Standard deviation of run power samples by Jim Vance
13 June 2018
One of the common practices and beliefs in power data, is looking to have some insight beyond just average power. Some companies have tried to create an algorithm to normalise the power data, and even Today’s Plan has what is called “Adjusted Power”.
While Adjusted Power helps give a representation of what the load on the athlete would be if they road perfectly steady, there’s a better unit of measure which isn’t considered by most, but is vital, the Standard Deviation, SD for short, or in the charts as STD DEV.
Standard Deviation is a measure used to quantify the variation or dispersion of samples within a given sample size. In the case of power data, (bike or run), it would represent how the samples compared with the average power value. This gives a much better idea of how much variance of intensity a workout or race had.
One of the great benefits of SD is that it uses the same units it is comparing to, so it is a fair and relevant value for any sample size.
A small SD represents a much steadier effort, with a majority of the recorded values within that interval or workout, being close to the overall average, or mean, of power values recorded. A large SD value would represent a very large dispersion between the highest and lowest sample power values, meaning a lot of time spent in higher and lower intensities, such as a fartlek run, or a bike race with a lot of attacks and sitting in the pack.
You can find these values within the Activity Viewer window, in the chart of intervals. If you find yourself with large SD values, especially in your harder workouts or races, you will likely need more recovery time, as higher variance in hard workouts tends to mean a greater stress on the body.
If you’re looking to judge pacing skill, SD is a great way to see how well an athlete kept the sample sizes together for a given duration or effort, as a low SD value would represent a very steady effort.
When reviewing your workouts, be sure to pay attention to SD, to better understand the quality of pacing of an effort, or the intensity variations, which may require more recovery time than a steady effort. Using this data can help athletes enhance performance and recovery.