Drive Phase: Peak Force Arrives at the Middle of the Stroke

The drive is the part of the stroke where you apply force to the footboard. The standard description runs legs, then trunk, then arms, and the standard correction for a messy drive is to go back and start the stroke with your legs. That advice is correct and almost useless on its own, because it names an order without saying what the order is for.

The context

The usual framing treats the drive as a pull that ends at the finish. Under that framing everything about a good drive is judged by the last third of the stroke: how high the handle comes, whether the arms finish clean, whether the body stayed connected. It also produces a strange difficulty, because people who reach the finish in a mess often have perfectly good legs.

The manufacturer's own technique material sets out the order the same way most coaching does, as legs, then trunk, then arms ([1] Concept2), and practitioner biomechanics writing has coached it that way for years ([11] Kleshnev). That is why the sequence arrives as the first thing you are told and rarely as the thing you are asked to examine.

There is a second framing, older and less quoted, in which the drive is a chain rather than a pull. The rowing-physiology literature describes the stroke as a coordinated sequence through legs, trunk and arms rather than a single effort ([10] Secher), and the biomechanics reviews reach the same conclusion from the mechanical side ([2] Soper, [4] Baudouin). Under this framing, breaking the order does not just look untidy; it changes how much force reaches the blade.

The core idea

Peak handle force does not arrive at the end of the drive. It arrives near the middle of the stroke cycle.

In a cross-sectional study of fourteen adolescent rowers, peak forces were measured during 2000 m ergometer time trials split into start, middle and end phases. Male rowers produced their maximum force at 46 percent of the cycle and female rowers at 43 percent, with the start phase carrying a high force burst in the initial quarter ([5] Ledergerber). The figure is the whole argument: the stroke's biggest contribution lands at roughly the halfway mark, which is to say in the middle of the leg and trunk work, not during the arm finish.

That reframes what the stroke is doing. The drive is not a single pull that ends at the finish. It is a race to get force up quickly, followed by a decision about how much of it to keep. A rower who arrives at the catch fresh and unhurried has time to build that peak. A rower who spends the first part of the stroke setting up will arrive at the middle of it with nothing left to produce, and no amount of arm effort will make up the difference, because the handle force curve is already past its peak by then.

This is also why the sequencing order is not arbitrary. Movement-competency work in rowing describes the trunk as a lever running the length of the drive and as the major link in the kinetic chain between legs and arms ([3] Legge). A lever transmits force; it does not generate much itself. If the trunk is not braced and swinging by the time the legs have finished, there is no rigid structure to transmit that leg force into the arms, and the leg work dissipates into the seat rather than reaching the handle.

The shift across the drive is measurable too. Trunk extensor activity dominates roughly the first sixty percent of the drive, alongside the hip extensors, and trunk flexor activity takes over for the remainder, performing a braking action as the stroke runs into the finish ([3] Legge). So "trunk second" is not a rule about staying still. It is a description of a lever that is loaded on the way and unloaded on the way out.

The how-to

Work through this at a rate low enough that the middle of the stroke is separable from the end of it. Around 16 to 18 strokes a minute is usually enough.

1. Start the drive at the catch, not after it. The stroke's force peak is a mid-stroke event, so time spent compressing before the push is time taken directly out of the window where the peak has to happen. If you find yourself rushing to get moving at the front, you have already spent part of the budget.

2. Push the footboard until your legs are nearly straight, and only then let the trunk open. Not "push, then open" as two separate moves, but a genuine overlap where the legs are still finishing as the swing begins. Practitioner biomechanics writing has long coached the sequence this way ([11] Kleshnev), and it matches the lever description above.

3. Read the force curve, not the finish. A Concept2 monitor plots handle force across the drive rather than watts against stroke position, so what you are looking for is the shape of that curve, not a number on a timeline. A tall curve that peaks early and stays high is a drive that loaded quickly. A curve that starts low and climbs late is one still building force when it should already be past its peak. If the shape changes as the rate rises, the drive is getting slower to load, and no amount of leg strength will show up in the split.

4. Stop training the drive at a rate you cannot hold. Work on the sequence at a low rate until it is unremarkable, then raise it. Field work across stroke rates found that the balance of blade losses and velocity-fluctuation losses inside the stroke shifts with rate, and that the net effect still favours a higher rate ([7] Hofmijster).

5. Do not use efficiency as the reason to slow down. A common piece of folk advice is that rowing fast wastes energy through internal losses. Measured directly, that does not hold. In seventeen well-trained female rowers at 28, 34 and 40 strokes per minute with power to the flywheel held constant, gross efficiency showed no significant differences across rates, and the researchers concluded that internal power losses within the stroke cycle are unrelated to cycle frequency ([8] Hofmijster). Slowing down to be efficient is not, on this evidence, a real intervention.

6. Treat a late collapse as a fatigue signal. In sixteen school rowers, lumbar flexion increased significantly across a trial while erector spinae activity showed the classic fatigue pattern of rising magnitude and falling median frequency, and the authors put forward the fatigue as a partial explanation for the extra flexion ([9] Caldwell). If your drive is crisp for four minutes and then falls apart, fatigue is the first thing to check. That study does not rule out a sequencing fault underneath it, so use this to decide where to look, not to stop looking.

The trade-offs

One limit is worth naming honestly. The strongest evidence for phase-specific strength work comes from a small sample. Fourteen adolescent rowers is a narrow group from which to draw prescriptions for masters rowers, masters competitors and beginners, and the authors are careful to call for intervention studies before training recommendations follow ([5] Ledergerber). What transfers is the shape of the finding rather than the numbers: trunk extension strength and mid-thigh pull rate of force development carry the start, leg press strength carries the middle, and trunk flexion carries the end.

The wider evidence base is more robust and points the same way. A systematic review and meta-analysis of rowing-specific strength and ergometer performance found strong correlations between maximal leg press and back squat strength and ergometer performance, and significant positive associations between explosive strength and rowing power ([6] Jacobs). Eleven studies met inclusion criteria, so this is a small literature, but it is a consistent one, and it puts leg strength at the centre of the drive rather than at its margins.

What to do tomorrow

Pick a piece and row it at a rate you would describe as easy. Watch the monitor rather than your reflection, and note where in the split the peak power lands. Then row the same piece at a rate you would describe as hard, and note it again.

If the peak arrives earlier in the stroke as you speed up, your sequencing is holding and you have room to push. If it arrives later, or the number is lower at the same split, the drive is loading more slowly and the rate is ahead of the technique.

That is a ten-minute diagnostic and it separates the two problems people confuse constantly: a drive that cannot produce force, and a drive that produces force at the wrong time.