Curling Delivery: Balance and Leg Endurance for a Stable Slide

Ask a club-level curler why their draws keep sailing through the house instead of dying on the button, and most will point at the ice or the sweeping crew. Ask a coach who has actually watched slow-motion video of the same delivery, and the answer is usually simpler: the sliding leg gives out before the rock does. Somewhere between the third and fifth second of a full-length delivery, the quadriceps and adductors of that leg are holding roughly 40-55% of bodyweight at a knee flexion angle near 145 degrees and a hip flexion angle close to 90 degrees — a position most gym-based leg training never comes close to replicating. When that muscle group starts to fatigue mid-slide, the hips creep upward, the shoulders square off a beat late, and the hand releases the rock on a slightly different line than the one the skip called. On video it looks like a technical fault. In the muscle, it is a straightforward local endurance and balance failure that never got trained on purpose.

The modern flat-back delivery loads one leg far more heavily than most curlers realize. As the thrower pushes out of the hack, the sliding leg (the leg carrying the slider) extends forward and bears the bulk of body weight while the trailing leg extends behind for counterbalance, contributing relatively little vertical support once the slide is established. Bradley (2009), in a widely cited review of curling sport science published in the Journal of Sports Science and Medicine, describes the physiological profile of the sport as dominated not by cardiovascular demand but by sustained local muscular loading of the sliding leg — specifically the quadriceps, hip adductors, and hip flexors — held in a low, flexed position for the duration of each delivery.

A full-length delivery covers roughly 2.5 to 4 meters of slide over 3 to 5 seconds depending on ice speed and the weight of the throw being called. That is not a long time in isolation. The problem is that a single game involves 15 to 20 deliveries per team, spread across 8 to 10 ends and up to two and a half hours, and a competitive bonspiel weekend can mean three or four games in two days. The sliding-leg position is never trained to true fatigue in a single delivery — it accumulates, delivery after delivery, end after end, day after day.

The relationship between contraction intensity and how long a muscle can hold a static position is one of the oldest and most reliable findings in exercise physiology. Rohmert's classic isometric endurance curve, established in the 1960s and confirmed repeatedly since, shows that endurance time drops sharply as the required force climbs above roughly 20% of maximal voluntary contraction (MVC) — largely because intramuscular pressure at higher intensities begins to restrict local blood flow, cutting off the oxygen supply the muscle needs to keep contracting. At the 40-55% MVC range typical of the curling delivery position, sustained continuous holding is generally limited to somewhere in the range of 60-90 seconds before failure in an untrained muscle — far longer than a single delivery requires, which is exactly why the problem does not show up on delivery one or two.

What actually happens across a game is cumulative. Coaches who review video across a full ten-end match commonly notice that knee flexion angle at release has drifted upward by 8-12 degrees by the final ends compared with the first — the sliding leg is quietly losing its ability to hold the trained position, and the body compensates by standing up slightly earlier and higher. That drift is rarely visible to the thrower in the moment. It shows up as a released rock that is a fraction of a second early and a fraction of a degree off line, and by the time it is diagnosed it usually gets blamed on ice reading rather than leg fatigue.

Endurance is only half the equation. The delivery position also requires the sliding leg to control lateral hip drift, knee tracking, and rotational stability across a moving contact point on low-friction ice — a balance demand that gets harder, not easier, as the muscle fatigues. Zech et al. (2010), in a systematic review of balance training published in the Journal of Athletic Training, found that structured balance interventions of 4-6 weeks produced meaningful improvements in postural control measures across the studies reviewed, with reported effect sizes generally falling in the moderate-to-large range. The review's honest limitation for curling purposes is that almost none of the included protocols involved a sustained, low-position, single-leg hold anything like the delivery position — most used standing balance boards or brief dynamic single-leg tasks performed from an upright stance.

That gap matters. General balance training builds the sensory and reflexive machinery for staying upright, but it does not automatically transfer to staying stable at 90 degrees of hip flexion with one leg gliding on ice for four seconds while the quadriceps is already fatiguing. The practical implication is that balance and endurance cannot be trained as separate qualities for this population — they need to be trained together, under fatigue, in the actual joint angles the delivery requires, or the carryover to ice is limited.

The following eight-week progression builds sliding-leg endurance and balance together rather than as isolated qualities, moving from a supported hold through to a fatigue-matched simulation of a full game's delivery count.

The rep counts in Phase 4 are deliberately built around 15-20 repetitions per session because that mirrors the number of deliveries a curler actually throws in a game — the goal is not maximal single-hold duration but the ability to reproduce the same clean position for the fifteenth time as for the first.

Three mistakes show up repeatedly when curlers start this kind of training. The first is holding the isometric position with the hips higher than the delivery actually calls for — 100-110 degrees of hip flexion instead of the 90 degrees a real delivery requires — because the higher position is simply less demanding on the quadriceps. Coach to a mirror or have a partner check hip height against a reference photo of the athlete's own delivery, not a generic target.

The second is training the sliding leg in isolation and never loading the trail leg's contribution to balance, which leaves athletes stable in the gym and unstable on ice the moment the trail leg needs to make a small correction. Phase 3 above exists specifically to close that gap.

The third, and most common among experienced curlers who assume their years on the ice have already solved the problem, is skipping the fatigue-matched volume in Phase 4 and stopping at single long holds. A curler who can hold the position for 60 seconds once but has never practiced holding it cleanly for the sixteenth time in a session has trained the wrong quality — durability across repeated efforts matters more here than peak single-hold time.

Because delivery balance failure is a fatigue effect rather than a static deficit, testing it once at rest tells a coach very little. The useful test is a within-session comparison, and it takes about five minutes with an IMU sensor.

Equipment: PoinT GO sensor (or any 100Hz+ IMU) secured at the sacrum with the provided belt, a stable surface or slide board matching the athlete's usual delivery position.

Procedure: After a general warm-up, the athlete performs 16 delivery-position holds of 5 seconds each, spaced 45 seconds apart, in the PoinT GO app's stability test mode. The sensor logs hold duration and lateral/anteroposterior sway (in mm of displacement) for each rep.

Normal ranges: Trained curlers typically show sway under 8-10mm on early reps. A sway increase of more than 40-50% between reps 1-3 and reps 14-16 in the same session indicates meaningful fatigue-related instability; an increase under 20% suggests the current endurance base is adequate for the athlete's competition volume.

Interpretation: An athlete showing a large late-session sway increase should stay in Phase 3-4 of the protocol above longer before adding on-ice volume. An athlete with a small, stable sway increase across the full 16 reps is a good candidate to progress toward twice-weekly sessions or added competition load. Retesting every 3-4 weeks during an off-season build, or monthly in-season, is sufficient to track whether the training is producing real change rather than guessing from feel alone.