Running With Power

In my view running with power (in other words, using watts to gauge intensity) is superior to all the other metrics - but does that superiority matter?

I have had a Stryd running power meter since it was first released in 2016 as the first generation non-wind model. In 2019 they released the ‘wind model’ which reported the extra power required to overcome air resistance. This allows you to know how much power you need to run into a headwind, the power saved when running with a tailwind, or the power saved when drafting off a pack of runners. In September 2022 the Next Gen Stryd was launched with improved responsiveness for the most accurate pacing and finally, at the end of 2023, the Stryd Duo was released allowing you to combine two Next Gen pods which provides left-right running balance metrics to empower runners with deeper insights into their running.

The watch manufacturers have attempted to incorporate power into their products but if you are looking for the gold standard, Stryd is still the leader, not only for the accuracy and responsiveness of the pod, but an entire power-based training system and analytics platform supported through its app and web Power Center.

Two things drew me toward using power for running. One was simply practical. I wanted to get accurate pacing on my treadmill and other situations that lacked a GPS signal. The second was that I had experienced firsthand how superior power-based training was in cycling, and intuitively I thought it must also work for running.

Why Power? Let’s compare the different metrics

The main benefits of using power are:

1. Objective Measurement: Power output provides an objective measure of the actual work being done by the runner, regardless of external factors like terrain or weather conditions. It is the closest proxy to the actual metabolic load that you can use.

2. Instantaneous Feedback: Power responds quickly to changes in effort level, providing immediate feedback to the runner during intervals, hills, or changes in pace.

3. Consistency: Power is consistent across different terrains and conditions, making it easier to maintain a consistent effort level regardless of factors like elevation or wind.

4. Training Zones: Power-based training allows for precise targeting of training zones, ensuring that workouts are tailored to the individual's fitness level and goals.

5. Less Susceptible to External Factors: Unlike heart rate, power is not affected by factors like dehydration, fatigue, or caffeine intake, making it a more reliable indicator of effort.

6. Useful for Indoor Training: Power can be particularly useful for indoor training on treadmills, where GPS-based metrics like pace can not be used.

So given 8 years of using power in running am I still a convert?

The honest answer is both yes and no. I still believe in the superiority of it over pace or heart-rate, but I guess my problem is I’m still a romantic when it comes to running. I like the freedom of no metrics at all and I believe in race day magic! The idea of running every session under clinical scrutiny, or racing to a prescribed power number just doesn’t float my boat. Neither does getting into the weeds of statements like using an “extension to the hyperbolic CP model to include a slow component or Riegel-like effect directly into a variant of that model” particularly interest me. Now you can certainly avoid the complexity by using Stryd’s auto-critical power calculator and other things built into the ecosystem, but still, something about reducing run performance to numbers isn’t for me (even if it works!).

I run mainly on feel or RPE (Rating of Perceived Exertion), triangulating feel and pace (I almost always ignore heart rate) this is something I enjoy. Then seeing that this triangulation gives a consistent power number is testimony to getting it objectively right.

Marathons are hard to pace. Let’s be honest, they are not easy in any regard, and I tend to do one every 4 years as that’s how long it takes for me to forget the experience. My last attempt at the distance was the Uppsala Marathon at the end of 2021 at the age of fifty-one. My taper week was not by the book. I had a boozy week in the UK taking my son to visit family and pubs. I landed back in Sweden less than 12 hours before the race started. I had also given myself a lofty goal of a PB (I had run 3hr8 previously) and the race target was sub three hours. I shared the race goal with a run buddy so there was no backing out - we were in this together and I was committed.

These are the 5k splits with Pace, Cadence, HR and Power:

Note that I was not pacing on power. I had my watch on auto-lap at 1km and I had one metric on my watch face - average lap pace. To hit my goal I knew I had to average 04:15 or lower. It was the only metric that mattered! That, and keeping my RPE in check so that my central controller wouldn’t shut me down.

The first 10km felt easy. We had to continually pull ourselves back from getting carried away, even so, it took a while to dial in the pace. Half way it still felt good. By 30km it was becoming a struggle and I can say that holding on through 40 km is one of the hardest things I’ve done. The last km or so was uphill with the finishing line at the castle on top of a steep incline (some evil race director did that to us). In the last 200m, I had the highest power of the race (333w) albeit one of the slowest splits.

Pacing on a race target pace and RPE worked great. I can see that I didn’t get stressed and managed my effort, for example, kilometer 31 was 4'10/km at 287w and kilometer 32 was 4’30” at the same power output. My RPE was dialed into the metabolic cost of the terrain. Racing on power would have taken away any doubts in situations like that, but I don’t think in this case it would have altered the outcome. Cadence was consistent. I don’t think I could have executed the race plan any better.

I had an average power of 290w which is 4.2 w/kg. From the book Secret of Running they predict 4.0 w/kg would give a marathon time of 3:02:51 and 4.25 w/kg would give 2:52:05 - so 4.2 w/kg resulting in 2:59:31 is about right considering the course is not the fastest; and would predict a time closer to 2:53:00 on a course like Berlin or Valencia. As for my running buddy, he nailed it too. Respect. Stryd’s race calculator recommends 89-91% of Critical Power (CP) for a marathon. So that would put my critical power at 318 to 322w. My Critical Power at the time was 320w. Synchronicity!

I’ll continue to use power as an additional training metric, and who knows, maybe one day I’ll go all in. It would be an interesting experiment to see whether the superiority of the metric would translate into superior results for me. I’m pretty sure if you are one of the many runners that finds pacing and distributing effort in a race difficult then running with power is for you.