Running With Power
Some Observations on Native Power Monitoring in the Coros Pace 2
A few months ago I was in the market for a new running watch. My daughter started her freshman cross-country season at Patrick Henry High School here in Roanoke and I wanted to encourage her to adopt a data-focused approach to her training. Assuming (probably correctly) she would never keep up with a chest strap AND a watch, I gave her my Garmin Forerunner 55 with its wrist-based heartrate monitoring.
While shopping for a replacement I came across Coros, a newbie in the wearables market. One of their early products is the Pace 2 and their value-add of integrated power monitoring at a reasonable $199 price caught my attention. Power? Sure, I’ll take some of that.
Pro cyclists have trained with power data for decades because the technology is straightforward. A mechanical strain gauge integrated into the crank or pedal does its magic. And mass adoption by weekend warriors and tri-girls and guys has driven the price down to a somewhat-affordable-for-cycling $600 entry price.
Power is not new in the running world either. But it’s not super-advanced. It’s generated in the stride push-off, and measuring that force while maintaining shock absorption in your cushy shoe soles is tricky enough that it’s seen as the money-grail of running. The current best (affordable) tech seems to be an external pod that attaches to the shoelaces or waist strap. The Coros Pace 2 that uses an internal accelerometer and fancy algo-magic to derive power is not the first watch to do so, but it’s the initial entry in the “budget” category.
I did a lot of reading and ‘rithmetic, attempting to learn the dark secrets of Coros power. So far I’ve failed. And they’re not offering an explanation on what’s in the tiny black box. They just asked me to run with the watch for a month or so before I started relying on their power numbers. I guess their bots wanted to get to know me before any long term commitment.
After the break-in period, my first real test was a turkey trot 5k, a fun little Ohio outing with my nieces.
Time: a little over 19 minutes. Good enough for P1 Masters in a small Thursday morning race, but I have faster days ahead.
Average Heartrate: 166 bpm. This tracks what I’ve been seeing the past few months. I redline at around 170 these days.
RPE (Perceived Exertion): 10. All out. It was a 5k, after all.
Average Power: 348 watts. Multiplied by 95% yields a Functional Threshold Power of 331 watts.
All this data allowed me to reset my zones, start a new training cycle, and “run to power”. Or at least start making some power-based judgements and comparisons. Over the next six weeks, I varied my running. Trail, long and short. Road, flat and hilly. Track, round and round. Sometimes with power as the target. Other workouts with heartrate, pace, and exertion as the target and power as a derivative.
I recently took another test and ran the annual Frozen Toe, a 10k trail race that feels like it goes up the entire route.
Time: a little under 46 minutes. Good for P1 Age Group, but again, hopeful for faster days ahead.
Average Heartrate: 164 bpm. On point, given the amount of technical downhill where it’s hard to go max heartrate effort.
RPE: All out.
Average Power: 294 watts. Hmm. Feels low.
After three months with the Coros Pace 2 here are my initial takeaways on running power:
It’s data and it requires context.
This isn’t different from heartrate or pace or even RPE. An understanding of the fundamentals is required and even a seasoned understander can get crossed up on a given day. Intuitively, a runner exerts more power on an uphill and less on a downhill. But it’s not linearly related to heartrate. I can maintain a HR close to my Functional Threshold on a downhill but I’ve found it difficult to maintain Functional Threshold Power (FTP). On a recent interval workout I threw in some downhill reps at the end; my pace was close to max but my power output was low Zone 3.
It might be useful in controlled workouts.
Most serious runners head out the door with a plan. Distance or duration. Pace, heartrate or exertion targets. A route, an understanding of the elevation changes, and a rough idea of how to close the loop and get home. Power data can be used as a target in some situations. But so far I’ve found it more useful for Zone 3,4 & 5 stuff. For the broader range of the slower Zone 1&2 workouts, power and heartrate appear to track linearly, rendering the power data extraneous. On trails, the variability is even less controlled. In my second race test at Frozen Toe, my low average power of 294 watts demonstrates this. It’s very difficult to deliver significant and consistent power on trails. There might be runners out there that can run to max power on a technical, rocky, rooty southwest Virginia downhill. But I hope they wear a GoPro. They’ll deliver some spectacular crash reels at some point.
It doesn’t add a lot of value.
I’m planning to mine my power data for a few more months with some varied and controlled workouts. But so far I haven’t seen anything mind-blowing or revolutionary. If you primarily train on the road and track there may be value in the extra data points. But a good coach looking at the combinations of heartrate, pace, and RPE can get you to the same finish line. Algo-derived power might just another way to sell more wearables.