How rowing on machines is related to on-water technique?

Rowing Biomechanics Newsletter No: 300, March 2026

Did you know that… Biomechanical indicators of rowing on common types of ergometers determine corresponding indicators of on-water rowing by about 45% on average? This is the second part of the study on the comparison of rowing on different types of machines with on-water single sculling.

It was found that maximal seat velocity on all rowing machines was significantly higher than in the single scull. The only exception was the FirstDegree water rower at medium and high resistance. This contradicts the common opinion that ergometers require slower leg speed than on-water rowing, and suggests that rowing machines could be used as a tool for developing fast leg drive, especially at low and medium resistance levels (except for the FirstDegree). However, peak seat velocity in the boat occurred earlier (“bouncing” leg drive) than on the machines; therefore, the latter may promote an ineffective movement pattern.

The Rowing Style Factor (RSF, %) was defined as the ratio of seat displacement to handle displacement during the first 20% of the drive length. RSF represents the coordination of the legs and upper body segments during the beginning of the drive, from the catch to the “transition point”. The target zone for RSF was set at 80–100%, and in the boat, the average values were found within this zone (80–82%) at all stroke rates. Rowing on all types of machines featured a significantly lower Rowing Style Factor than on water, with values ranging from 57% to 76%. The RSF values closest to those in the boat (75–76%) were observed on the Concept2 RowErg and PowerFun at low stroke rates and high resistance. The reason for the lower RSF on machines was that handle velocity increased much more rapidly after the catch than in the boat, requiring rowers to use more upper body movement to pull the handle at higher speed. Typically, rowers demonstrate earlier “body opening” and more “shoulder grabbing” while rowing on machines than in a boat. This effect increases with stroke rate, so the average RSF decreases from 69% at 20 strokes per minute (spm) to 64% at 32+ spm.

The Catch Factor (CF, ms) was defined as the time between the moments of direction change at the seat and at the handle at the catch. The target zones for CF were set from –5 to –25 ms for sculling and from –15 to –35 ms for sweep rowing, meaning that the seat should change direction slightly earlier than the handle. The average CF value in the boat was found outside the target zone (+6 ms). It was found that rowing on all machines demonstrated significantly more negative Catch Factor than on water. This could be explained by the necessity to change the direction of the heavy rower’s mass during rowing on machines with a fixed stretcher, which takes more time and requires earlier seat acceleration at the catch. On water, the light boat mass accelerates more than the rower’s mass at the catch, so changing direction at the seat could be completed later. Surprisingly, ergometers with a mobile stretcher (RP3 and C2 Dynamic) demonstrated similarly negative CF values. The most negative CF values were measured on the Concept2 RowErg and PowerFun machines, which fell within the target zone at all stroke rates and resistance levels. CF on machines increased with stroke rate and resistance level, which may be related to an earlier handle pull relative to the direction change at the seat.

An important question is: How well can rowing technique on machines predict on-water technique? To answer this, values of the main biomechanical indicators on each machine were correlated with the corresponding indicators on water for each rower at each stroke rate (n = 30). Rowing power showed a high correlation with in-boat power: the coefficient of determination (R² = 0.85) indicates that power on the ergometer explains 85% of the variation in power on water. The Catch Factor had a much lower determination – only 30%. The highest average coefficient of determination was found for rowing power (69%), maximal seat velocity, and RSF (51%), while the lowest values were found for the finish force gradient (11%) and the ratio of average to maximum force (25%).

Acknowledgements: Thanks to Russian Rowing Federation for organising this study and personally to Alexander Kleshnev for his extensive work with these measurements.

This is a short version of the Newsletter. To access the full text, please subscribe to BioRow membership here: https://biorow.com/membership/

©2026 Dr. Valery Kleshnev