Blog

Power Meter Accuracy and Consistency are not the same thing.

The key advantage of power-based training is the range of power metrics that allow us to measure our progress against a structured training programme, or informally. Clearly the metrics depend on the quality of the raw data recorded by our power meters. As they say, Garbage in, Garbage out (GIGO)!

With many new brands of power meters coming to market, the debate has returned to ‘accuracy’ vs. ‘consistency’, much as in the pioneering days of the ‘gold standard’ SRM and PowerTap meters over ten years ago.

If our primary goal is to measure ourselves against ourselves, then consistency might seem more important, since it’s only ‘my watts’ that are important to me. But human nature mostly doesn’t work this way and inevitably we all compare ourselves against our rivals or check how we rate against best-in-class benchmarks of power/weight ratio, FTP or other key metrics. When do this, accuracy becomes critical. If your power solution is not both accurate and precise, it is not really ‘complete’ because it’s not reporting ‘true watts’.

“If your power meter measured in ‘Awesomes,’ would it really matter to you? Or is it more important that you know if you were more or less awesome than the month before? Accuracy and precision are not the same thing.” BikeRadar

The developers of SRM and PowerTap recognised this core requirement and devoted the majority of their research effort to building the best possible data collection platform based on a carefully designed array of up to 16 high spec. strain gauges, which, though costly, guaranteed the accuracy required for reporting in narrow power bands, as advanced programmes require, and also to establish an industry standard benchmark of accuracy. It doesn’t matter that SRM chose to mount its strain gauges in the crank ‘spider’ and PowerTap the hub. The fact is that the inherently stable data collection platforms of both have ensured the survival and growth of both brands in an increasingly fickle market where many others have tried and failed.

The way top range power meters work is by converting high quality raw strain gauge data into torque using the universal stress/strain relationship (remember Hooke’s Law and Young’s modulus ?), after applying corrections for temperature and other known ‘environmental’ variances. Then, by applying Newton’s Laws:

Power = Torque X Angular Velocity

Both SRM and PowerTap use essentially the same calculation of Power, but since the sensors on each system are located at a different position on the bicycle (SRM being at the crank, closer to the point of power delivery) the SRM measured power is a little higher than the PowerTap, which measures at the hub. But both report true watts, with any differences attributable to small drive train losses.

As an analogy, consider the testing of a vehicle’s power – the power measured at the flywheel will always be higher than the power recorded at the tyres, the difference being the power loss through the drivetrain. In bicycling terms, the power that is being recorded via PowerTap is as close as you will get to the power being delivered to the wheel.

Why should coaches or self-trained athletes need to worry about absolute accuracy over just consistency?

The obvious answer is that it allows realistic comparisons across different platforms known to be accurate, and comparisons against best-in-class benchmarks.

“Accuracy & Consistency: SRM was the originator of the main stream power meters. Its owners have been satisfied with the accuracy. The same is true for PowerTap. Stages is new and questionable. It only measures the wattage generated from one crank and has been proven to be tricked by hard, asymmetrical pedalling. It also has a large smoothing factor and fewer readings per second as the top players. Accuracy, in my mind is questionable. I’m not one to buy into the marketing hype hocus pocus.” (sic) iBikeBlog

Because of the importance of accuracy, both SRM and PowerTap chose to seek independent verification since the accuracy and consistency of some cheaper (and some more expensive) devices is known to be questionable, as in-depth testing is beginning to show. Single crank devices, for example, ‘measure’ the ‘wattage’ generated from one crank and assume that total power is simply double this figure. This is not necessarily true, especially for riders with biomechanical imbalances, poor set-up positions or inconsistent pedalling technique. Such devices also rely on complex software to smooth and manipulate the raw data, which may compound the accuracy problem. As we said earlier – GIGO !

In practice there are dangers associated with changing from one power solution to another that is neither accurate nor consistent. Coaches working on periodization training programs which compare new power data to that of the previous periods may find higher readings and thus assume the athlete is peaking early. This may lead to misleading advice to rest, resulting in the athlete not reaching optimal performance. Conversely, if a new device is under-reading or is inconsistent, an athlete may inadvertently be training outside his/her physical capabilities or target zones. Year-on-year performance gains or losses can be as low as 10%, but this could mean the difference between a mid-field finish and a podium position.

The solution ? Make sure to get the best possible raw data collection platform that you can afford, preferably with independently verified accuracy and precision. Only that way will you be confident that your power/weight ratio of 6,5 (we can all dream !) is comparable with the top Pro Tour riders.