Running power meters: why 'watts' in running, and where they're overrated

Power has long been the main metric in cycling: athletes pace races by watts, not by heart rate. Logically, manufacturers brought "watts" into running too — the Stryd footpod, as well as Garmin, COROS, and Polar watches now show running power right on the display. It sounds tempting: an objective number that doesn't lie. But before building your training around running watts, it's worth understanding one thing — these are not at all the same watts as in cycling.

How running power differs from cycling power

The difference is fundamental and lies in the way the number is obtained.

In cycling, power is measured directly. In the pedals, crank arm, or hub there's a strain gauge that captures torque (force) and multiplies it by the angular speed of rotation. This is a physical measurement of applied force, and a good cycling power meter hits real power with an accuracy of about 1.5%.

In running there's no such force sensor under your foot in a normal workout. Running power is not measured but modeled. The accelerometer and gyroscope in the footpod capture movement, and then the algorithm calculates the presumed power from your mass, speed, gradient, and stride characteristics. This is an estimate, not a direct measurement of force.

The practical consequence is simple, but often ignored: each brand has its own model. Stryd, Garmin, COROS, and Polar watts are computed by different formulas and are not comparable to one another. You can't compare your watts with someone else's or carry numbers between devices — these are different scales.

What the measurements say

Let me be honest right away: the evidence base for running power is weaker than in "big" physiology. The studies are small, often in controlled conditions, and no scientific consensus has yet formed. So, without exaggeration.

The key validation study (Stryd on an indoor 200-meter track, speeds of 8–19 km/h) found a strong linear relationship of modeled power with oxygen uptake (R² = 0.82) and with external mechanical power (R² = 0.88). But at the same time Stryd systematically underestimated absolute values of power — the authors had to propose a linear correction. The ground contact time and leg-spring stiffness metrics turned out to be valid.

It's worth distinguishing two different properties separately:

Reliability (repeatability) for Stryd is generally good: coefficient of variation <3%, mean error <2.5%, correlations >0.76 between repeats. That is, under similar conditions the device produces a stable number.
Absolute validity — how much this is "true metabolic power" — is limited. The device stably repeats its own estimate, but that doesn't mean it equals real energy expenditure.

There are caveats too: some research on five commercial technologies casts doubt on consistency between devices, especially uphill (power, contact time) and on descents (vertical oscillation). A 2024 trail study directly advises interpreting the data cautiously and measuring with one and the same device. The sensitivity to wind and terrain in some devices also hasn't gone anywhere.

Who it's actually useful for

The main practical plus of running power is the metric's instant response. Heart rate lags by tens of seconds and "drifts" toward the end of long work. Watts respond almost immediately. Hence the clear scenarios where the device genuinely helps:

Trail, hills, wind. GPS pace deceives on a climb, heart rate lags — but power responds instantly and accounts for the gradient.
Evenness on a long tempo. It's easier to hold a steady effort rather than accelerating downhill and "dying" on the climb.
Intervals without heart-rate lag. It's hard to nail the pacing of a short segment by heart rate; by power the response is instant.

How to use it correctly

Build YOUR zones from a test. Don't take someone else's "threshold watts" from the internet — your model is different, and so is your physiology. Calibrate your zones from your own test.
Don't mix brands. Switched from Stryd to Garmin — you can throw out the old numbers, you need a new test.
Remember the conditions. Strong wind and loose terrain can distort the estimate on some devices.
Treat the watt as a model. It's a convenient effort guide, not absolute truth in joules.

And who doesn't need it? If you run on a track or a flat road, where GPS pace and laps are already accurate, running power adds little — you're overpaying for a metric that a stopwatch replaces.

The bottom line

Running power is modeled (accelerometer + mass + speed + gradient), not measured by force as in cycling.
Watts from different brands are not comparable — you can't compare yours with someone else's or mix devices.
Stryd's reliability (repeatability) is good (CV <3%), but absolute validity as "true power" is limited; the link with VO₂ is strong (R² ≈ 0.82), yet absolute values are underestimated.
Myth: running watts = cycling watts. No — these are different quantities.
Myth: power is more accurate than heart rate in everything. No — more accurate in response, but it's a model estimate, not a gold standard.
Myth: FTP from running = FTP from cycling. No, you can't carry it over.
Genuinely useful on trail, in hills, in wind, for evenness and intervals; on a flat track it's excessive.

Sources: Cerezuela-Espejo et al., "Are we ready to measure running power? Repeatability and concurrent validity of five commercial technologies"; Stryd validation study on an indoor track (Frontiers in Physiology), https://pmc.ncbi.nlm.nih.gov/articles/PMC7404478/; García-Pinillos et al., "Reliability and validity of the Stryd Power Meter during different walking conditions", https://pubmed.ncbi.nlm.nih.gov/34896839/; "Validity of the Stryd Power Meter in Measuring Running Parameters at Submaximal Speeds", https://www.mdpi.com/2075-4663/8/7/103; TrainingPeaks, "The Differences Between Running and Cycling Power", https://www.trainingpeaks.com/blog/the-differences-between-running-and-cycling-power/