COROS Pod 2 In-Depth Review: A Missed Opportunity

COROS has announced a new ‘Pod’, now becoming the second edition of such Pod units. This pod is a blend between a glorified footpod for smoother and more instant running pace, as well as a running efficiency metrics pod that gathers added data like ground contact time and vertical oscillation. The previous COROS Pod was attached just to your running shorts, but this new pod can be placed on your shoe or running shorts. And perhaps one day they’ll allow alternative placements for alternative sports. But today it’s all about running.

The main goal of the POD 2 is simply faster instant pace, and more stable instant pace. That’s because while GPS continues to get better, there are some cases where instant pace from a footpod is faster in big pace shifts (like intervals). Of course, unsaid is there are also scenarios where footpod-based pace isn’t as awesome (like very steep terrain that doesn’t have a consistent stride, or sometimes changes in surfaces – like running on sand). Nonetheless, the aim for the COROS Pod 2 is to provide better pacing in tough GPS signal environments, such as city running, or perhaps some dense trees. Additionally, they say it’s also targeted for indoor running, which otherwise would depend on wrist-based pace (usually quite good, but again, there’s always caveats).

With that, let’s talk about what’s new.

What’s new:

The new Pod 2 is a bit more expensive than the Pod 1, coming in at $99 versus the existing $69. There are both new features, but some removed features. In the case of the feature removals (running power), they won’t matter much since on COROS watches running power still comes from the wrist anyway.

– Adds footpod based running pace & distance
– Adds footpod based running cadence
– Adds footpod based stride length
– Adds pod-based temperature
– Adds pod-based altitude
– Adds pod-based elevation gain/loss
– Unrelatedly adds COROS Effort Pace
– Added magnetic compass

Further, it maintains the following data types from the POD V1, when worn on the waist

– Keeps Left/Right Balance
– Keeps Ground Contact time
– Keeps Stride Height
– Keeps Stride Ratio

Here’s a quick specs sheet overview with all the geekiness included:

So in short, it can be used in both footpod mode (on your running shoe), as well as on your running shorts/waist (where it does running efficiency metrics mode). The price is $99 (the previous POD1 was $69). It includes two shoe clips, and one waist clip.

Using the Pod:

This section will be silly short, because frankly, once it’s paired up, there’s nothing additional to see as a footpod – it just provides more stable pacing data. When worn on the waist, you will see additional data for running efficiency metrics (the same as with Pod 1). Further, in both cases it provides data in a blended manner for stats like altitude and temperature. In fact, now’s a good time to look at this chart to figure out which data source comes from which thing. Note that the pod does not provide running power, that comes from the wrist. Also note, this chart is probably confusing AF at first glance. That’s actually OK, it’s more of a reference thing:

So, after tapping/whacking/shacking/whatevering your pod, the thing will wake-up and the little light will blink. At this point, it’s ready for pairing.

You’ll go into the accessories menu on your COROS watch (it’s only compatible with COROS watches, except the Pace 1, which isn’t compatible at all). Note that you’ll need to update your watch firmware if you haven’t done so since the date of this post shown above (Oct 27th, 2022), in order to pair with the pod.

Once paired, it’ll show as connected, the pairing happens via Bluetooth. There are two options, one to change which orientation it’s in (footpod or waist, but I just left it on auto), as well as a orientation calibration option (for the internal compass)

Now, we’ll go ahead and start a run. You’ll notice a new run screen while waiting for GPS that shows the Pod 2 status on the left side:

After that, we can press start and begin running. You might at this point also notice the pace data field has a slightly different name now, to indicate the Pod is connected and providing that data source. This is a nice touch, even it probably unnecessary. If the pod is not connected, it won’t show this added text.

In addition, if you’ve got other data fields like temperature or altitude added/shown, the data is used including the pod data, as noted above in that fancy chart. At this point, everything else is the same. There’s no other differences here from a running footpod standpoint.

Instead, let’s switch it up and run with it on our running shorts. In this configuration we’ll trade the pace data, and instead get the running efficiency metrics data: Ground Contact Time, Stride Height, Stride Ratio, and L/R balance. We still maintain temperature/elevation/altitude data, but lose pod-based distance, pace, cadence, stride length, and L/R balance. Further, we also lose the so-called ‘GPS track enhancement’.

In this configuration, there’s nothing for you to do, it automatically figure out which position it’s in, and provides data streams accordingly. Now you’ll see that data on the watch, as well as in the app afterwards:

Finally, if looking at a treadmill scenario, you’d want to use it as a footpod (obviously). However, if you want to use it on a platform like Zwift Running, you can’t pair directly to Zwift (as the Pod doesn’t transmit data via any open standards). Instead, you’ll broadcast with your COROS watch to Zwift, and the COROS watch will use the inputs from the Pod and re-transmit them to Zwift. This works fine of course, but kinda undercuts the entire point of a faster responding pace device, since now we introduce another roughly 1-2 seconds of lag in there for the re-transmission.

Finally, for charging, you simply snap it into its little charging coffin, and plug in a USB-C cable. I found that while it charged just fine connected to my Mac, for some odd reason, it flickers when connected to my Lenovo Windows laptop (any port – also providing endless connected/disconnected messages). I’ve had no other device do this, so I’m not sure what’s going on there, but I have tried multiple USB cables without luck.

It has a small charging light on the top to let you know what the current status is. Battery life is claimed at 28 hours of running usage, and 50 days of standby usage.

Effort Pace – Let’s Chat:

Effectively unrelated to this entire Pod situation is a new ‘term’ they’re also announcing today, which is ‘Effort Pace’. COROS says the following:

“Due to the limitations surrounding Running Power, COROS has made the decision to keep Wrist Based Running Power as a metric for its users, but not develop this metric any further at this point. Rather, we are introducing a new metric, Effort Pace, that COROS can personalize to athletes, and continue to develop well into the future. Effort Pace will surpass Running Power as the metric becomes customized to each individual regardless of environmental factors.
Athletes can still pair with Stryd for a full Running Power experience if preferred.”

So what is Effort Pace? Well, Effort Pace is simply the previously named ‘Adjusted Pace’, now renamed. As COROS themselves notes:

“Adjusted Pace will now be named Effort Pace and will be further developed to take into account many environmental factors other than just uphill and downhill gradient.”

To which you might ask, what is “Adjusted Pace”? No problem, from their support site, “Adjusted Pace” is just what everyone else in the world calls “Grade Adjusted Pace” (GAP).

But wait, I thought Effort Pace was different? No, not yet. COROS says that at an unannounced time in the future, Effort Pace will include new data components that make it different than the current Effort Pace, and in turn, will do different things:

“Currently Effort Pace has the same data inputs as GAP, so while the algorithms are not identical they are incredibly similar measurements at this time. In the future we plan to include: temperature, humidity, altitude, and more.  Before end of the year: Effort Pace will be individualized to distinguish runner’s personal strength or weakness running flat, uphill or downhill.”

So, to complete this circle: Effort Pace was Adjusted Pace, which is just another word for Grade Adjusted Pace. The only difference is COROS is putting a crapton of marketing behind this Effort Pace concept, without actually defining what the heck it is. I could rename it OreoPace or BeerPace and it’d carry as much usefulness of a term (except OreoPace would likely be trademarked by Mondelēz International, and thus have to get renamed again to BlackAndWhiteCookiePace).

If you’ve been around the block here long enough, you know that nothing gets me more annoyed than companies “re-imagining” existing sports tech terms without any benefit to either consumers, or at least the sports tech industry. Further, COROS seems to think the industry is going to follow them on this term, but there’s exactly 0% chance of that happening. First of all, all their competitors are…well… ‘competitors’ – and have no desire to see COROS succeed. And secondly, most of them are already using the industry agreed upon term for this metric. Thirdly, COROS isn’t doing themselves any favors here by introducing a pod that’s compatible with (also) exactly 0% of the pre-existing industry standards.

The new COROS Pod 2 does not support the Bluetooth Smart footpod profile used by every app and almost every watch out there (even COROS’s own watches). Nor does the Pod 2 support the ANT+ footpod profile also used by many apps and any watch that supports ANT+. The BLE profile has been around about a decade, and the ANT+ profile has been around probably 15 years or so now. Not to mention the COROS Pod 2 doesn’t support the already established ANT+ Running Dynamics profile either.

Look – I’ve got no problems with new metrics as long as said metrics can be defined, provide value to consumers, or provide standardization to the industry, or heck, just do something at all. At this point, Effort Pace doesn’t do anything except confusingly rename a very functional term.

A Data Dive:

Now, there’s two basic buckets of data types the pod produces:

A) Pacing related bits (e.g. as a footpod)
B) Running efficiency related bits (e.g. ground contact time, etc..)

Practically speaking the pacing portion is where most of the (marketing) focus from COROS is on the new Pod 2, and technically speaking, comparing things like VO/GCT/etc is really tricky. Plus, it’s been done before.

So I really wanted to dig into a couple of broad claims that COROS made on the Pod 2, to see if they held water. These claims are roughly:

A) That GPS pacing is too slow to respond
B) That the Pod 2 is far faster to respond

And no better way to do that than line-up their competitors, and themselves on a not-so-simple 2KM circuit and do some tests in different configurations with different devices to see how things handle. This circuit included two passes under a three-bridge 10-12 late highway/train overpass, as well as passing some tall buildings (and trees). It was perfect for a scenario of ‘lost GPS’ signal. In total, I was looking at pacing sources from the following units:

1) COROS Pod 2
2) COROS Vertix 2 (no external sensors)
3) Garmin Forerunner 955 (no external sensors)
4) Stryd as a footpod/pacing data
5) Garmin HRM-PRO Plus (for pacing data)

These were attached to a variety of watches, and all the data pulled together on the DCR Analyzer. Now, because I’d rather just show you data, let’s get straight into three scenarios that are easy to demonstrate:

Scenario 1: Zero to Steady-State Pace

In this scenario, I started from a stand-still, and then simply started running at my long-run easy pace (roughly 4:40/km or 7:30/mi). It takes about 1 second for me to stabilize on this pace once I push off. If there’s one thing I’m good at, it’s running incredibly steady without looking at a watch. Said differently, if a device can’t get stable pace results on me, it’s got issues. I did each of these tests about half a dozen times, and the below figures are essentially the average of those attempts, with outliers noted. The time noted is time until it reached my actual pace/stabilization

COROS Pod 2: 6 seconds
COROS Vertix 2 (no sensors): 8-10 seconds
Garmin Forerunner 955 (no sensors): 6-7 seconds
Garmin HRM-PRO Plus strap: 6 seconds
Stryd V2: 5-6 seconds

Now, at this point you want to see my demonstrate this on a graph, and that’s actually surprisingly tricky. It’s essentially a ‘tree falls in the forest’ problem. That’s because a graph of pace won’t show pace when standing (duh, it’s zero), but consequentially, won’t show pace when the sensor is lagging/thinking from a standstill. Thus, it’s impossible to show what’s zero versus what’s “thinking zero”. Think of it like Jeopardy. There’s no difference in the answer time between a contestant that hasn’t pressed the button, versus one that presses the button but hasn’t given an answer. Both answers are still zero. Finally, adding even more complexity to it, is that while GPS watches do sync their time of day via satellite & Bluetooth (which is used for the chart alignment), the reality is that’s only accurate within about 1-2 seconds in most cases – which isn’t quite enough for our purposes. So I’ve re-aligned the charts based on what I actually recorded while running, so data alignment matches to observed reality. Got all that? Good.

Here’s the first chart, showing Zero to Steady State, with the COROS Pod:

And then showing the Vertix 2 without the COROS Pod:

As you can see, the TLDR here is that the COROS Pod/FR955/HRM-PRO consistently landed at stabilization in about 6 seconds. Sometimes the FR955 was 7 seconds, sometimes actually quicker at 5 seconds. But usually 6 seconds. Stryd tended to be about 1 second faster, at around 5 seconds. The Vertix 2 was consistently always the slowest (by at lot) at 8-10 seconds.

Scenario 2: Already Running to Interval Sprint:

In this scenario, I was at either an easy recovery pace or a steady-state long-run pace, and then accelerated up to an interval pace. In this case I did intervals at roughly 3:30/km or 5:36/mi.

COROS Pod 2: 3 seconds
COROS Vertix 2 (no sensors): 6-14 seconds (or worse)
Garmin Forerunner 955 (no sensors): 4 seconds
Garmin HRM-PRO Plus strap: 6-8 seconds
Stryd V2: 3 seconds (but then overshooting)

In this case, we can see the benefit of the footpods by both Stryd and COROS, reducing that time to about 3 seconds. However, the Forerunner 955 was no slouch, a mere 1 second behind. Interestingly, I found that in the acceleration test, the HRM-PRO was actually a bit slower – around 6-8 seconds in total (or about 3-5 seconds behind). The Vertix 2 without a pod? In the best case scenario it was 6 seconds, but it was often upwards of 10-14+ seconds. It seemed to depend whether it was a slower ramp (which it did worse), or a shaper transition (it did better).

Here’s the first chart, showing steady-state to interval, with the COROS Pod:

And then showing the Vertix 2 without the COROS Pod. I’ve made markers showing the Stryd overshoot, stabilization of most, and then the COROS Vertix 2 overshoot staying high on pace for another 10 seconds after I slowed down. Heck, so did Stryd.

This is very hard to show on charts due to the scale. Instead, it’s easier just to see the lag on the screen in real-life. But the long and the short of it, which you can even see above, is the Vertix 2 takes forever to change pace without the pod.

Scenario 3: Stopping:

This was my favorite scenario. I simply stopped running abruptly, then recorded how long it took each watch to zero-out. I did this multiple times, and each time the results were +/- 1 second, so the graphs will very slightly for each iteration. But the general gist of it was:

COROS Pod 2: 4-5 seconds
COROS Vertix 2 (no sensors): 6-8 seconds
Garmin Forerunner 955 (no sensors): 5-6 seconds
Garmin HRM-PRO Plus strap: 5-6 seconds
Stryd V2: 4 seconds (but sometimes stayed weirdly above zero)

This one was pretty straight forward, and basically matched what we see in earlier tests. The two footpods were very similar, though I will note that while Stryd showed a zero value in pace on the screen, in reality the watch was writing very low non-zero pace numbers to the file. I don’t know if this is a Stryd issue, or a Garmin issue, because I don’t know precisely what Stryd was transmitting. From a practical standpoint while running it didn’t matter, since the display showed 0:00 (which basically happens once you fall below ~20 mins/kilometer), meaning it tripped the lower pace threshold.

Anyways, as with the other tests, the Vertix 2 by itself was the slowest of the bunch.

Scenario 4: Pace Stability:

Next, let’s just look at some of the steadier state sections where I’m running along. Both without a tunnel, and with a tunnel. As you can see here, without a tunnel (mostly open sky), it’s perfectly fine. I thought it was interesting to see the Stryd differences in a few spots, especially since those spots didn’t match what the others were doing.

And then a tunnel circuit without the pod, the Vertix 2 is also stable. The Stryd seems a bit more frisky this time. That could be real, or not.

As I noted above, I tend to run very stable, and as a result, watches tend to be pretty stable with my numbers in this area. Maybe I just need to stand by the side of the running trail with a sign that says “Help Wanted: Inconsistent Runner Needed”.

Scenario 6: GPS Track Enhancement:

One of the items that’s included in the Pod 2 is GPS track enhancement. Meaning, it’s likely using gyros inside the pod to provide a secondary source of data for the COROS watch to figure out which direction you’re running. This can then act as a sounding board in bad GPS data environments (e.g. cities, bridges, tunnels, etc…). So again, let’s look at that. Here’s two back to back images of running under this umpteen lane bridge where we lose GPS. We can see how well it performs there with and without the pod.

First the tunnels, comparing the Pod 2 (orange) with non-pod. Seems mostly a wash, though, the Vertix 2 actually did slightly better exiting the tunnels than the Pod 2 did, as it went into the water a bit.

 

And then on a straight away section with a few buildings, the Pod 2 track did considerably worse, being across the street and in some buildings.Positions of all watches were identical between the two circuits and exact path run would have been within 1 meter at any point for each circuit.

In fact, in looking at all the data on these circuit sets, in almost every location the Pod 2 assisted GPS did worse. I’m not sure it’s something I’d want connected at this point for GPS assistance.

Scenario 5: Treadmill Stability:

In this test, I went through three different treadmill paces. Basically just presets on the treadmill. One can argue all day on whether or not a treadmill is accurate, though, this particular one has largely been pretty close to accurate. Notably though, this specific Forerunner 955 was just factory reset yesterday, and as such hasn’t had much time to accumulate pace-learning for each pace level. Thus, I wouldn’t overthink that specific line for this chart. Here’s how each of the three levels performed:

Steady-State Run – 12.1KPH (Peloton Tread):

COROS Pod 2: 11.5KPH
Stryd V2: 11.3KPH
Garmin HRM-PRO Plus: 11.6KPH
Forerunner 955 Wrist: 13.5KPH

Sprint Pace – 20.1KPH (Peloton Tread):

COROS Pod 2: 17.8KPH
Stryd V2: 18.3KPH
Garmin HRM-PRO Plus: 20.0KPH
Forerunner 955 Wrist: 21.7KPH

Walking Pace – 4.0KPH (Peloton Tread)

COROS Pod 2: 4.0KPH
Stryd V2: 4.0KPH
Garmin HRM-PRO Plus: 3.2KPH
Forerunner 955 Wrist: 3.6KPJ

For all these values, I took it once the treadmill belt speed showed that it had stabilized. Since this treadmill doesn’t quite ramp up as fast as some, there was no meaningful difference between the ramp rates of the sensors and the treadmill itself. There was of course pace differences.

Accuracy Summary:

So in summary, these results show a few interesting – if not somewhat surprising things. Much of the reasoning from COROS to buy a pod is that it’s more responsive. And yes, that’s true – but it’s actually most impactfully true for their own watch, and less so for the primary competitor. In the case of the equally-priced Garmin Forerunner 955, it either was *at worst* 1 second behind the pod, but often either equal – or even occasionally a second ahead. Compared to Stryd, the COROS Pod was usually about 1 second behind, or equal in some cases.

What about pace stability? Well, again, using equally modern watches, the FR955 pace stability, even in the ‘tunnels’ was perfectly stable – matching that of the footpods. Because again, modern watches don’t just use GPS for pacing. They use wrist-based measurements to blend with that satellite data, something Suunto for example used to call FusedSpeed.

Undoubtedly, at this point someone will comment “One time at band camp my XYZ watch can’t show pace right, burn them at the stake!”. And that may be true. It might be true that that particular watch might be less optimal. Or, your particular running stability might be less optimal. Or, some blend of the two. But if we look at the direction of GPS technology in 2022 about to go 2023, with good multiband implementations, we’re seeing astounding pace stability and accuracy. That combined with existing wrist-based counterbalancing, leaves little reason for newer watch owners to buy a footpod. It may however leave good reason for older watch owners to buy a footpod.

Wrap-Up:

Look, footpods are hardly new. In fact, they are arguably about the oldest thing in the running sports tech book. Before there were GPS watches, there were footpods connected to non-GPS watches. It was literally the first way people started tracking pace/distance on watches, many years before watches could do so-called ‘wrist based pace/distance’. The accuracy of said footpods has definitely improved over the years, though, so has GPS technology.

GPS watches no longer use just GPS for pacing. They now use satellite signals from multiple constellations, blended together with both accelerometer and gyro wrist-based watch data to form a far more accurate real-time picture of what you’re doing. When one data stream is impacted, it silently transitions to another. This isn’t new either, it’s been doing this for years. It’s just that those sensor inputs have gotten dramatically better in the last 1-2 years. Combine that with watches now supporting things like running track mode (one of the hardest things for GPS running watches to tackle), there’s even less of a need for footpods.

Still, there are many older watches out there that simply aren’t that accurate, or might need some loving assistance for better data. And, the COROS Pod 2 could have helped there. Same goes for indoor running on treadmills. But unfortunately, the Pod 2 is only compatible with COROS watches (except the Pace 1). As such, you can’t pair it with an older Garmin/Suunto/Polar watch that might need that assistance. Nor can you pair it directly with Zwift Running, or any other indoor treadmill app. Thus in a nutshell, the people that need this pod the most aren’t the ones that can get it. Though, based on the test results above, one might argue the people that need it the most are the COROS Vertix 2 users…

Look – had COROS made this thing dual ANT+/BLE compatible following the industry standards for running footpods (and maybe even existing running efficiency standards/protocols, or the existing temperature sensor protocols), I’d say it’s a great thing. But at this point, it’s essentially only there to overcome COROS’s lesser instant pace responsiveness. All at a time with every bit of running sports technology is moving away from pods for these sorts of tasks.

With that, thanks for reading!