2016 W2x POL
2016 M8+ St.Paul
2016 LM4- DEN
2016 W8+ Cornell

Erg Power vs. On-Water Power

Erg Power vs. On-Water Power

How does a rower’s erg score correlate with their in-boat power?

This is one of the most common questions asked by rowers and coaches. Recently, we have received interesting data, which may help to answer this question. 47 university rowers in six crews were tested in the same eight instrumented with the BioRow telemetry system: 34 women (averaging 1.73m, 67.9kg, where one rower was tested twice and her best performance data was used), and 13 lightweight men (1.79m, 74.0kg,). The standard BioRow test protocol was used consisting of a single 2000m piece with increasing rate steps (RBN 2013/03). Each rower’s season-best 2k erg was obtained from the head coach of the squad.

A very high correlation r=0.814 was found between the erg power and average power production in the boat (Fig.1). This means 66.3% - two thirds of power variation in a boat is explained by the erg scores of the rowers.


As power depends on stroke rate and significantly varies during the step test, the following regression equation was constructed using only the racing stroke rate (the second last sample from the test protocol, at an average stroke rate of 35.4 spm): in-boat power Pb relates to erg power Pe as

Pb = 0.753 Pe + 11.4                                  (1)

To find factors affecting the remaining 33.7% of in-boat power variation, and to be able to analyse data at various stroke rates, the work per stroke WPS was derived, which is independent of stroke rate. The erg power Pe was converted into WPSe=Pe (60/SR), and an assumption of stroke rate SR being 36spm was used. A regression relating WPSe with in-boat WPSb was constructed (WPSb=0.600WPSe+100.5) and deviations ∆WPSb from it for each rower were derived, which indicate the difference between real on-water power production from the prognostic power based on the erg score. Then, ∆WPSb values were correlated with all other variables and obtained factors ranked by magnitude.

The highest correlation r=0.63 was found between ∆WPSb and power to weight ratio. This means that lighter rowers produce higher power in a boat compared to their erg score. On the other hand: heavier rowers have better erg scores but similar on-water power production compared to their lighter crewmates. This is confirmed with a negative correlation r=-0.22 between ∆WPSb and rower weight.

The seat number in the boat had no effect on relative rowing power, but timing relative to other crew members did: a higher power was produced by those who managed to “connect” to the water earlier (r=-0.39 with time after “slip”), to achieve maximal seat velocity earlier (r=-0.33), and increase force earlier up to 70% of maximal force (r=-0.27). This could be related to the effect of “power transfer through the boat” (RBN 2012/04): rowers accelerating their mass earlier after the catch have an advantage in oar power production.

Longer rowing angles helps to produce higher power relative to the rower’s erg score (r=0.35). Longer catch angles had a slightly higher effect (r=0.26) than finish angles (r=0.22).

Some indicators of the force curve affected ∆WPSb: relatively higher on-water power was produced with a shorter force “slip” (r=-0.23) and “wash” (r=-0.21). Other force curve indicators had insignificant effects (ratio average/max force r=0.11), or no effect at all (position of the peak force, gradients).

Relative power production had negative correlation with the blade efficiency (r=-0.33) and positive correlation with wasted blade power (r=0.37), which means rowers producing higher power in a boat have to “pay the price” of more energy wasted in blade slippage.


Fig.2 shows data from two rowers from the same eight at 35spm: Rower 1 (red) was at the bottom of ∆WPSb ranking (-2.8 SD), and Rower 2 (blue) at the top (+2.0 SD). R2 changes direction at the handle (1) and seat (2) much earlier than R1. At the catch, R2 buries the blade earlier (3) and deeper (4), and also has earlier force growth (5) and peak seat velocity (6). As a result, having 12.3% higher erg score (247.1W and 279.4W), Rower 2 produces 32.9% more power (172.0W and 239.7W) in the boat.

Acknowledgements: Thanks to Cambridge University Women’s Boat Club for cooperation with this study.

©2018 Dr. Valery Kleshnev www.biorow.com