How to increase bowling speed: what the research shows

How to bowl faster: 10 evidence-based steps

1. Measure a baseline before changing anything. Bowl six full-effort deliveries off your full run-up, record both the average and the fastest, and repeat exactly the same set-up every time you test. The bowling speed calculator turns a slow-motion phone clip into km/h.
2. Build a faster run-up you can control. Run-up speed is the most consistent technique predictor of ball speed in the research. Accelerate smoothly into the crease rather than sprinting flat out: researchers expect a point beyond which a bowler can no longer hold the action together.
3. Carry the momentum through the jump and back-foot contact. Faster bowlers spend less time on the back foot, and elite men keep more of their run-up momentum between back-foot and front-foot contact than elite women, who in one study relied more on rotating the pelvis and trunk.
4. Plant the front leg and brake hard. Faster bowlers produce a larger braking impulse at front-foot contact, helped by planting the front leg further ahead of the body, and release the ball over a straighter front knee.
5. Let the chest drive forward over the front leg. The fastest bowlers in England’s elite group flexed the upper trunk further between front-foot contact and release.
6. Keep the bowling arm back longer. A bowling arm that is still further back when the front foot lands — delayed arm circumduction — was part of the best model of speed in the same group.
7. Line the hips and shoulders up at back-foot contact. Twisting the shoulders back against the hips has shown no measurable speed benefit and is linked to lower-back injury: see bowling action types.
8. Train sprint speed and lower-body strength alongside bowling, not instead of it. Sprint times and lower-body strength correlate with ball speed, but an eight-week strength block without bowling improved strength and left speed unchanged.
9. Treat heavy and light balls as an optional extra, not the core. Controlled trials found trivial or inconsistent gains, and one found worse accuracy.
10. Add workload gradually and stop if your back hurts. Junior bowlers should keep to the spell and over limits for their age, set out in average bowling speed by age, and lower-back pain while bowling means seeing a sports doctor or physiotherapist before bowling again.

What the research links to ball release speed

The broadest summary is a meta-analysis of 18 correlational studies (Ramachandran and colleagues, Sports Biomechanics). Run-up speed correlated strongly with ball release speed (r = 0.50) and the front-knee angle at release moderately (r = 0.40). The front-knee angle at front-foot contact (r = 0.26) and the centre of mass’s speed at release (r = 0.16) were weaker, and peak vertical ground reaction force barely related at all (r = 0.13). Correlations show what faster bowlers do, not what will make a given bowler faster, but they are where the evidence starts.

The most useful single study tested 20 bowlers in the England and Wales Cricket Board’s elite fast-bowling group, who released the ball at 118–143 km/h. Four technique variables explained 74% of the differences in speed between them: a quicker run-up, a straighter front knee at release, more upper-trunk flexion between front-foot contact and release, and a larger shoulder angle at front-foot contact, meaning the bowling arm was held back longer (Worthington, King and Ranson, 2013). Their run-ups ranged from 4.8 to 6.8 m/s.

The front-foot contact is where run-up speed is turned into ball speed. In the same elite group a larger braking impulse at the front foot went with faster balls (r = 0.57), and planting the front leg at a larger angle from the vertical went with a larger impulse (r = 0.71) (King, Worthington and Ranson, 2016). In 20 Australian professional, academy and premier-grade bowlers, run-up velocity (r = 0.62) and a shorter back-foot contact (r = −0.61) went with faster balls (Kiely and colleagues, 2021). A 2024 systematic review of 15 studies found agreement on four things: a faster run-up, a sequential transfer of momentum from the body to the arm, quick deceleration with an extended front knee from front-foot contact to release, and a delayed bowling arm (Bhandurge and colleagues, 2024).

None of this is universal. In Kiely’s group the front-knee angle at release had a trivial correlation with speed (r = −0.06), and in 18 junior bowlers run-up speed at back-foot contact was the strongest predictor while front-knee angle at contact had only a small, non-significant effect (Ferdinands and colleagues, 2026). That study also found the front knee is controlled mainly by braking (eccentric) muscle action, and concluded that chasing a braced leg through isolated knee-extensor strengthening may be ineffective.

What the evidence does not support

A longer delivery stride. Of seven studies in the 2024 systematic review that tested stride length, six found no association with ball speed, and one study in the review linked longer strides with poorer accuracy.

A bigger twist between hips and shoulders. The evidence on hip–shoulder separation is inconsistent. In Australian Institute of Sport data, faster bowlers reached their maximum separation later in the stride, a weak relationship (r = 0.34); another study in the 2024 review found no correlation, and in Kiely’s group separation at front-foot contact had a trivial one (r = −0.09). What looks like separation is often shoulder counter-rotation, which had virtually zero correlation with ball speed (r = 0.009) in the same Institute data and is the main technique factor linked to lumbar stress fractures (Portus, 2001).

Side-on versus front-on. Neither has been shown to be faster; how the shoulders are aligned at back-foot contact has not been linked to ball speed. The action types and their injury record are covered in side-on, front-on and mixed actions.

Hitting the crease harder. Peak vertical force barely correlated with speed in the meta-analysis, and in England’s elite group the faster bowlers tended to have lower peak vertical forces and lower loading rates: speed came with a larger horizontal braking impulse, not a bigger vertical slam (King and colleagues, 2016). An earlier Australian analysis did find higher forces in faster bowlers (Portus and colleagues, 2004), so the safe reading is that speed does not require a harder landing.

Strength, sprint speed and body size

Physical characteristics correlated with release speed more strongly than technique did in the meta-analysis (r = 0.65 against 0.42). Total arm length (r = 0.65) and shoulder strength (r = 0.58) stood out. In South African high-performance bowlers, shoulder extension strength correlated with speed (r = 0.39) (Wormgoor and colleagues, 2010), although in schoolboys no isokinetic strength measure did (Loram and colleagues, 2005).

Size matters most when comparing juniors with seniors. Australian first-class bowlers bowled a best ball of 126.7 km/h against 99.6 km/h for junior representative bowlers and carried about 12 kg more estimated muscle mass. Among the juniors, static jump, bench throw, body mass, percentage muscle mass and height together predicted speed closely (multiple r = 0.86); among seniors, static jump and arm length did (r = 0.74). The authors concluded that lower-body strength contributes more at senior level (Pyne and colleagues, 2006).

Sprinting and lower-body strength show up in high-level bowlers too. In Kiely’s group the 10–30 m sprint split (r = −0.55), the 30 m sprint (r = −0.48) and the isometric mid-thigh pull (r = 0.47) correlated with ball speed; vertical jump and bench-pull strength did not. Bowling-action duration and the 10–30 m split together explained 54% of the variation in speed.

Training methods that have been tested

Heavy and light balls. In the original trial, 16 senior club bowlers trained three times a week for ten weeks, either with regulation balls or with a mix of overweight (161–181 g), underweight (131–151 g) and regulation balls. The modified-ball group gained 4.0 km/h and the regulation group 1.3 km/h; with 5 km/h as the smallest worthwhile change, the authors put the chance of a real benefit at about 1% and concluded the programme was not useful for club cricketers (Petersen, Wilson and Hopkins, 2004). An eight-week programme with a wider range of ball weights produced a worthwhile gain in only two of five bowlers (Wickington and Linthorne, 2017), and warming up with balls 10% heavier or lighter made no difference to speed (Feros and colleagues, 2021).

Strength training without bowling. Ten elite-academy pace bowlers who completed an eight-week off-season strength programme got stronger, but ball speed did not change (31.6 to 31.8 m/s). The authors noted the programme was done in the absence of skills training (Callaghan and colleagues, 2024).

Strength and power alongside bowling. Nine club bowlers who added two weekly sessions of core and lower-body strength, plyometrics and weighted-implement work for four weeks gained 5.1 km/h, about 6%, while nine who kept to normal cricket training gained nothing (Maker and Taliep, 2021). In a randomised trial of 20 medium-fast bowlers who all kept up their technique training, eight weeks of velocity-based French contrast training raised ball speed by an average of 2.08 m/s (about 7.5 km/h) and beat traditional resistance and plyometric training (Zhao and Li, 2026). A combined programme using 250 g and 300 g balls and resisted sprints raised peak speed by 1.2 m/s, but accuracy got worse (Feros, Young and O’Brien, 2020).

Read these as encouraging rather than settled. The groups were five to ten bowlers, the programmes four to ten weeks, and no trial followed bowlers for a season. Technique is slower to change than strength: in an England and Wales Cricket Board group coached individually for two years, coaching aimed at the knees produced no measurable change in knee angles, and the 3 mph average gain in speed was put down to physical maturation (Worthington and colleagues, 2009).

Measuring progress honestly

Decide in advance what counts as real improvement. The training studies above treat about 5 km/h (1.5 m/s) as the smallest change a good batter would notice. A different run-up, camera position or instrument also changes the reading, so judge a gain against that benchmark and a consistent set-up, not one evening’s best ball.

Test the same way every time: same run-up length, same instrument and position, six full-effort balls, with average and fastest recorded separately. Without a radar, check bowling speed on your phone and use the bowling speed calculator; film at 120 or 240 fps if your phone allows it.

Track technique alongside speed, because technique is what you are changing. Run-up speed and front-knee angle at release are both visible from a side-on clip, within the single-camera limits explained in cricket biomechanics from video, and bowling action analysis covers how to film them and what else to measure, phase by phase. Stance measures ball speed from phone video, calibrated from the regulated size of the stumps, and analyses run-up mechanics from phone footage too, so a faster ball can be checked against a faster approach or a firmer front leg rather than guessed at.

Pace, juniors and the lower back

In teenagers, bowling faster goes with a higher risk of injury. Cricket Australia’s 2026 guidelines say around one in three adolescent pace bowlers suffers a lumbar bone stress injury, that the lower back is most vulnerable in the two to three years after the main growth spurt, and that bowlers who are taller or faster than their peers are more vulnerable. In 222 Australian pathway bowlers aged 15 to 19, the risk of such an injury rose about 1.1 times for every km/h of extra ball speed, and about three times for every year younger, after allowing for other factors (Sims and colleagues, as reported in Farhart and colleagues’ 2023 systematic review).

Some of what adds pace also adds load. In 40 adolescent bowlers, a more extended front knee, a faster ball and more shoulder counter-rotation were all related to higher loads on the lower back (Crewe and colleagues, 2013), and in the Australian Institute of Sport data the bowlers without trunk injuries released the ball over a more flexed front knee (Portus and colleagues, 2004). Pace does not, though, require higher peak landing forces (King and colleagues, 2016).

Not every practice ball needs to be flat out. When pace bowlers bowled at 70% and 85% of perceived maximum effort, peak horizontal force fell by 7.9–12.3%, impulse by 15.8–21.4% and loading rate by 7.4–12.7%, against a 5.4–8.3% drop in ball speed (McGrath and colleagues, 2022), which makes sub-maximal bowling a sensible way to practise skills. Keep juniors within the ECB and Cricket Australia limits listed in average bowling speed by age.

The riskiest shortcut is forcing extra pace by twisting the shoulders against the hips. That creates a mixed action, the technique most often linked to lumbar stress fractures, with no evidence it adds speed; bowling action types explains how to spot one. Whether the elbow stays within the 15-degree legal limit is a separate question, covered in how to check a bowling action.