When it comes to GPS running watches, there are generally three metrics that people care about:
-
GPS accuracy (distance, pace, track)
-
Heart rate accuracy (optical wrist-based or chest strap)
-
Training load metrics (VO2 Max, recovery time, training effect)
For some, cadence and running dynamics are also important, but I’ve found that most readers ask me to recommend a watch with the above three sensor types. While I’ve discussed GPS chipsets and HR sensors at length, I thought it would be interesting to discuss running dynamics sensors in GPS running watches because they have a few caveats that few people know. They shouldn’t be trusted with absolute values — only with identifying trends. Why? Let’s discuss.
At the moment, the two most popular running dynamics sensors you will find in many GPS running watches are the Garmin Running Dynamics Pod and the Stryd footpod. Across all of the watches I’ve reviewed in the past few months, I’ve seen one of these sensors referenced in every watch that supports advanced running dynamics. In other words, they’re very common and referenced all across the industry.
Many people don’t know that these sensors show a running power estimate by default — not an absolute value. Below is an explanation of the power estimation used by default for the Garmin Running Dynamics Pod.
From Garmin GPS running watch training brochure.
In short, since the absolute values from running power sensors aren’t very reliable (more on that soon), the companies that produce the sensors have generally opted to use relative indexes instead. In Garmin’s case, the running dynamics pod adapts and calibrates itself frequently to set new baselines. This means a power of 250W should always mean ‘typical effort at threshold’ and 300W ‘above threshold’ regardless of your terrain. In other words, this number helps identify trends but not absolute values.
Polar running dynamics data from Polar RC3 GPS user manual.
Above is an explanation from Polar that recommends that GPS watch manufacturers don’t use the raw output from their running dynamics sensors but rather use the running power algorithm. This algorithm functions the same as Garmin’s and sets 100W as the baseline ‘easy running’ power over a set period (which manufacturers can adjust). Therefore, any number below the baseline indicates an easier effort, and any number above represents a harder effort.
So, why do these manufacturers recommend this approach? It’s because consumer-grade running dynamics sensors aren’t very accurate unless you know the exact biomechanical profile of the runner. This is because running power sensors are more sensitive to certain gait patterns and less sensitive to others.
The above image from a running watch accuracy study shows how a running dynamics sensor has different accuracy for different running cadences and paces.
As you can probably guess, since running power sensors have different accuracy at different cadences, and we can never know for sure the exact gait profile of every runner, an absolute running power figure is not particularly useful or accurate. While many older running power sensors (and some current ones) do give absolute watt values, and even the sensors mentioned in this post can be adjusted to show absolute values, they’re not fully accurate and hence the choice by Garmin and Stryd to use relative indexes and calibrated estimates instead.
Based on this information, I don’t think anyone should choose a consumer-grade GPS running watch based on running power performance alone. None of them are perfectly accurate, and the running power recorded by the device should be used only to identify trends in training load.
The Garmin Fenix 7 shows only a running power estimate (increasing, decreasing, or stable) for running dynamics. While I didn’t realise it then, this makes a lot of sense with the limitations of running power sensors in GPS running watches.
One final note is that since these sensors regularly readjust their baselines (depending on the period set by the GPS watch manufacturer), overall trends shouldn’t be trusted — only short-term trends over a run or less. Some watches will allow you to adjust the running dynamics calibration period, but it’s not common.
At the end of the day, having a running dynamics sensor in a GPS running watch is useful, but it’s important to know the limitations and caveats of such sensors. If you have any questions, please don’t hesitate to ask below!
