Revision as of 15:30, 25 March 2017

Stryd is one of the most useful and impressive running technologies available. While a lot of the Stryd marketing focuses on their "running power meter" functionality, I see the greatest value coming from their accurate measurement of pace and distance. If you look at my testing of GPS Accuracy, you will see that Stryd gives a far more accurate measurement of distance than any GPS watch. It accurate enough that it really doesn't need any calibration, and unlike GPS, it should work in any situation. Not only is it providing accurate distance measurement, but it provides real-time pace, something that's impractical to achieve with any GPS device. While I believe Running Power Meters are quite different to cycling power meters and are more of a "power estimate" than a meter, I think that if you understand their limitations they can still be useful. At $200 I think that Stryd is good value for money, and it's one of the few running devices that I would replace unhesitatingly if I lost it. Currently Stryd is only available via their web site. (I don't have any affiliate or sales link with Stryd.)

1 The Metrics

Stryd provides an array of different metrics, which vary both in accuracy and usefulness.

• Pace and Distance. My testing has shown that the Stryd is stunningly accurate in its measurement of distance. It's the only Footpod that is accurate enough that it doesn't require calibration, something that greatly improves its usability. If you look at my testing of GPS Accuracy you'll see that the Stryd is right at the top of the chart, and is far more accurate than any GPS device I've tested. This accuracy is also reflected in its measurement of current running pace. I think that this feature alone makes the Stryd good value for money.
• Elevation. I suspect that Stryd is providing remarkably accurate elevation information. Due to limitations on data export I've not been able to perform any statistical evaluation, but anecdotally the data looks really good. Obviously, the Stryd cannot provide any absolute altitude information, but it does seem to be able to measure relative changes far better than either GPS or barometric altimeter's. At the moment, this information is not as directly usable as I'd like, as it's only available in real time via their, not their watch integration.
• Running Power. As I talk about in my page on Running Power Meters, I believe that they are a flawed concept, but if you understand the limitations they can still be useful. So, Stryd it does not actually measure running power, but attempts to estimate what the power would be based on the things it can measure. My assumption is that it is mostly using pace and elevation change, possibly along with Cadence or Ground Contact Time. As discussed in the section below on the testing running power, I have found that the Stryd seems to give a reasonable estimate of relative intensity for flat and uphill running, but seems to underestimate relative intensity on the downhill sections. I've found that with some testing to understand the limitations, I've been able to use Stryd to more effectively pace myself on hilly runs.
• Form Power. Stryd describing this metric as the "running in place power", but it's unclear what that really means more if it has any value. They suggest that a decrease in this value represents improved Running Economy, but there is nothing to back that up.
• Cadence. Cadence is fairly trivial to measure for even a far simpler Footpod, so Stryd nails this easily. It's arguably one of the most important running metrics, so you should pay attention to this. There are far cheaper ways of measuring cadence (like MilestonePod) but it's nice to have this included in the Stryd.
• Ground Contact Time. Ground Contact Time is how long each foot spends on the ground, and it's frequently suggested that a lower value represents a better Running Economy though the research is mixed.
• Vertical Oscillation. Because Stryd is a Footpod, it has no way of measuring Vertical Oscillation. Therefore, Stryd is estimating vertical oscillation from things like Cadence and Ground Contact Time. My testing indicates that Stryd is underestimating my vertical oscillation, which is perhaps not a surprise given this is only a mathematical model. Stryd suggests that this is because they are measuring the vertical movement of the runners' center of gravity rather than just the torso, but the research suggests that should be no difference between the two measurements while running.
• Leg Stiffness. It's possible to model a runner with their legs representing a spring and the rest of the body as a mass. When a runner lands, they decelerate their bodies vertical movement and the rate of deceleration can be used to estimate the stiffness of the "spring". The stiffness of this hypothetical spring is related to Cadence, with a higher cadence having a stiffer spring. The preponderance of evidence suggests that a stiffer leg stiffness is more efficient, and that fatigue tends to soften the stiffness. There are various ways of estimating leg stiffness, such as measuring the vertical ground reaction force and vertical movement. This gives the force applied and the amount of deformation of the "spring", resulting in a reasonable estimation of stiffness. Another approach is to use ground contact time and cadence, which is how I suspect Stryd is estimating stiffness. Personally, I suspect that this is an overly simplistic model.

Note that Stryd provides no information on foot strike parameters such as pronation, or foot landing position.

2 The Pod

The Stryd pod is similar to other Footpods. It's slightly larger than most, and has an LED status light on the top. Stryd uses a rechargeable battery that they claim lasts about 20 hours. That's good enough for most runners, but may be a problem for longer ultramarathons. Stryd recharges wirelessly, which is nice, as I hate fiddling around with connectors or losing specialist cables.

The pod has an LED on top (the white area in the middle) that's used to convey status.	Here's the Stryd pod on the shoe.
Stryd is my first device that supports wireless charging.	I don't normally show packaging but the design is so minimalist yet elegant that I thought it was worth a picture.
Here's a visual comparison with some other Footpod's. From top left to bottom right there is the Stryd, Garmin Footpod, MilestonePod v3, Polar Stryd Sensor, Adidas, and MilestonePod v2.	This is a view of the Stryd and Garmin Footpods with their cradle to clip into the shoe laces.
Looking sideways you can see the Stryd and Garmin Footpods have similar openings for shoe laces.

3 Watch Integration

Stryd has remarkable levels of integration with a wide array of running watches. It supports both Bluetooth and Ant+, and it can look like a standard Footpod, or a cycling power meter. There are a large number of combinations of watches and configurations to consider, and I will return to this issue and update this section in the near future. I'll also include some recommendations of the best watch is to use with Stryd.

4 The App

The Stryd app is rather limited, and I mostly use it just to sync data with the web site.

The home page of the app shows you the recent runs and the "start training" button to run with the app.	If you click on a previous run you'll see this brief summary. For more information you'll need to use the Stryd web site.	If you run using the Stryd app, this is the screen you'll get. It shows some basic data, but given a phone screen it would be nice to have some graphs and more data, customized fields, etc.
When running on a treadmill you can tell the app how fast your running and what the incline is. I haven't been able to work out what this is for, so please let me know if you find out.	There's a calendar view to pick a run for a given day.	The setting screen allows you to set the Stryd to transmit power as Cadence. Due to limitations on the cadence field, you'll get power divided by 100, so 215 watts will be 22.

5 The Web Site

I rather like the Stryd website. There's the usual calendar views of your runs, but there is some useful analysis you can do within their website. The biggest shortcoming I've found is that you can only show data by time, not by distance. If you look at the first graph below the hill shown by the purple line looks asymmetric because I'm going down much faster than I'm going up.

This view allows you to overlay various metrics in a different colors. Here you can see one of the hill repeats I analyze further down this article, so I have elevation, power, and pace shown. You'll also see some average values in the circles above the graph. If you look at the very bottom of the image you can see a slider where you can zoom in on a subsection of a run. Here I'm showing you just one he'll repeat.

If you click on the circles some extra data pops up up. I'm not quite convinced by this approach to displaying data, and I feel that it all could be presented in a smaller space and still be quite legible.

This shows the pop-up selection for various attributes.

The comparison view is rather uninspiring, just allowing you to see your estimated power for two separate runs.

6 Data Analysis

Because Stryd is compatible with so many watches, it's fairly easy to get most of the data into an app for analysis.

• Pace and Distance. If you have your watch set to take a distance from the Stryd, you'll naturally have access to distance in pretty much any application you can get your data into. Because you have accurate distance, you should have accurate pace to analyze as well.
• Running Power. Getting power into an application for analysis will depend on which watch you are using and how you got it configured. If you're pretending this is a cycling power meter, then everything should work just as it would for cycling power. If you're using the Garmin Connect IQ data field, then you'll need to have an application that supports importing customer data. I found that Golden Cheetah worked fine. If you're getting power as Cadence, then most applications should be able to handle that, but obviously you'll have it in the cadence field and it will be 10x lower (210w will be 21 steps/minute.)
• Cadence. Cadence is well supported by pretty much any analysis application, unless of course, your using the cadence field for power. Note that the Garmin Connect IQ data field will record the cadence value from the Footpod in addition to any other Cadence device.
• Elevation. I've found it tough to get elevation data into an analysis application. I'll update you if I find a good way of doing this.
• Form Power, Ground Contact Time, Vertical Oscillation, Leg Stiffness. These fields are all available via the Garmin Connect IQ data field, and can be viewed using Garman Connect.

A view of the data collected using the Garmin Connect IQ data field with Stryd also connected as the Footpod, using a Garmin Fenix 3.

7 Testing Distance and Pace

I tested the distant accuracy of the Stryd using the same basic methodology as I do for my GPS Accuracy testing. (See GPS Testing Methodology for details.) I used a Polar M400 configured to use the distance from the Stryd footpod. As you can see from the results, the Stryd is remarkably accurate, far better than any GPS device I've tested. My testing of pace is more anecdotal, but I've been able to hit my target paces using guidance from my Stryd. I've tested using a range of paces from 10:00 min/mile to 6:30 min/mile and the Stryd allows me to cover a given distance in just the right time.

An infographic of the accuracy of the GPS running watches. The top right corner represents the most accurate watches. (This graphic uses ISO 5725 terminology.)

8 Treadmill Problems

The Stryd is so accurate outdoors, I couldn't understand why it was so bad on a treadmill. To my chagrin, I discovered that the Stryd Footpod was right, and my treadmill was wrong. Accurately calibrating a treadmill proves to be rather more complex than I'd have expected, as the treadmill will slow down when your foot is in contact with the belt, then speed up when you're in the air. For details on how to evaluate your treadmill have a read of Treadmill Calibration.

9 Testing Running Power

I don't view Stryd as a "power meter", but a way of estimating relative training intensities, rather like Heart Rate. There are some important benefits to using Stryd over heart rate for estimating relative training intensity. The graph below is showing a run up and back down a local hill. It's not a very long Hill, but it is fairly steep, averaging 6% with sections nearer 10%. The first graph below shows me running hard up the hill, then resting on the way back down. You can see that both my heart rate and power estimate rise on the uphill and decline on the downhill. You'll notice that my heart rate response rather slower to the change in intensity than the power estimate, making it more useful for providing a useful, real-time estimate of intensity.

Traditional Hill training, focusing on high intensity on the way up and easy back down. Power data from Stryd, Heart Rate from Wahoo TICKR Run, elevation/distance/pace from GPS using Suunto Spartan Ultra.

Another interesting use for Stryd is to provide more even pacing on hilly courses. The graph below shows me attempting to keep an even effort based on the Stryd power estimate. My goal was to keep my intensity at about the same level as a running at 8:30 min/mile pace on level ground, which at the time of this run is about 200-210 watts and about 135-140 BPM heart rate. You can see that I was moderately successful, though both my heart rate and estimated power were a higher than my target. I was surprised by just how slow I had to go up the hill to compensate for the slope. You can see this more accurately in the lower image that's looking at the Stryd pace data. Even though my pace has dropped to 12:18 min/mile my heart rate and estimated power are both well above target. This means that to go up this hill with the same effort as 8:30 min/mile on the flat, I'd need to drop to quite a bit slower than 12:30 min/mile pace. Here the estimated power data is far more usable than heart rate, as I can tweak my pace moment by moment rather than waiting for my heart rate to adjust. I found that in practice, the Stryd estimated power output is remarkably effective at giving me real-time feedback of my exercise intensity on level ground and uphill. Though this is still a modeled, estimated power intensity, for uphill and level ground, it's far better than anything else available to us and is good enough for real world usage.

Trying to maintain an even effort up the hill. Power data from Stryd, Heart Rate from Wahoo TICKR Run, elevation/distance/pace from GPS using Suunto Spartan Ultra

The same hill repeat with data from Stryd.

Sadly, things are not quite so rosy on the downhill sections. The graph below shows a hill repeat where I took it easy on the way up, and then pushed the pace hard on the way down. You can see that during the downhill section, my heart rate has risen a little, but the estimated power output has dropped massively. You can see my pace a little more accurately on the lower chart that uses the Stryd data, and my pace is hitting 6:30 min/mile. This matches up with my overall experience of Stryd, which consistently underestimates my exercise intensity for downhill sections of my run. This is a little limiting, as it means I come to use Stryd to pace myself when running downhill. I'm reasonably confident that Stryd will update their model to improve this, but it is currently a noteworthy shortcoming.

Easy up the hill, and high intensity fast downhill running. Power data from Stryd, Heart Rate from Wahoo TICKR Run, elevation/distance/pace from GPS using Suunto Spartan Ultra.

The same hill repeat with data from Stryd.

10 Stryd Power Estimate or Heart Rate?

The use of heart rate for training has been established for many years, and heart rate based training has some useful advantages, as well as some significant shortcomings. I think the Stryd estimate of power output overcome some, but not all of the issues with heart rate based training. Personally, I don't see this type of power estimate completely replacing heart rate based training, only augmenting it.

• Heart rate responds to exercise intensity with a delay, while Stryd's power estimate is much closer to real-time.
• During longer exercise, Heart Rate Drift occurs that generally causes a higher heart rate for a given intensity. The reasons for this drift are complex, and include dehydration, fatigue, carbohydrate depletion. Using Stryd's power estimate ignores this drift, though it's unclear to me when to use heart rate and when to use the power estimate. In some situations, it seems likely that the drifted heart rate is a better estimate of intensity than an unmodified power estimate.
• There is a widespread myth that Maximum Heart Rate can be calculated, leading to some erroneous assumptions of how a given heart rate relates to the percentage of exercise capacity. In practice, both Heart Rate and maximum estimated power require a practical test.
• A common use of heart rate data is to allow an athlete to train at their Lactate Threshold, often referred to as Tempo Runs. The belief is that this training intensity is especially beneficial, though the available science indicates the opposite. If Tempo Runs made sense, then Stryd's power estimate would be quite valuable for hitting that pace accurately. I'm sure that many runners will use Stryd this way, even though the science indicates it is ineffective.