Fast bowling action types: side-on, semi-open, front-on and mixed

The four action types, defined

Researchers classify an action from above, at the moment the back foot lands. Draw one line through the two shoulders and another through the two hips, and measure each against the pitch. On the scale used in Australian and English studies, 180° means the line points straight down the pitch, fully side-on, and 270° means it lies parallel to the bowling crease, fully front-on (Worthington and colleagues, 2009 and 2013).

The criteria most often cited come from the Australian Institute of Sport (Portus, 2001). Side-on: shoulders under 210° at back-foot contact. Semi-open: 210° to 240°. Front-on: over 240°. All three also require the hips and shoulders to be within 30° of each other at back-foot contact, and the shoulders not to counter-rotate by 30° or more afterwards. Mixed: a hip–shoulder separation of 30° or more at back-foot contact, or shoulder counter-rotation of 30° or more.

Shoulder counter-rotation is the turn of the shoulders back towards side-on during the delivery stride: the shoulder angle at back-foot contact minus the most side-on angle the shoulders reach later in the stride, usually just before the front foot lands (Portus, 2001; Farhart and colleagues, 2023). Hip–shoulder separation is the shoulder angle minus the hip angle (Worthington, King and Ranson, 2013). Elliott’s review describes the result: the lower body of a front-on action combined with the upper body of a side-on one (Elliott, 2000).

The thresholds are research conventions rather than laws of nature, and they have changed. An early classification counted shoulders under 190° as side-on, and early work used 40° of counter-rotation to define a mixed action (as reported by Stockill and Bartlett, 1992); Elliott’s review cites counter-rotations of 12–40° as predicting injury; a New Zealand study used modified thresholds and a different angle convention (Ferdinands and colleagues, 2010). The back foot gives a visible clue: it typically lands parallel to the crease in a side-on action and pointing down the pitch in a front-on one (Arora and colleagues, 2014), and its angle correlated with the hips (r = 0.73) and shoulders (r = 0.51) at back-foot contact in first-class bowlers (Stockill and Bartlett, 1992).

Side-on, semi-open or front-on: which is faster?

None, on the evidence available. Shoulder alignment at back-foot contact — front-on, mid-way or side-on — has not been linked to ball release speed, and neither has the amount of shoulder counter-rotation (as reviewed by Worthington, King and Ranson, 2013). In the Australian Institute of Sport data, counter-rotation had virtually zero correlation with speed (r = 0.009) (Portus, 2001); in 20 Australian high-level bowlers its correlation was small and not significant (r = 0.24) (Kiely and colleagues, 2021). In 17 international and county bowlers filmed in England, the effect of back foot, hip and shoulder orientation on speed was inconclusive (Stockill and Bartlett, 1992).

The side-on action was once taught almost without exception as the correct technique (Stockill and Bartlett, 1992), but top bowlers do not follow the textbook. Those 17 first-class bowlers averaged a shoulder angle of 228° at back-foot contact, more front-on than the coaching manuals of the day, and of England’s 20 elite fast bowlers tested in 2013, six were front-on and 14 mid-way at back-foot contact (Worthington, King and Ranson, 2013).

The evidence is thin — a handful of studies, mostly small and some old — so the honest conclusion is that no legal action type has been shown to be faster, not that action type is proven irrelevant. What does predict speed is the same in every type: run-up speed, the front leg, trunk flexion and arm timing, covered in how to increase bowling speed.

The mixed action and the lower back

The case against the mixed action comes mainly from Western Australian and Australian Institute of Sport studies. Of 24 selected 16- to 18-year-old fast bowlers, 54% had pars interarticularis defects and 63% disc degeneration, and actions involving counter-rotation were associated with more of both (Hardcastle and colleagues, 1992). In 82 young high-performance bowlers followed for a season, 11% suffered a vertebral stress fracture (Foster and colleagues, 1989); those with fractures counter-rotated their shoulders by more than 40°, against 15.7° in uninjured bowlers, and back-injured bowlers landed with shoulders over 240°, more front-on, at back-foot contact (as summarised by Farhart and colleagues, 2023).

Over 2.7 years, disc degeneration in 19 young bowlers rose from 21% to 58%, and progression was related to using the mixed technique at both testing sessions (Burnett and colleagues, 1996). Among 42 high-performance bowlers, those with lumbar stress fractures averaged 41° of counter-rotation against 19° in bowlers without trunk injuries (Portus, 2001). A 2018 review reported a twelve-fold higher risk of lumbar injury with excessive counter-rotation (Senington and colleagues), and a 1996 review judged the link between the mixed technique and lower-back injury strong (Bartlett and colleagues).

Mixed actions are common even at the top. Of the 42 Institute bowlers, 31 were mixed — and 17 of those only because their shoulders counter-rotated after a well-aligned back-foot contact (Portus, 2001). Ranson and colleagues found a high percentage of 50 professional fast bowlers were mixed, attributable to counter-rotation over 30° (2008); 64.7% of 34 New Zealand bowlers were mixed (Ferdinands and colleagues, 2010); and 19 of 26 elite Australian women fast bowlers used a mixed action (Stuelcken and colleagues, 2010).

What newer studies add, and the limits of the label

Later research is less clear-cut. A 2023 systematic review of 25 studies rated 23 at high risk of bias and found that associations reported for front-on shoulder alignment, counter-rotation and the mixed technique were replicated in only one of six subsequent studies (Farhart and colleagues, 2023). In 48 English county professionals, a quarter of whom suffered an acute lumbar stress injury, action type made no difference to who was injured (Ranson and colleagues, 2008).

Part of the reason is that the label is a proxy. Counter-rotation and hip–shoulder separation were only modestly related to measured spinal motion in 35 elite bowlers, and the authors called counter-rotation a very modest proxy — while noting it remains the only such metric yet linked to injury (Senington and colleagues, 2018). Mixed and non-mixed professionals used similar proportions of their available lower-trunk movement, with the greatest extension, side-bending and rotation used during the front-foot contact phase (Ranson and colleagues, 2008). In elite women, counter-rotation was not related to a history of low back pain, but greater side-bending of the trunk was (Stuelcken and colleagues, 2010).

Juniors show the same pattern. In 60 junior bowlers most differences between action types occurred around back-foot contact, in trunk rotation and the thoraco-lumbar joint, and mixed and front-on bowlers loaded the ground faster at front-foot contact than semi-open bowlers (Schaefer and colleagues, 2020). In 40 adolescents, more counter-rotation, a straighter front knee and a faster ball were each related to higher lower-back loads (Crewe and colleagues, 2013). Counter-rotation is still worth reducing, but a non-mixed label is not a clean bill of health: side-bending at front-foot contact and workload matter as well.

How to identify your action type from video

1. Set up two phones at 60 fps or faster, ideally 120 or 240 fps. Put one side-on, square to the pitch and level with the crease, to time the foot contacts. Put the other on the line of the pitch, either far enough behind the bowler to clear the run-up or behind the batter’s stumps facing the bowler, behind a net, raised as high as you safely can: research classified actions from a side camera and one directly overhead, and a raised view along the pitch is the nearest practical substitute.
2. Wear a top that contrasts with the background and put a strip of tape on the tip of each shoulder and on each side of the waistband, so the shoulder and hip lines can be seen from behind.
3. Film six deliveries at match pace off your full run-up.
4. In the side-on clip, find back-foot contact: the first frame in which the back foot’s movement changes because it has met the ground. Find front-foot contact the same way, then locate both moments in the clip filmed along the pitch.
5. At back-foot contact, look at the back foot. Parallel to the crease points towards side-on; pointing down the pitch, towards front-on.
6. At back-foot contact, compare the shoulder line with the hip line in the clip filmed along the pitch. Shoulders and hips around 30° or more apart — usually shoulders more side-on than the hips — is a mixed set-up.
7. Step forward frame by frame to front-foot contact and find the frame where the shoulders look most side-on. If they have turned back 30° or more from where they were at back-foot contact, the action is mixed by the research definition, however neat it looked when the back foot landed.
8. Classify only when the deliveries agree, and treat any reading close to a threshold as unresolved.

Reading the shoulder line on video, and the limits

Filmed from directly behind or in front, on the line of the pitch, a perfectly side-on shoulder line points at the camera, so one shoulder hides the other; a front-on line is seen at full width. In between, the apparent width is roughly the full width multiplied by the sine of how far the line has turned from side-on: about half at 210°, about 87% at 240° and all of it at 270°. That is our geometry, ignoring perspective and trunk lean; get the full width from a frame of the bowler standing square to the camera at the crease.

Two things follow. Near side-on the apparent width changes quickly, so side-on and semi-open are easy to tell apart; between 240° and 270° it barely changes, so semi-open and front-on are hard to separate in this view. Counter-rotation is easier to see than absolute alignment: shoulders turning from 250° at back-foot contact to 220° before front-foot contact shrink from about 94% to 64% of full width while the hips stay open.

Sideways trunk lean, the front arm crossing the chest, camera height and lens distortion all bend the picture, and no single camera recovers an angle seen from above precisely; laboratories use three-dimensional motion capture. Stance’s run-up analysis works from the same kind of phone footage and the same geometry applies to it, so use the view along the pitch to spot clear cases and large counter-rotations, not to settle a few degrees. Why depth is the weak axis is explained in cricket biomechanics from video, and what else the same two-phone set-up can measure, phase by phase, in bowling action analysis.

Action type is also a separate question from legality. Any of the four types can be legal or illegal: legality depends on how much the elbow straightens, judged against the 15-degree rule, and measuring that needs a camera side-on to the bowling arm instead, as described in how to check if a bowling action is legal.

How coaches remediate a mixed action

The aim is to make the upper and lower body agree at back-foot contact and stay together until the front foot lands — not necessarily to make every bowler side-on. Bowlers who land very front-on rarely hold that alignment through the stride, and those trying to be side-on or semi-open often fail to line the shoulders up with the lower body before back-foot contact; both end in counter-rotation (Portus, 2001). Elliott’s review recommends reducing counter-rotation (Elliott, 2000), and the New Zealand authors suggest the semi-open action as an alternative to front-on (Ferdinands and colleagues, 2010).

Coaching can change it, slowly. In Western Australian development squads with a mean age of about 13, a yearly half-day clinic and six small-group sessions a season over three years reduced counter-rotation, and the incidence and progression of disc degeneration fell with it (Elliott and Khangure, 2002); the review that summarised the study reports counter-rotation falling from 35.4° to 21.3° (Farhart and colleagues, 2023). In England and Wales Cricket Board bowlers coached individually for two years with verbal feedback, video feedback and part drills, the coached group’s shoulder angle at back-foot contact moved from 254° to 233° and counter-rotation from 52° to 32°, while uncoached bowlers did not change (Worthington and colleagues, 2009; Ranson and colleagues, 2009).

Shortcuts are less convincing. A bowling harness reduced hip–shoulder separation in 13-year-olds while they wore it, but left no lasting change once it came off (Wallis, Elliott and Koh, 2002). An eight-week exercise-based injury-prevention programme in community adolescent bowlers reduced counter-rotation by 3.8° relative to controls, whose counter-rotation increased (Forrest and colleagues, 2020).

Remediation need not cost pace on current evidence. The injury-prevention programme had no detectable effect on ball speed, and the England bowlers gained 3 mph on average over the two years, which the authors attributed to maturation (Worthington and colleagues, 2009). Even after successful coaching, side-bending of the lower trunk did not change, so technique work belongs alongside sensible workload limits, such as the junior limits in average bowling speed by age, not in place of them.