---
title: 'Fitbit Ace - One Month of Scientific Testing!'
source: 'https://youtube.com/watch?v=vvBOlKT1bRI'
video_id: 'vvBOlKT1bRI'
date: 2026-08-14
duration_sec: 1528
---

# Fitbit Ace - One Month of Scientific Testing!

> Source: [Fitbit Ace - One Month of Scientific Testing!](https://youtube.com/watch?v=vvBOlKT1bRI)

## Summary

This video presents a detailed, data-driven review of the Fitbit Ace (referred to as 'Fitbit Air') after one month of scientific testing. The reviewer, a biological data scientist, compares the device's heart rate, heart rate variability, and sleep stage tracking against reference devices like the Polar H10 ECG chest strap and a ZMax EEG headband, testing both wrist and biceps placement. The results show excellent performance, especially when worn on the biceps, making it a top contender in its price range.

### Key Points

- **Introduction to the Fitbit Ace Review** [00:02] — The video is a deep dive into a month of data collected with the Fitbit Ace, focusing on biological data analysis and comparing it to reference devices. The reviewer also briefly discusses the app experience.
- **Heart Rate Tracking During Indoor Cycling** [00:43] — During indoor cycling, the Fitbit Ace on the wrist shows a correlation of 0.99 with the Polar H10 ECG reference. When worn on the biceps, the correlation is perfect at 1.00, showing a cleaner signal.
- **Biceps Band Recommendation** [03:31] — The reviewer recommends wearing the device on the biceps for best performance. A biceps band is not yet widely available, but the reviewer made one himself and expects third-party options to appear soon.
- **Running Performance** [03:43] — For running, the wrist correlation is 0.99, and the biceps is slightly better. Minor details like small heart rate dips are sometimes missed on the wrist but detected on the biceps.
- **Outdoor Cycling Challenges** [04:36] — Outdoor cycling is harder to track on the wrist due to handlebar grip, resulting in a correlation of 0.92. On the biceps, it improves significantly to 0.98.
- **Weightlifting Tracking** [05:32] — Weightlifting is challenging due to wrist tension. The wrist correlation is 0.93, and the biceps is 0.95. The device often misses heart rate peaks on the wrist but detects them better on the biceps.
- **Hiking Results** [06:48] — Hiking on the wrist shows a low correlation of 0.88, but on the biceps it looks almost perfect. This suggests biceps placement is superior for this activity.
- **Overall Performance Percentiles** [07:17] — On the wrist, the Fitbit Ace outperforms 64% of devices for indoor cycling, 57% for outdoor cycling, 80% for running, and 71% for weightlifting. On the biceps, it outperforms 94% for indoor cycling, 83% for outdoor cycling, 94% for running, and 83% for weightlifting.
- **Cardio Heart Rate Scores** [10:03] — The average correlation for cardio exercises (indoor cycling, outdoor cycling, running) gives the Fitbit Ace a score of 97 on the biceps (perfect) and 99 on the wrist (great). This is among the best performance seen for any device.
- **Nighttime Heart Rate Tracking** [11:44] — During the night, heart rate tracking is excellent on both wrist and biceps, with correlations around 0.89. The narrow range (40-70 bpm) makes this a strong result.
- **Heart Rate Variability (HRV) Tracking** [13:41] — HRV tracking is also very accurate, with a few outlying points likely due to low data quality. The device follows the reference well, making it reliable for nightly HRV measurements.
- **Sleep Stage Tracking** [15:15] — Using a ZMax EEG as reference, the Fitbit Ace shows excellent sleep stage agreement: deep sleep ~85%, light sleep ~80%, REM ~73% on the wrist. Biceps results are similar (82%, 75%, 76%).
- **Comparison to Competition** [18:15] — The Fitbit Ace ranks among the top five brands for sleep stage tracking, alongside Whoop, Apple Watch, Oura Ring, and Eight Sleep Pod. It offers the cheapest way to get good sleep tracking.
- **Sleep Scores** [20:41] — The overall sleep score is 79 on the wrist and 78 on the biceps, both very good and similar. The score is the average sensitivity over deep, light, and REM sleep.
- **App and Value Assessment** [21:24] — The app is improving with updates, though text is still large. The free version provides all necessary data. At 99 euros/dollars, it's a bargain, especially compared to older Fitbit models with the same sensor.

### Conclusion

The Fitbit Ace is a highly recommended device for most people, offering excellent heart rate and sleep tracking, especially when worn on the biceps. Its performance rivals top-tier devices at a fraction of the cost, making it a standout value in the wearable market.

## Transcript

shelves the last few months and there's good reason for this. It's a really good tracker at a very affordable price. Today, we're going to take a deep dive into all the data I've collected over the last month using this device, but
also using it on the biceps. Now, for those of you are new to the channel, my scientist specializing in biological data analysis and as always that will be the main focus of this video. Data and
data science to see how good this device actually performs by comparing it to reference devices. Now, at the end we'll also briefly discuss the app and app experience, which luckily has been improving though it's far from perfect.
But as always, I want to start with that data and data science and the first thing I want to look at is heart rate tracking performance during exercise. We're going to look at a bunch of different exercises and I've collected a
everything, but I can show you all the overviews that contain actually all that data in there. And let's start with the tracking performance during indoor cycling, which is supposed to be one of the easiest exercises for a device to
track. And in this case, this is the data for tracking it on the wrist. Now, as a reference device, we'll use the Polar H10 ECG chest strap, which can genuinely record my heart rate very accurately because it actually uses ECG,
which is the most reliable way of tracking your heart rate. And that reference here is along the horizontal axis and the Fitbit Air as worn on the wrist is along the vertical axis and these are all the data points for I
think about six or seven indoor cycling sessions. And if they would perfectly agree, all points should be on or close to the blue line right here. And luckily, they are. Now, there were a few moments of some small deviation, but
this was very minor. But where it does even better is when you wear it on the trackers can get a much cleaner signal on the biceps. You can see those results right here. So now all points are
basically on or super close to the blue line and the correlation value this R value up here is now 1.00. That is correlation cannot be higher than one so 1.00 is basically perfect. And if you look at the results for the
And if you look at the results for the wrist, the score is 0.99. So very close already to that 1.00 but on the biceps it's definitely doing better. Now we can look at some individual spinning sessions to see
from. And here we have one of those spinning sessions. With long the horizontal axis the time and the heart rate along the vertical axis with the Fitbit Air in red and the Polar H10 ECG chest strap in blue green.
And as you can see most of the time for this example session at least, the well though there were some minor moments of deviation. But for the exact same session when worn on the biceps it's basically perfect. For the second
session on the wrist a few days later, again some small moments of deviation in the beginning, but when worn on the biceps versus the wrist it's definitely doing better on the biceps. So very good performance. So here I definitely
recommend wearing it on the biceps. Now I saw there was one available biceps band on AliExpress already. I don't know how good or if it's even real that is. This is one I made myself but I suspect that in a few months it will be really
easily available on Amazon. So if you really want the best performance you have to wait for that biceps band to come out or make one yourself like I did in this video right here. Now running is also something a lot of
us are into right now. So how did it perform for that? Well already quite good for wearing it on the wrist with a correlation of 0.99. And actually on the biceps even though it's definitely better. So if we compare
the two, you can see it is closer to the blue line on the biceps. The correlation was still at least rounded 0.99 as well. So the improvement was a little bit smaller there. If we look at some individual sessions, we can see they
look more or less the same. Though we sometimes these kind of small details are missed on the wrist. So not the full dip in heart rate was detected here, whereas it was detected when worn on the biceps. But this is so minor that here I
have nothing to complain about even when it's worn on the wrist. So that's running, which is still relatively easy for devices to track. But what about biking? This tends to be even harder for devices to track because there tends to
when you're holding on to those handlebars, which can make it harder to track your heart rate on the wrist. And we can indeed see that when worn on the wrist, the correlation isn't quite as high anymore for biking at 0.92.
And it is quite a lot higher when worn on the biceps at 0.98. So here again, if we compare the two, we can see that for the biceps almost all points are close to the blue line and this is not completely true for when
it's worn on the wrist. So in this case wearing it on the biceps had a major example sessions, we can see what's going on. Here we have one for on the wrist and we see some moments with some minor deviation. And the exact same
session on the biceps looks a lot better. Same for this session right beginning. On the biceps right here, there is still some deviation, though not quite as severe. So here again, I
would recommend getting a biceps band. Now to close off, let's take a look at the results for weightlifting. Now weightlifting tends to be quite hard for devices to track because here there's the most tension on on wrist, especially
when doing chest for instance, and pushing up on the bench, there's just a to detect those peaks in your heart And indeed, we can see that when the reference was detecting a very high
heart rate, or for weightlifting high between 120 and 150, let's say, it was to detect my heart rate correctly, and it tended to detect a too low heart
this is definitely a lot better, so in many cases it was detecting the correct heart rate, but it still sometimes detected a slightly too low heart rate, but there's still some deviation right here with a correlation of 0.95 for the
here with a correlation of 0.95 for the biceps and a correlation of 0.93 for the And if we look at one of those individual sessions right here, we can indeed see that on the wrist, the peaks are basically missed, so part of the
peak is detected, but not the full peak. And on the biceps, this just looks a lot better, so the peaks were basically all correctly detected. Now, to close off, I want to look at some hiking results, which weren't so good, weirdly, in my
initial test, and I did another hike, and here the results overall are getting a bit better when we're on the wrist, but still the correlation is quite low at 0.88. Interestingly, for the biceps, this
looks almost perfect. So again, for hiking as well, we see quite some actually summarize all these results together in sort of an overview plot for both the wrist and the biceps, and actually compare to all the devices I've
tested in the past, and that's displayed in this overview right here for the one worn on the wrist with the four main exercises I do, so indoor cycling, outdoor cycling, running, and weightlifting along the horizontal axis,
and that correlation value we're looking at before along the vertical axis, where this red dot and line indicates where the Fitbit Air is compared to all the other devices. So the other devices are the black dots, where we have its
correlation value and its percentile. Now, the percentile basically indicates what percentage of the devices the Fitbit Air is better than. Now, in this case, the Fitbit Air on the wrist is better than 64% of devices that I've
tested before for indoor cycling with a correlation of 0.99. So, I would say it's good enough, but a lot of devices are doing well. For biking outside, it's at the 57th percentile, so around the middle at a correlation of 0.92, so not
great, but still okay. Now, for running, it's doing really well at a correlation of 0.99 at the 80th percentile, so doing better than 80% of devices. And for weight lifting, it's at the 71st percentile at a correlation of
0.93, so not good enough probably, but still better than many devices out there. However, if we now take a look at the Fitbit Air on the biceps, here we see that it's doing super well. So, for indoor cycling, it has correlation of
1.00 at the 94th percentile, so better than 94% of devices I tested. For outdoor cycling, a correlation of For outdoor cycling, a correlation of 0.98 at the 83rd percentile.
For running, a correlation of 1.00 and it's at the 94th percentile. 0.97 at the 83rd [music] percentile. So,
overall, on the biceps, it's definitely doing better than on the wrist. Now, on enough, at least on my body, but it definitely is far from perfect. devices out there. Now, again, this is
differently on other people. I'm actually already testing it on some of going to do a bigger test where I'm at least going to test it on three, maybe four other people, but that's in my final full review of the Fitbit Air.
This is already a lot of data in here, And next I actually want to move to some night time measurements because that's a very important application of a screenless wearable like the Fitbit Air, which tracks you all day and all night
and actually much of the valuable data comes at night. And here there were some But before looking at those results, let's actually give the final scores for according to my standard scoring system. So, if we take a look at my overall
scoring system, calculate the average correlation over the cardio exercises, so indoor cycling, outdoor cycling, and running, then the Fitbit Air, when worn on the biceps, gets a score of 97,
of great. However, if you wear it on your biceps, the heart rate score moves to 99, which means it gets a heart rate rating of perfect. So, this is some of the best performance I've ever seen for any
device if you wear it on your biceps. So, I'm super happy with this. I really think for the price, the Fitbit Air for heart rate tracking during exercise is doing amazing when worn on the biceps. But as I said, during the night many of
the important measurements are taken by the Fitbit Air. So, let's see how well it performed for that both on the biceps again and the wrist. But first, I because running this channel next to my full-time job as a scientist is neither
easy or cheap. And there's really easy ways of supporting. The most direct way of financially supporting is by using my general Amazon affiliate link before you make that your default way as a bookmark of going to Amazon, each time
you click it and buy something, it doesn't cost you any extra. For it gives me a small kickback which supports the channel. You can also early access to some of my videos like this one right here. Or I also have some
specific affiliate links to, for instance, the Oura Ring, the Whoop sleep improvement device, especially during hot days like this. And all of those also give you the best discount possible. Now, thanks so much for
considering, but let's go back to the data. Now, what I did here is look at my heart rate during the night and see how that compares to my heart rate as measured by the Fitbit Air during the night, in this case on the wrist. Again,
the reference is along the horizontal axis and the Fitbit Air along the vertical axis, where again here we're looking at heart rate, so not heart rate variability, even though it says Elite HRV here. That's just the app I'm using.
measurement, and we see a really nice correlation, especially given the narrow range of heart rates we're looking at, so between 40 and 70 bpm. So, looking
really nice, nothing to complain about. And on the biceps, it looks more or less And on the biceps, it looks more or less identical, potentially a tiny bit worse. So, it tends to take slightly higher heart rates right here, but only by a
few bpm. Overall, the correlation is mostly similar, so the R on the wrist is 0.89, or the R squared 0.8. And we have mostly similar values right here on the
biceps. So, potentially it's a tiny bit worse. Now, this could be by chance, that's mostly what I suspect, but it is interesting that the best-fitting line, which is this blue line, has a slightly different slope than the actual values
that we should get. Overall, though, you will still get the this case, there doesn't appear to be a difference between the heart rate on the wrist and the heart rate on the biceps, or potentially even slightly better on
the wrist. We can actually take a look at some example plots, so this is for one night, the 15th of June, for both the wrist and the biceps, with the reference in red in this case. And you can see that both the Fitbit Air on the
biceps and the Fitbit Air on the wrist follow along really well with those values as measured by the reference device. So, nothing to complain about. What is actually much harder to measure is heart rate variability. Now, you
something you want to be as high as possible because it can indicate both better fitness, but also better recovery for a specific night. And right here we have the results for the wrist. So, again, the reference device along the
horizontal axis and the Fitbit Air is worn on the wrist along the vertical axis. Where you mostly see a super nice correlation. There's two outlying points. Now, I didn't filter for data quality, so it could be that these were
really low quality data points. Overall, though, the majority of points is really accurately estimated. And for the biceps, we see a very similar correlation except for those outlying points. But again, these could be the
exception rather than the rule. Overall, super good results, I would say. And if we look at it over one night right here, so again, the 15th of June right here, so again, the 15th of June as an example, we mostly see that Elite
HRV in red and the Fitbit Air on the wrist or biceps in blue green follow along really well. Maybe some of the details are different and maybe on average the Fitbit Air detected a slightly lower HRV. Overall, very
whatever data you get out is mostly reliable. So, I'm very happy with these results. I would trust it as a measurement for my heart rate and heart rate variability during the night as well. So, really
this case again, I don't see a major difference between the wrist and the biceps. It actually makes sense because during the night it tends to be quite easy to get reliable, not so noisy signals. Now, as the next metric, I want
to look at the sleep stages. So, your deep sleep, your light sleep, and your REM sleep and how reliably the Fitbit Air tracks those on both your wrist and your biceps. Now, in this case, I'm using an EEG device as a reference, so
something that can measure my brain waves. I'm using the ZMax EG device, scientific sleep stage tracking, which is sort of a silver standard, not perfect, but good enough to get a general impression of the tracking
performance of the Fitbit Air. Now, on the wrist, we get super nice results. So, on top, we have the sleep stages as tracked by the reference, so deep sleep, light sleep, and REM sleep. And on the vertical axis, we have the
results of the Air, in this case, the one on the wrist. And if they would perfectly agree, all values on the diagonal right here should be 100%. And actually some of the best tracking performance I've ever seen. So, deep
sleep agreement is about 85%, light sleep agreement about 80%, and REM sleep agreement is about 73%. So, just really good performance. the results for the one on my biceps, and they're very similar. So, first of
all, we have deep sleep agreement at about 82%, light sleep agreement at 75%, and REM sleep agreement at about 76%. So, super similar. I think those small deviations are not significant. They're overall just both super good trackers.
comparing it to the competition, but I just wanted to show you two example nights right here, or actually a single example night for the two trackers, on the wrist and on the biceps. So, let's first take a look at deep sleep for this
night, where the deep sleep is marked in purple right here, as measured by the And as you can see, this tends to match very well with the deep sleep as have detected a little bit of extra REM at the end of the night. Otherwise,
the wrist. And on the biceps, it looks basically identical. So, it detected the three main segments. It just missed some of the one on the wrist is detected. But I think this is more or less algorithm
instability and not a difference between the biceps and the wrist. And also REM sleep agreement tends to looks really good. So I sometimes have a super short cycle. Now this wasn't detected by the
two main REM sleep segments were detected and near the end when I was sort of alternating between awake time and REM sleep. The Fitbit Air on the detected mostly awake time with a little bit of REM sleep, but this is still very
good. Now the other Fitbit did detect a bit night, so this is the one on my biceps. And it also detected this very short REM sleep segment here in the beginning of the night. Overall, very good
performance by both of them. Again, I think it's probably random variation and not better performance on either position because it's just such a small Overall, if we compare it to the competition right here, we see it doing
really well. So we have the Fitbit Air on the wrist right here and the one on the biceps right here. Now along the horizontal axis, we have the average stages. And on the vertical axis, we have the
three because we want it to be good at every sleep stage. So the further to the top right the device is, the better is its agreement with the reference device. And we can see that the Fitbit Air is amongst the best devices out there. Very
close to the performance of the Pixel Watch 4 with the new algorithm. Basically identical performance for all those devices from Google. At the moment, I would say there are five brands that have the best sleep stage
tracking out there. Those are Whoop. We have the Apple Watch. The Amazfit Bip Max was actually only based on two nights of data. It will likely end up a little bit lower, so right here. But more on that in that full video. So we
had Whoop, Apple, there's also the Oura Ring. Then we have the Eight Sleep Pod, Now, the Eight Sleep Pod is super expensive, so only get it if you can
really afford it, but it is my lifesaver on hot days like this because the main bed. So, it's not a wearable, it goes around your mattress, and it's also a blanket that is cooled by little water channels going through them. Works super
well, and it actually helps me fall asleep so much better and stay asleep as if it wasn't super warm at night. So, it's really been a lifesaver, but again, you can actually afford it. But if you're interested to try it out for a
to get the best discount possible, use my affiliate link up here or down here. Of course, I also made a full review. Now, those are the first four brands, and the fifth brand is any device from Google. So, that includes both Fitbit
and Google Pixel Watches. You can see both Fitbit Air did really well, and also the Pixel Watch 4. They're the five really top performers And the Fitbit Air or any Fitbit device at the moment with that same sensor is
honestly the cheapest way of getting good sleep stage tracking at the moment. Now, I am also making a video about the Amazfit Helios strap, which does seem to have slightly improved sleep stage tracking now, though not quite at the
level yet of the Fitbit Air, but it's really getting there. That video will drop soon as well. Generally though, very good performance. So, what score do the Fitbit Air on the wrist and the biceps get for sleep stage
tracking? Now again, we can also calculate that overall sleep score for both the Fitbit Air on the biceps and on the wrist, and they're just super similar. When worn on the wrist, it got a score of 79, and when worn on the
a score of 79, and when worn on the biceps, it gets a score of 78. Both of those of you who want to know what the score comes from, it is the average sensitivity over deep sleep, light sleep, and REM sleep.
Generally, very similar, great performance both on the wrist and the biceps for the sleep stages of the Fitbit Ace. So, really happy. So, in actually a real top performer, especially when worn on the biceps. On
the wrist, it's not perfect, but for the price, it's definitely doing really well. And also, the app has been improving. Let me actually show you. Now, Google has been listening to both reviewers and all your comments, and
And I'll actually discuss it in a lot more detail in another video. I think more detail in another video. I think the latest update is 5.02 or 03, which also allows for more customization. But overall, the app does seem to be
slowly improving. So, the text parts are still relatively big, but getting a little bit smaller. But I'm getting used to it a bit. It's still not as clean as I would want it to be, and I still think the text parts are just way too big and
should only pop up if I expand them. But as more customization is becoming possible, hopefully, I can also set that up. Still, it has everything you need. Like, I did a quite long run today. It has that data in there. And also, for
all your health metrics, you can see what is in range, what is out of range. You can look at your heart rate, your weekly cardio. Now, personally, I'm not focusing too much on this weekly cardio load because for me, my workouts are
more on a daily basis than weekly basis. But this is a device that was built more for the masses and less so at the moment, at least, specifically for >> [music] >> Now, with that customization, that might
become better and better over time. But at least in this initial view, that's less of a focus for me right now. And only if I go here, weekly cardio is sort of a smaller part of that. So, I'm quite happy with my experience.
The AI or premium part is something I don't need and I wouldn't pay for it at the moment. I think where the Fitbit Air shines is that free version. You don't get all the guided exercises, I think, and things like that. But all the data
much better than some of the competition. And at 99 euros, that's really a bargain, or 99 dollars. And this is even true if you get an older Fitbit, like a Fitbit Inspire 3 or Fitbit Charge 6. As I'll show in another
video, it does seem that those older Fitbit models, which still have the same sensor as the Fitbit Air, do seem to be performing in a similar range to the better deal. During Black Friday, I think you could get an Inspire 3 for
under 50 bucks. And then with the Google Health app, that's just an extremely good deal. And I think this is where the competition is going to start to suffer. Now, it's not like other devices like the Helios strap or the Whoop strap will
not be the best choice for some other people, but I do think a large group of people that just want a decent app with great performance in terms of raw data will go for the Fitbit Air in the end. I hope that biceps band will be available
soon by third-party manufacturers on Amazon. It's probably the cheapest way by Google releasing a bunch of their specifications for the Fitbit Air, making it easier for third-party manufacturers to create these kinds of
biceps bands. So, overall, definitely for most people, the Fitbit Air is a relatively easy device to recommend. Performance on the wrist, again, is decent. On the biceps, it's great. But at night, you're going to get good
performance for both, at least if the data on my body translates to other people as well. Now, if you do decide to get yourself a Fitbit Air, a new [music] coffee mug, maybe you want to get yourself a Whoop strap or an
Amazfit Helios strap, another device like my favorite sleep improvement these hot days, or anything else on Amazon, even something as small as coffee beans or toilet paper. Want to get the best discount possible in many
cases and at the same time support the channel. I have a bunch of affiliate don't forget to subscribe. Now, given that you watched this video, I think you will like this video on the Amazfit Helios strap or this video on the AC
Helios strap or this video on the AC pod.