COROS POD Data: Running Power and Dynamics in Comparison

You may be curious to see what sort of data the COROS POD gives and how it compares. Well, here goes…

Power

Running power is one of the more interesting measures that have come to some prominence recently.

Here, I want to just have a quick look at the power numbers given by COROS POD, Stryd, and Polar Vantage V:

Example power recordings from COROS POD, Stryd, Polar Vantage V
Zooming in on a selected range (where I never stopped)

I’ll just leave this like so, though, because power needs a much more detailed discussion… including on why power numbers should not necessarily even be compared like that.

Power Graph in COROS App

What makes rather more sense is the look at what more the COROS app can make of the recorded power data:

For one, there is simply the graph of the recorded power – which is, of course, a lot like the one above. Or actually, should be the same, just with a bit less detail.

Power graph from COROS app

(There are some more differences because I did leave out a bit of the recording in the graphs I exported from quantified-self.io)

Power Analysis

Where the COROS app and its use of data from the POD is really interesting is in the analysis of horizontal vs. vertical vs. lateral power.

Stryd does have “Form Power” which sounds to be a somewhat similar measure, but otherwise, there is nothing (to my limited knowledge) that is the same.

Running Power Analysis in Coros App
Running Power Analysis in Coros App

Maybe it was because of the focus on cadence? Maybe it’s my running style? Whatever it is, if that analysis is correct, then a third of the power I produced in my running was just for bounce, not really for propulsion…

Vertical Oscillation

Talking of bounce… COROS POD and Stryd both measure vertical oscillation, so there’s something to compare – and it’s something that should be related to that vertical power reading. (That relation would only become clearer or more useful if I looked at it over more training runs and saw about changes…)

Vertical oscillation over (nearly) the whole run

As everything, this looks a bit shifted but overall, quite comparable.

Stryd appears to measure less vertical oscillation than the COROS POD; that might well be due to their different position/location (on shoe vs. waistband).

Zoomed-in view of vertical oscillation data from COROS POD and Stryd

Not much else to say from a closer look at a section of that recording. It looks shifted, and it looks quite similar (especially if that shift weren’t there… which is what has me so puzzled; the time should have been the same for all devices, after all, not different by two minutes, in any case).

Stride Analysis in the COROS App

Vertical oscillation is not shown in the COROS app’s running (power/dynamics) analysis, but there is quite a bit of data on one’s stride.

Stride data analysis in the COROS app

There is the stride ratio, the ratio between stride height and stride length, which is a measure of running efficiency. It tells how much of each step goes into the length – i.e., forward motion – or the height – i.e., the bounce. This is also, of course, a reflection of vertical oscillation and reflected in vertical power.

Cadence

The type of run I was doing there should become clearer from a look at the cadence data. It was a running cadence / frequency-focused training session, so that had better be reflected…

And it is.

Times I was stopping are making things a bit jaggier and harder to read, but all in all, it is clear – including from the stops – when I had recovery breaks and when I was running, focused on a higher step frequency.

There are a few oddities here. In the basics, the recordings reflect the actual; in some details… I don’t know.

Vertix (i.e., COROS POD) and Forerunner 945 are pretty much in agreement. Polar Vantage V and the “Unknown Device” (which is the Racefox Run app, using data from the Polar H10 HR sensor) are rather similar to the others, as well.

The Stryd data is odd; I keep re-checking if maybe it was just set to a wrong time, but I don’t see how that could have been the case. The Stryd data looks like it was lagging behind, though.

A closer look at a part of the recording just adds to the impression that the Stryd was somehow lagging behind.

Racefox just doesn’t filter the data, that’s why there are more spikes from that.

Everything else is pretty similar (as far as one can tell).

Cadence in the COROS App

Cadence is actually something that the COROS app has always shown, iirc. It’s towards the top of the running workout data, even before running efficiency – and here, it very nicely reflects the type of run this was.

Ground Contact Time

Ground contact time, here labeled as stance time (if I’m not very much mistaken) and/or ground time is said to be related to cadence and another one of those elements of running technique one is supposed to have a good look at in training.

How does that measurement look?

Stance/ground time, from COROS POD and Stryd

Big surprise, the data looks like it would be very similar if only it weren’t shifted. Or similar, but shifted.

In instances where the stance time became higher, the Stryd pod sees a higher stance time than the COROS POD; in other instances, their readings are quite similar.

Selection of stance time graph/recordings from Stryd and COROS POD

A zoomed-in closer view just reinforces that point. Shifted, similar, with higher readings in instances of higher stance time from the Stryd.

Again, that difference may well be due to the position of the Stryd pod on one’s shoe, the COROS POD on the waistband.

Ground Time in the COROS App

Again, we also find this data set in the COROS app:

There, it is found (as it is on the watch display during a run) as “Ground Time”.

The inverse relationship of ground (contact) time to cadence is actually nicely visible here. I.e., when I was running at higher cadence, my ground time was lower.

Speed and Pace

Cadence should perhaps have an impact on speed and pace, but might not. Speed and pace should fit together, though – even if this should be data coming from GPS, not a pod… or should it?

Speed over the whole frequency-training run
Pace over the whole frequency-training run

Nicely instructive, but only about data issues, these two:

Speed being in m/s makes it a pretty sensitive measure; the min/km pace might already be less sensitive just from the units used – and with the time where my pace was extremely slow (because I was basically stopped), there seems to be a big difference just because the pace data gets compressed into a graph that looks like I was always running at a constant pace, by and large.

A closer look at the speed graph shows that most devices recorded quite similar speeds, actually. It’s just that the Racefox Run’s unfiltered recording hides everything else behind its jagged lines…

Speed for just a section of that run…
… and the same with pace

Looking only at parts of this run drives that point home. The Racefox Run app records without filtering, thus has very jagged graphs here.

Hidden behind it are the other device’s recordings, which are mainly in quite some agreement. (Stryd is still lagging behind; whatever was responsible for that.)

This data doesn’t really help think about the COROS POD more, though…

L/R-Balance

One final set of data from the COROS POD and app, which I have otherwise only found in the Racefox Run app (and I’m sure is in a device like the Running Dynamics pod from Garmin), measures left-right balance.

This shows if one foot is favored over the other, one leg does more than the other, which is an indicator of some asymmetry (in strength) or injury.

(Unless if you’re constantly running along a slope. It could also be due to that – which one should perhaps consider in case one’s balance suddenly seems to be off.)

Final Thoughts… for now

All in all, it’s hard to compare, it all requires a closer look over time and with a focus on developing one’s running technique, preferably. But, there was nothing that looked bad and a lot that looked good about the data given by the COROS POD.

The mixture of dynamics and power data, and power data being differentiated into horizontal, vertical, and lateral power, makes the COROS POD well worth considering.

I’ll actually have to ask if it would be compatible with any watch other than a COROS, but it might even be worth a step into the world of COROS by using it with the smartphone app alone, e.g. if you want to start trying out running by power and don’t want to pay up for a Stryd…

2 responses

  1. ChrisTexan Avatar
    ChrisTexan

    I was skimming this at work so may have missed it, but how did you “start” the Stryd recording? Using the app on a phone, or from a watch/device? I only ask, because Stryd indicates in some of their documentation that if you do an “offline sync” from a device post-workout, it will typically have data from before and after the desired time if recording also to a watch, when it syncs up at Stryd’s site. That’s what it looks like here at least to me, the “shift” you are seeing is “pre-start” data recorded, pushing over the rest of the data set.

    I’m still new to Stryd (using about a month now) but have had to learn some of it’s quirks. Basically when the pod starts moving in any fashion, it starts recording, and that’s independent of a device it’s paired to “recording the data from the pod”. Then the Powercenter, if both are synced, tries to match them back up, but per Stryd, the above tends to happen (some pre-post start/stop data is kept).

    If however you are actively recording (start/stop/record) within the app on a phone, no idea, I haven’t done that yet, I use my v800 for recording.

    1. This was the data that is started recording when the watch recording is started… or should have been. The offline sync data, yes, includes even more time, and it was even more wrong. I kinda think it was because of my changing locations between Europe and China.

      Starting the recording in the app starts it at that time, and that actually sounds like a good idea for this!

Leave a Reply

Your email address will not be published. Required fields are marked *