg separately.
• Hamstring length (passive straight leg raise).
• Hamstring and lower-back flexibility together (sit and reach, cm).
• How far the knee travels over the toes with the heel down (weight-bearing ankle dorsiflexion, “knee to wall”, mm) — both legs.

On-field technique (13 parameters):

• Performance-linked: ball release speed; run-up speed; how straight the front leg is at release (front knee angle at BR); how much the trunk folds forward from front foot contact to release (thoracic flexion FFC→BR); how late the bowling arm comes over (bowling shoulder angle at FFC and BR).
• Injury-linked (all seven taken from Alway et al. 2021): rear knee angle at BFC; rear hip angle at BFC; front hip angle at FFC; thoracolumbar side flexion at BFC and BR; thoracolumbar rotation at BFC; pelvic tilt at FFC; lumbopelvic angle at FFC.
• Convention: anatomical position = 180°; below 180° = flexion, contralateral side flexion, contralateral rotation.

Findings

1. Ball release speed correlated with NOTHING. Not one of the 18 ROM measures was related to ball release speed (range 32.0–39.8 m/s, mean 35.7 ± 1.6). Flexibility does not make a bowler fast.

2. Twenty-three significant correlations were found overall (r = 0.300–0.452, p < 0.05) — ten with performance-linked technique, thirteen with injury-linked technique. Shoulder ROM had the most.

3. Normative ROM values for elite male fast bowlers (Table 2) — useful as a screening reference:

The bowling shoulder is 13° tighter into internal rotation and 11° looser into external rotation than the front shoulder — but the total arc is identical (187 vs 188°). The authors call this “potentially a protective shift in ROM.”

4. Which ROM restrictions predict which injury-linked technique faults. This is the table a coach and physio should care about. Direction is stated in plain language; “injured profile” means the direction Alway et al. 2021 found in bowlers who went on to sustain LBSI.

5. ROM and the performance-linked technique parameters:

More bowling-shoulder internal rotation → a more delayed bowling arm at release, which is a known ball-speed enhancer. Bigger shoulder and front-hip arcs → more trunk flexion from FFC to release, another known ball-speed enhancer.

6. The authors’ central argument. Shoulder ROM had by far the most correlations with technique. Their reading: a bowler whose shoulder cannot get into the required position will compensate with the trunk and pelvis — adopting “potentially injurious movement strategies in the bowling action to ensure the bowling arm is orientated to deliver the ball towards the target at ball release.” The front (non-bowling) shoulder is argued to be a good proxy for what the bowling shoulder was like before it adapted — i.e. it reflects the bowler’s underlying constitution.

7. Ankle dorsiflexion is the second usable finding. Bowlers with more ankle dorsiflexion generated their trunk lean higher up (more contralateral thoracolumbar side flexion at BR) — the uninjured pattern from Alway et al. 2021. Dennis et al. (2008) independently found LBSI-free bowlers have greater ankle dorsiflexion. The authors’ proposed chain: restricted dorsiflexion → more knee and hip flexion at back foot contact → the technique pattern that predicts LBSI.

All correlational. These are r values between 0.30 and 0.45 — that is 9% to 20% of shared variance. No causal claim, no intervention, no injury outcome measured in this cohort.

What a coach should look for on video

The point of this paper is that you can screen off the field — but there are two genuine video cues in it too.

The off-field screen (the main deliverable)

Screen 1 — Front (non-bowling) shoulder internal rotation. The single most informative measure in the paper.

• The measure: Bowler lies face up, non-bowling shoulder abducted 90°, elbow flexed 90°, forearm mid-prone. Physio passively rotates to end of internal rotation; angle read on the back of the forearm just above the wrist. ICC 0.73 (the weakest reliability in the protocol — do it carefully, and twice).
• Reference value: 75 ± 8° (range 57–90) in elite male bowlers.
• What restriction predicts: Less front-shoulder IR correlates with three separate injury-profile technique faults — more ipsilateral thoracolumbar side flexion at BFC (r = −0.424), more contralateral thoracolumbar rotation at BFC (r = +0.310), and more lumbopelvic extension at FFC (r = −0.301). That last one is half of the 88%-accurate LBSI prediction model in Alway et al. 2021.
• Intervention implication: This is a physio-addressable restriction. It is the cluster’s clearest non-video lever.

Screen 2 — Bowling shoulder external rotation and total arc.

• Reference values: ER 125 ± 11°; total arc 187 ± 12°.
• What restriction predicts: Less bowling-arm ER and a smaller total arc correlate with a more flexed rear hip at back foot contact (r = +0.378 and +0.320) — the strongest single LBSI predictor in the 2021 paper, where each 1° of extra rear-hip flexion multiplies injury odds by 1/0.88.

Screen 3 — Ankle dorsiflexion, both legs (knee to wall).

• Reference values: rear leg 109 ± 36 mm, front leg 103 ± 32 mm (huge ranges: 30–200 and 10–185 mm).
• What restriction predicts: Less dorsiflexion correlates with less contralateral thoracolumbar side flexion at ball release (r = −0.343 rear, −0.301 front) — i.e. the injured pattern, where the trunk lean is sourced lower down at the lumbopelvic junction instead of higher at the thoracolumbar joint.
• Corroborated independently: Dennis et al. (2008) found LBSI-free bowlers had greater ankle dorsiflexion.

Screen 4 — Front leg hip internal rotation. Reference 32 ± 8°. In this paper more front-hip IR went with technique in the injured direction (more flexed front hip at FFC, r = −0.387). But the Keylock 2022 adolescent study found the injured bowlers had less contralateral hip IR (32.3° vs 39.5°). These point opposite ways. See the contradiction section — treat hip rotation screening as unresolved.

The two video cues

Cue A — How late the bowling arm comes over at release

• The cue: Where the bowling arm is relative to vertical at the moment the ball leaves the hand — i.e. whether circumduction is delayed.
• Camera view + frame: Side-on, scrubbed to ball release.
• What “good” looks like: A more delayed arm. In this cohort bowling shoulder angle at BR was 220.1 ± 16.5° (range 182.2–253.5).
• What the fault looks like: The arm arriving early, coming over before the trunk has folded.
• Why it matters: Ball speed. Delayed circumduction is an established speed enhancer (Felton et al. 2020; Worthington et al. 2013), and this paper shows it is enabled by bowling-shoulder internal rotation ROM (r = 0.351, p = 0.018). If a bowler can’t delay the arm, check their shoulder before you coach the arm — you may be asking for a position their joint cannot reach.

Cue B — How much the trunk folds forward from front foot contact to release

• The cue: The amount the ribcage flexes over the front leg between the front foot landing and the ball leaving the hand.
• Camera view + frame: Side-on, scrub FFC → BR.
• What “good” looks like: Substantial fold. In this cohort thoracic flexion FFC→BR was 30.3 ± 7.5° (range 16.2–48.5°).
• What the fault looks like: The trunk staying upright, with the arm doing all the work.
• Why it matters: Ball speed — the most heavily replicated speed-enhancing characteristic in the literature. This paper shows it correlates with front-hip total arc (r = 0.429), front-hip IR (r = 0.427), front-shoulder total arc (r = 0.417), front-shoulder ER (r = 0.370), bowling-shoulder total arc (r = 0.327) and rear-hip IR (r = 0.317). A bowler who cannot fold over the front leg may be limited by hip and shoulder rotation, not by intent.

What NOT to expect

Do not screen flexibility to predict pace. No ROM measure correlated with ball release speed at all. Sit and reach and combined shoulder elevation correlated with nothing in either table.

Caveats and limits

• Correlations are modest. r = 0.300–0.452. The strongest single relationship (front-hip IR vs front knee angle at BR, r = −0.452) explains 20% of variance; most explain 9–13%. These are signals, not determinants.
• No injury data in this study. The “injury risk” reading is entirely borrowed from Alway et al. 2021. The chain is: ROM ↔ technique (here) + technique ↔ injury (2021) ⇒ ROM ↔ injury (inferred, never tested).
• The authors flag exactly this: “Another potential limitation concerns investigating ROM measures versus previously identified discrete performance and injury technique characteristics rather than investigating these within the current cohort.”
• No multiple-comparison correction. 18 ROM × 13 technique = 234 correlations, at α = 0.05, unadjusted (deliberately, to avoid Type 2 errors). You would expect roughly 12 significant results by chance alone; 23 were found. The authors say the findings “should be considered cautiously.” Treat the r ≈ 0.30, p ≈ 0.04 results as weak and the r > 0.40 results as the ones worth acting on.
• Direction of causation is unknown. Does tight shoulder ROM force a compensatory trunk pattern, or does years of a particular trunk pattern shape the shoulder? The paper cannot say.
• ROM is variable within a person — time of day, prior activity, warm-up state. Bowlers used self-selected warm-ups, and ROM was measured pre-warm-up to control for this. But “some players may achieve a greater increase in ROM after warming up than others,” so the pre-warm-up numbers may not reflect what the joint can do when bowling.
• Sample: 45 elite male bowlers from one nation. The authors note ROM varies between genders and ethnicities.
• Shoulder internal rotation reliability is the weakest in the protocol (ICC 0.73) — and it is the measure carrying the most findings.
• One trial per bowler (fastest with minimal marker loss).
• Discrete instants, absolute angles — same limitation as the rest of the cluster.

Relationship to other Felton work

• This paper exists to serve the 2021 Alway MSSE paper (2021 Alway — Cricket Fast Bowling Technique and Lumbar Bone Stress Injury). All seven of its “injury” technique parameters are lifted directly from that study, and the entire injury interpretation rests on it. It is the practical follow-through: given that these technique characteristics predict LBSI, what physical constraints produce them?
• Uses the musculoskeletal screening protocol from Keylock et al. 2022 (2022 Keylock — Lumbar Bone Stress Injuries and Risk Factors in Adolescent Fast Bowlers), with Steve McCaig co-authoring both, and cites its reliability figures.
• Cites Felton, Yeadon & King (2020) on optimising the front foot contact phase, and Felton, Lister, Worthington & King (2019) on male-vs-female bowlers, for the performance-side interpretation — both in the simulation/technique clusters.
• Cites Bayne et al. (2016), Dennis et al. (2008) and Stuelcken et al. (2008) for the prior ROM-injury literature it is trying to improve on.

CONTRADICTION: Throwing may be actively bad for a fast bowler’s back. Fast bowlers’ shoulders adapt exactly like throwers’ — gaining external rotation and losing internal rotation on the bowling side (Sundaram et al., 2012; Sauers et al., 2014), and this cohort shows it (bowling arm IR 62 ± 10° vs front arm 75 ± 8°). But this paper finds that the ROM profile associated with non-injurious bowling kinematics is greater internal rotation and less external rotation. The authors state the conflict directly: “The ROM requirements linked with non-injurious kinematics conflict with previously observed bilateral shoulder adaptations in fast bowlers… it is possible that throwing may be counterproductive to developing and maintaining a safe fast bowling technique due to conflicting movement patterns and ROM adaptations.” For a coach this is a live question about fielding practice volume in young fast bowlers.

CONTRADICTION (internal to this paper’s own conclusion): The abstract and conclusion both state that “increased internal rotation, less external rotation, and greater total arc of rotation were associated with technique characteristics… [linked to] decreased lumbar stress injury risk.” But the paper’s own Table 4 shows greater bowling-shoulder external rotation correlating with a straighter rear hip at BFC (r = +0.378, p = 0.010) — the single most protective technique characteristic in the 2021 model. The blanket statement holds for the front shoulder and not the bowling shoulder. Read the two shoulders separately.

CONTRADICTION: Hip internal rotation points in opposite directions across the two 2022–2023 papers. Here, greater front-leg hip internal rotation correlates with the injured-profile technique (more flexed front hip at FFC, r = −0.387, p = 0.009). In Keylock et al. 2022, adolescent bowlers who went on to be injured had 7.2° LESS contralateral hip internal rotation than uninjured bowlers (32.3° vs 39.5°, Hedges’ g = 0.987). Add Dennis et al. (2008), who found reduced ipsilateral hip IR reduced injury risk. Felton et al. acknowledge the mess: “studies investigating the relationship between hip rotation ROM and injury occurrence have found conflicting results.” A coach should not act on hip rotation screening for LBSI risk on current evidence.

TENSION: The ROM profile that helps you bowl fast overlaps with the one that goes with injury-profile technique. Greater front-hip internal rotation and total arc, and greater front-shoulder external rotation, all correlate with more trunk flexion FFC→BR — the key ball-speed characteristic (r = 0.370–0.429). The same measures also correlate with a more flexed front hip at FFC and more lumbopelvic extension at FFC — both injured-profile characteristics from Alway et al. 2021. The paper does not resolve this, and it should temper any simple “get more flexible” message.

TENSION: Flexibility does not make you fast. Ball release speed correlated with zero of 18 ROM measures across 45 elite bowlers spanning 32.0–39.8 m/s. Whatever flexibility work is for, it is not for pace.