Ask someone to describe the indoor stroke and you will get four words back. Ask them to say where each one ends and the next one begins, and the answers get vaguer. That gap is the whole problem with how the stroke is usually taught, because it teaches recognition rather than sequence — and a stroke you can recognise is not a stroke you can reproduce.
The context
The four phases are a real and useful division. Almost all of the coaching literature and the manufacturer's own material organise the stroke this way, and the federation coaching material does the same [10] [11]. Nothing about that is wrong.
The problem is what the framing implies. Presented as four shapes, each phase becomes something you can achieve at a point in the cycle and then relax from. The biomechanics literature describes something different: a sequence in which each part of the stroke exists to set up the next, and where the trunk's job is specifically to be in the right place at the right moment for the legs to take over [1]. Secher's physiology review frames the same thing as one coordinated sequence through legs, trunk and arms rather than separable phases [8] — which is why fixing one phase in isolation so often moves the problem somewhere else.
There is a second thing almost nobody tells you, and it is specific to the machine you are on.
The core idea
A linear ergometer cannot give you the boat's finish.
Cardoso and colleagues put sixteen rowers with international competitive experience on two machines — an angular ergometer built to approximate boat geometry, and a Concept2 — and compared them across the whole intensity range [2]. Oxygen uptake was similar. Rate and effort differed. But the kinematic differences showed up across every intensity domain, and the one that matters most for technique is the thorax angle at the finish: about 19.4° on the angular machine against 27.5° on the Concept2 at heavy intensity, the authors attributing the difference to the angular machine's closer alignment with on-water technique.
That is roughly eight degrees of extra chest opening on the machine most people learn on, and it is structural, not a fault of yours. A rower who "gets the finish right" by matching the boat may end up, on a Concept2, over-opening to compensate.
This is why equipment exists to close the gap: Elliott, Lyttle and Birkett evaluated an ergometer built specifically as a training aid for on-water single sculling [9]. It is a reasonable conclusion that a machine will only ever approximate the water — and that the approximation is doing something to your finish whether you intended it or not.
Where each phase actually ends
Work through this once at a rate slow enough to separate the phases, around 14 to 16.
The catch ends when the body angle stops moving. Not when the shins reach a particular angle — when the angle settles. Soper and Hume describe an ideal technique in which the trunk arrives at its position on time [1], and an angle still moving when the drive starts is an unstable base for the legs to push against.
The legs open the drive; the arms close it. The drive is not over when the legs finish — the trunk swing and the arm draw carry it through to the handle. Cabrera, Ruina and Kleshnev's model reproduces observed rowing forces from stroke kinematics, with leg extension setting the trajectory [6]. The useful discipline is about order within the drive, not about where it stops: if the arms are already pulling while the legs are still going, the legs have been relieved of work they should have done. Baudouin and Hawkins' review of the factors affecting ergometer performance treats this as the central trade in the stroke rather than a detail — which variable you improve always costs you somewhere downstream [4].
The finish ends at the handle-to-knees point, and it is a decision. On a Concept2 this is where you may be tempted to open the chest further than feels necessary [2]. You do not have to. The angular machine's smaller finish angle is a useful reference for a tighter finish.
The recovery is unpowered, which means it has no excuse. It is the only part of the stroke that is unhurried by definition, and it is where rushed technique starts. There is nothing in it that needs to be fast.
Practising the transitions
The phases are easiest to feel if you isolate the joints between them rather than the phases themselves. Four drills, each on the same easy piece:
-
Pause at the catch. Row at 16, hold the front position for two seconds, then drive. What you are looking for is whether the angle is settled when you release it. If it is still moving, the pause makes that visible in a way a continuous stroke never does.
-
Row the recovery at half rate. Drive at 22, recover at 11. This is a clumsy drill and it is worth doing once. It removes the only reason the recovery gets rushed, which is that the drive is coming.
-
Finish small. Deliberately bring the handle in with a smaller chest opening than feels natural, and check whether the drive improves or degrades. On a linear machine this is the drill that matters, because the Concept2's finish angle runs wider than the angular machine's [2].
-
Change one rate at a time. Row the same piece at 20 and at 26 without changing anything else. The phases do not keep their proportions between those two, and seeing that directly is more instructive than being told it [5].
Two setup facts that move the stroke underneath you
The foot stretcher sets the catch. Engstrom and colleagues measured ergometer rowing kinematics across foot-stretcher positions and stroke rates [3]. Foot-stretcher height and angle change how the legs move through the stroke. A catch that has drifted is very often a stretcher that has moved.
Rate redistributes the stroke rather than scaling it. Hofmijster and colleagues found that as rate rises, power not captured at the blades falls — propelling efficiency improves — while the power lost to velocity fluctuations rises, and net efficiency improves only because the first gain exceeds the second loss [5]. That is why a rate ladder feels different rather than simply harder, and why the phases do not keep their proportions as you speed up.
The trade-offs
Two honest limits.
The first is that this article is mostly about the machine, and the machine is the easy part to measure. The on-water biomechanics literature is considerably larger than the ergometer-specific literature — a 2024 scoping review found a substantial body of instrumented on-water work that largely does not exist for indoor rowing [7]. Much of what is confidently said about indoor technique is extrapolated rather than measured on the machine you are reading this on.
The second is that Cardoso's comparison is one study of sixteen rowers, on one angular machine against one linear one [2]. It establishes that the machines are not interchangeable and that the difference is real and sizeable. It does not establish that the Concept2 finish is wrong, and it certainly does not say to stop training on one. The honest reading is that a Concept2 finish is a slightly different shape than a boat finish, and that a rower moving between the two is managing that difference rather than failing to notice it.
A third, smaller limit: the eight-degree finish difference is a group mean with a wide spread. The spread matters. At heavy intensity the angular machine's thorax angle was 19.4° ± 4.5° and the Concept2's 27.5° ± 7.6° [2] — which means individual rowers on the two machines overlap substantially. Treat the direction as solid and your own difference as something to measure rather than assume.
What to do tomorrow
Pick one easy piece and watch the transitions rather than the phases. Four things, once each: where the body angle settles at the catch, whether the legs start before the arms, how far the chest opens at the finish, and whether the recovery is unhurried.
If your chest is opening a lot at the finish on a Concept2, the machine may be part of why [2] — the group means differ by about eight degrees with enough spread that plenty of rowers overlap. Try a slightly smaller finish and see whether the drive feels better, not worse. And if the catch has moved without you changing anything, check the foot stretcher before you change your technique [3].
The phases are not four shapes to hit. They are one exchange, and the machine you are on is already shaping one of them.
Key points
- The four phases are moments in one continuous exchange. The useful question is where each one ends, not what each one looks like.
- On a linear Concept2 the chest is measurably more open at the finish than on an angular ergometer built to match the boat — about 27.5° versus 19.4° at heavy intensity.
- Foot-stretcher position changes ergometer kinematics, so the catch can move without the rower changing anything at all.
- Rate redistributes power through the stroke rather than scaling it: blade losses fall while velocity-fluctuation losses rise.
- A clean stroke on a Concept2 is not automatically the right shape for the boat. Do not chase a finish the machine cannot give you.
- The recovery is the only unpowered part of the stroke and the part that most often gets rushed.
- Effort ratings come out lower on an angular ergometer even at matched oxygen uptake, which is worth knowing before you read a monitor comparison.
Editorial & Coaching Standards
Guides in the MyNextRow Learn library are created by Concept2 athletes and coaches, synthesizing peer-reviewed sports physiology, biomechanics literature, and authoritative rowing guidelines. Every workout protocol and technical cue is tested for safety, repeatability, and PM5 monitor compatibility.
Sources and further reading
- Soper C, Hume PA. Towards an ideal rowing technique… Sports Med 2004;34(12):825-848— Canonical biomechanical account of the stroke sequence; where each phase actually belongs.
- Cardoso R et al. Angular vs straight linear rowing ergometers… Sensors 2024;24(17):5686— Sixteen international rowers on both machine types; the thorax angle at the finish differs by machine.
- Engstrom I et al. Foot-stretcher position and stroke rate… PLoS One 2023;18(5):e0285676— Foot-stretcher position and rate change ergometer kinematics, including at the front of the stroke.
- Baudouin A, Hawkins D. A biomechanical review of factors… Br J Sports Med 2002;36(6):396-402— Review of the biomechanical determinants of ergometer performance.
- Hofmijster MJ et al. Effect of stroke rate on the distrib… J Sports Sci 2007;25(4):403-411— As rate rises, blade losses fall and velocity-fluctuation losses rise; net power to the boat still improves.
- Cabrera D, Ruina A, Kleshnev V. A simple 1+ dimensional model… Hum Mov Sci 2006;25(2):192-220— A model that reproduces observed rowing forces and motions from stroke kinematics.
- Legge N et al. On-water rowing biomechanical assessment… Sports Med Open 2024;10(1):101— Scoping review; the on-water biomechanics base is much larger than the ergometer-specific one.
- Secher NH. Physiological and biomechanical aspects of rowing… Sports Med 1993;15(1):24-42— Rowing physiology and biomechanics review; the coordination the four phases exist to serve.
- Elliott B, Lyttle A, Birkett O. Rowing. Sports Biomech 2002;1(2):123-134— The RowPerfect ergometer as a training aid for on-water single sculling; equipment built to close the erg-to-boat gap.
- Concept2 — Rowing Technique— Manufacturer technique guidance; the four phases as the machine describes them.
- British Rowing — Online Coaching— National federation coaching material; the coaching-side framing of the sequence.