Last week Samsung touted a pretty bold claim in the middle of their keynote for their newest smartwatch: They could determine your cycling FTP (functional threshold power, the max power you can maintain for roughly an hour). To do this, you’d need a cycling power meter, 4 minutes, and…well…that’s it.
Thus, I decided to give this a whirl. Turns out, the fine print is pretty important here.
Doing The Test:
I’m going to save you many hours of troubleshooting, searching, and otherwise trying to figure out how on earth to do this test. Despite this being one of a handful of new features highlighted for the new watch in their keynote, there’s virtually no documentation on it. And by ‘virtually’, I mean ‘absolutely none at all’. The core of my challenge was figuring out first how to pair a power meter to the watch, and then second, how to trigger an FTP test.
But I went back and carefully watched the keynote over and over again, and read the ultra-small text at the bottom, along with some key tidbits in the keynote. While the screens behind the man showed a 4-minute timer, in reality, he mentioned we don’t pair our power meter to our watch, but rather, a Samsung Galaxy phone.
And sure enough, after poking around in the Samsung Health app settings, you can do exactly that (to seemingly every Bluetooth power meter/trainer I had nearby), via Samsung Health > Bike > … in upper corner > Accessories > Scan:
It seemed as though I was on the right track, and with the power meter paired to my phone, the next bit of trickiness, though, was figuring out how to trigger an FTP test. It seemed in the video that it was more of a post-ride scenario. Not really sure what to do, I went ahead and started a ride on my phone, another on my watch, and then did a hard 4-minute effort. Again, there’s absolutely *ZERO* guidance on this. So I was just making stuff up on the fly, trying to figure it out.
And frankly, if I can’t figure it out (someone who literally tests this stuff for a living), everyone else is hosed. In any event, I got everything on the phone app ready, and pressed start on the watch.
[Fun Things I Learned #1: You cannot re-search for new power meters or any Bluetooth fitness accessories without cellular coverage/signal. Literally, the page won’t enumerate. Ensure you do this before you leave cell service.]
Turns out, pressing start on the watch was the wrong thing to do. That’s because you instead press start on the phone, and then it takes over the watch (ending the watch session). Ok, fine – phone wins. I then tapped to close/lock the phone to stick in my back pocket. That was also the wrong thing to do, it immediately paused the session. Thus, I resumed the session from the watch (handy). Seriously folks, I can’t make up this stuff.
Also, you’ll note below it shows ‘POW’ in the upper left corner, indicating power meter connected, along with “HRM”, which is the Samsung Galaxy Ultra watch. It’s unclear to me if the GPS here is coming from the app or the watch in this scenario.
Now, for realz, off I went.
With zero instructions, I simply went balls to the wall down the rowing basin bike path. Unfortunately, being the first of two summer days in The Netherlands (yes, I know, it’s July 18th), there was a bunch of human and goose furniture at the start. Nonetheless, I quickly got up to speed and held a reasonably constant power of about 350-360w, save occasional goose dodging. After 4 minutes and 30 seconds, I stopped, got the phone out, and stopped and saved the session.
Astoundingly, it immediately gave me an FTP value. Said value was 272w:
I say ‘astoundingly’, because in that crazy-ass fine print up above, it technically said I needed one night’s sleep with the phone/watch combo, as well as some other unicorn dust (Samsung Health AI). I had neither. I hadn’t had bedtime relations with this watch yet, though, I had slept with its sisters in the past. Maybe that did the trick.
The second part of ‘astoundingly’, is that this wasn’t actually that far off of a value. My ‘real’ FTP floats between 290 and 310w, depending on the season/training/etc… Right now, Xert puts it at 308w (though, it would show 300w the next day). TrainerRoad at 302w, and Garmin also at 308w. I’ve generally found Xert and TrainerRoad to be within 10w of each other, and have proven to be historically pretty darn accurate to actual FTP tests I do when I’m super bored.
Yet, my day today was pretty busy:
1) Two hour photoshoot on the bike (not really a hard ride, just a lot of stop/go/hard/stop/go/hard/etc in the sun…)
2) A 30-minute interval run
3) Another 10-15 minute power meter calibration/bedding test (unrelated to the Samsung one).
4) Finally, this FTP ride.
In other words, I’d have guessed it would predict low, and indeed – it did.
So, let’s go sleep with the watch, and give it another whirl tomorrow – this time on super clean uninterrupted roads, with a good night’s sleep.
[Fast Forward to The Next Day]
Doing the Test Again:
The next day, with some good sleep behind me, I was ready to tackle it again. Albeit, still a bit fatigued from a busy week – but hey, the internet can’t wait. Once again I went out, and the Samsung app failed to connect to the power meter. I futzed with it for a while, probably 30 minutes all in, before it finally found it again. At this point, out I went:
This time I had less geese, less people, and a bit more power to my name. My route wasn’t quite perfect for this, as I was ending up slightly short on super-clean distance each time. Nonetheless, I kept it at 4 minutes and 30 seconds, including the time spent starting/stopping the app.
At this point, it bumped up my FTP estimate from 272w to 279w. Unquestionably I left a little bit in the tank, given the imperfect course (and my ability to properly pace a 4-minute all-out effort). Still, approximately ~280w isn’t that far removed from the lower 300w estimate from Xert, especially accounting for a week of fatigue (which would take Xert longer to account for).
However, and most critically, Samsung actually says you don’t need to do an all-out effort. They say a mid to high effort is apparently good enough. Of course, you’ll only see this text *AFTER* you’ve completed a power-meter-equipped ride, and then tapped on an FTP estimate. Further, it says to just keep riding with it, to improve estimates (like most other units). So I decided to go out for a more reasonable 40-minute ride, with perhaps one Strava segment tossed in for funsies.
Upon finishing my ride…I found that the power meter connectivity had apparently failed about two minutes into the ride and never resumed. Because you can’t see the power meter data on the watch itself, you don’t know until after the ride. Thus, this entire ride was a waste of time (save two windmill sightings).
But wait, there’s more. You see, even in cases where the power meter data does record, Samsung doesn’t send that data to Strava or seemingly any other platform. I had linked up my rides to Strava, and none of them show power meter sensor data.
[Note: This is different than how Strava handles the Apple Watch power meter data. In the Samsung case, Samsung is delivering the completed file to Strava as-is. Strava just processes it like any other data. Whereas for the Apple Watch, Strava themselves accesses the Apple Health data on your phone, and then decides not to pull the power meter data Apple added last summer…because…I don’t know why. TLDR: Samsung screws up their own power data, whereas Strava screws up Apple’s. In the case of Apple/Strava, you can use HealthFit to get that data in there properly.]
And while the Samsung Health App does show the power data, it seems highly smoothed (or outright blocky), so it’s a bit challenging to make use of that – at least in its current form.
Wrap-Up:
I’m genuinely surprised here on so many fronts. First, I’m actually kinda surprised that Samsung seems to get the final number in the right ballpark. Its most recent estimate (280w), seems like it’s probably aligned to a more fatigue-driven week I’ve had than my normal FTP that floats in the 290w-310w range. So, I’ll actually give it a pass on that.
However – and this is a massive ‘however’ – every other aspect of the Samsung cycling power meter implementation is totally dumpster fire status. Let’s just summarize:
A) There’s zero documentation on how to do any of this, in any way.
B) You don’t actually pair to the watch, you pair to your Samsung Health phone app
C) Samsung won’t work with Power meters if you lose cellular coverage
D) You can’t see the power meter data/metrics on your watch
E) The phone app will likely fail at some point to record your power meter data
F) The power meter data isn’t sent to any 3rd parties, including Strava
G) It does however, seemingly get a ballpark-ish correct FTP via the app
I guess I continue to be perplexed by Samsung’s health/fitness efforts. Someone, and some team, spent a lot of time on designing this feature. Heck, you can even export out a post-workout image from Samsung Health with your FTP overlaid on it (serious). But you can’t export out the power meter data, or see it on your watch. So close, yet so far away.
With that, stay tuned for my full in-depth review of the Samsung Galaxy Watch Ultra’s optical HR sensor, GPS accuracy, and more. Same goes for the Galaxy Watch 7, and Galaxy Ring. Hopefully, it’ll be better.
Thanks for reading!
FOUND THIS POST USEFUL? SUPPORT THE SITE!
Hopefully, you found this post useful. The website is really a labor of love, so please consider becoming a DC RAINMAKER Supporter. This gets you an ad-free experience, and access to our (mostly) bi-monthly behind-the-scenes video series of “Shed Talkin’”.
Support DCRainMaker - Shop on Amazon
Otherwise, perhaps consider using the below link if shopping on Amazon. As an Amazon Associate, I earn from qualifying purchases. It doesn’t cost you anything extra, but your purchases help support this website a lot. It could simply be buying toilet paper, or this pizza oven we use and love.
Hi Ray,
I’m curious, does the sensor pairing explicitly require cellular coverage or will any data connection (Wi-Fi for example) work just as well?
WiFi works as well (and how I initially did it, before switching to a SIM-enabled phone). And while technically you can use existing power meter connections without cellular/WiFI, that only works if it decides to remember the connection and let you connect to the PM.
So practically speaking, you’ll want cellular/wifi if doing any power stuff.
It seems that FTP is not a number that can be exactly quantified, more just a “ballpark range”, so I wonder how one would ever know if the watch was “correct” or not?
“With that, stay tuned for my full in-depth review of the Samsung Galaxy Watch Ultra’s optical HR sensor, GPS accuracy, and more. Same goes for the Galaxy Watch 7,…”
YES, PLEEEEEEEEEEEEEEEEEEASE! Dying out here, delaying my order until this comes out!
In terms of correctness, I think it’s really going to be ballpark-ish. Basically: “Is this number believable?” – Short of doing a full ~1hr FTP test, and even that, is iffy at best unless you know you gave it 100% and had 100% to give. (Hence, the entire somewhat silliness focus on FTP)
Hang tight on the other pieces, so far, at least for Ultra, things are actually looking pretty good…
OK, thanks! I’m really leaning towards the 7 over the Ultra, but they have the same HR and GPS sensors, if I understand correctly.
How about the ring?
I am not entirely convinced of the utility vs. a Garmin watch but who knows…
I was set to pickup the ring today, but alas, it wasn’t there. So…yeah, I wait.
(I had ordered Ultra & Watch 7 units that arrived a few hours before the Samsung media units arrived, so I didn’t need the media units. However, the Samsung Netherlands media/press office couldn’t get Ring’s since they don’t sell in the Netherlands, so I ordered one to France where I am now…but alas, it wasn’t at the store today, and they don’t really know why.)
Either way, I have other backup options, so…yeah, I wait.
I worked with Samsung directly in a previous job and the takeaway I had at the time is that they do a great job of recruiting talent (they are like the “ivy league” of South Korea in the sense of working there will gain you entry to a completely new societal status), and they produce a lot of interesting things, but they have a very poor ability to tie anything together into a cohesive user experience.
So you end up with stuff like this which is impressive, probably deep tech, but barely works for very basic reasons like the sensor drops out and it’s impossible to find out set up for ordinary mortals. I’ve seen it play out like this so many times that I just instinctively stay away from anything Samsung that isn’t a single use device like an appliance.
What I took from Ray’s piece here is that a great many very smart people designed this, and that precisely none of them are included in what they think the target market is, or what the UX needs to be for it to be a viable offering. It’s absolutely screaming “how do you do, fellow cyclists?”
*or have any understanding of what the UX needs to be
watched your YT video, thank you for sharing.
I have the latest Garmin Cadence and Speed 2 sensors and they work nicely with my Venu 3. Only after seeing your video and reading this article did I realize that I could possibly use the SH app to find these as accessories – thank you!
The app actually found both sensors. For indoor cycling, it is unfortunately simply ignoring the speed sensor and only reporting the cadence. For outdoor cycling it shows both. One annoying effect is that it uses some interval to pull the data for both sensors, ending up in a sawtooth graph (attached). Interestingly, it ignores the 0s and gives the correct average. The cadence was working simultaneously on both the SH app and the Venu 3. I thought that caused the intervals, but I confirmed that it still shows intervals when I remove the sensors from the Venu 3.
I am wondering if you have more past experience with this setup you could share. I keep looking.
Very much looking forward to your continued testing of the Ultra.
Scathing. Love it.
The algorithms coming from far east (Samsung,Huawei,Coros etc.) are fully inacceptable.
The have good hardware and that was.
I had thought I could not pair a chest strap for heart rate measurement with a Samsung watch. However, in this you didn’t pair the power meter to the watch but the phone. Could I get the same or at least similar effect as pairing a chest strap to a Garmin watch by pairing a chest strap with my phone and tracking an activity with a Samsung watch? I prefer biking, treadmill, and rowing for most exercise activities – could these be done with a chest strap reasonably well paired to the phone and using a Samsung watch. I am not sure its time to drop my Venu 3, its only a year old, but I was just wondering if I have been overlooking how to do this with a Samsung watch. My chest strap has bluetooth connectivity BTW.
I have the Venu 3 too. I wouldn’t give that up for its battery life. Looking at the Samsung Health it lists these heart rate devices, e.g. the Polar chest belt. I don’t know if these are all chest belts. I have not tried any of these for the Ultra because I do not own any. The Garmin sensors I tried (see above) were not listed under the app but still connected. Maybe any Polar belt will work, not sure.
Did it measure and get the FTP into a believable ballpark, or is FTP for all users a Gauss curve and taking into account age, gender, perhaps training load, any value near average will be ‘in the ballpark’?
I mean, it’s not far off of Garmin’s values, but then again, few people with a power meter will be below 200 W and few will be above 400 W, so taking the average and throwing in some factors will never be 100 W wrong?
How do you know it’s not simply taking a percentage of your best 4 minute effort?
The 4-minute “FTP” test is giving 7-minute abs energy.
link to youtu.be
I’m sorry Ray but you really dropped the ball on evaluating the correctness of this “FTP test”.
There are lots of tests that one can do to get an estimate of FTP. The original Coggan FTP test required at least a 40 minute max effort. The idea being that such a long effort is going to be dominated by aerobic fitness, which FTP is a measure of. Then someone made a 20 minute test as a less fatiguing alternative and used a correlation factor (0.95) to calculate FTP. This test is pretty accurate, assuming you follow the actual test protocol that includes an anaerobic effort first to make sure you’re testing your aerobic capacity in the 20 min effort. But still people wanted shorter, easier tests. So then came the 8 min test which isn’t accurate for many people because your anaerobic capacity has too large an influence to the result.
So Samsung comes along and says they can do it in 4 minutes, which is absurd, because your 4 minute power is heavily dependent on anaerobic ability. But again FTP is a measure of aerobic fitness. So you’re basically running a 400m all out and trying to estimate your 5k pace. Does that make any sense? No. Sure, you could take a bell curve of people and come up with some “typical” ratio of those times, but it will be wildly off for anyone that’s trained to be good at either distance.
Your sample size of one test is useless to prove if it’s reasonable or not. And a 20W error is actually quite large. So I would judge it a fail even for your n=1.
So what’s a better alternative? Simply upload your normal riding power data to Strava or intervals.icu. Assuming you’ve been doing some hard efforts on your rides, they will estimate your FTP fairly accurately, automatically.
“I’m sorry Ray but you really dropped the ball on evaluating the correctness of this “FTP test”.”
I get the feeling you didn’t actually read the post, at all. I literally talked about every caveat you listed, in my post.
That design is not aesthetically pleasing. They should have kept the body round as well.
I don’t suppose there’s any way that would work with an older Galaxy Watch4, but a Galaxy Fold5 phone?
– I’ll try it when I get home, but I don’t know if it’s dependent on the watch being able to pair with a power meter (Quarq in my case) or the phone