imagine an Apple Watch but with better battery life and close to half the price tag well that's what the Huawei watch fit 3 promises to be and it's also compatible with both Android and iOS on paper it seems to address many of the common complaints people have about the Apple watch but that's not why many of you have been urging me to review it the bus is all about huawei's claims of improved health and sports tracking So today we're diving into my initial test of the Huawei watchfit 3 and I have to say in terms of design at least this is my favorite Huawei product in a while I think the new square shape is just so much nicer than the elongated shape that the previous Huawei watch fit had but enough about the Statics and specifications let's jump straight into the testing and see how the Smartwatch really performs as always I want to start off with a heart rate tracking performance and here you can see an overview of that performance for one of the easiest exercises for a watch the track indoor cycling which I tested myself for three sessions now I already did more testing but one reason this review took longer than normal because I'm having issues getting to the raw data from the Huawei app so I have to rely on a limited data set for now however I still think the data is good enough for some solid First Impressions now to test the heart rate tracking performance I'll compare the measurements against the polar h10 ECG chest trap which can generally record my heart rate very accurately each dot in this plot is a single heart rate measurement with long the horizontal AIS the value according to the polar H and HG chest strap and on the vertical AIS the value according to the watch fit 3 now the closer the points are to this blue line the better the agreement and the darker black the color the more dots that there are and as you can see this generally looks pretty good actually most points are on or close to the blue line at least interestingly mostly in the lower heart rate range right here do we see a bit more deviation both below the blue line but especially above it also here in the medium heart rate range still this looks really good and the correlation is quite high at 0. 98 so that's this R value up here and this value cannot be higher than one so a correlation of 0. 98 is pretty good but let's take a look at the individual spinning sessions to see why we have a deviation especially here in the lower heart rate range and here we have the first example interval spinning session where we actually see a pretty decent agreement between the watch fit 3 and the ECG chest trap in the higher heart rate range is almost spot on but in the breaks between my segments there's a bit more deviation we can see some right here and also right here there's a tiny bit more deviation but overall honestly this looks really good and for this spinning session it looks even better honestly every time I took a break it caught it almost perfectly so that looks really good and for this final short example we see some deviation for instance right here and a little bit right here but again overall this would be good enough for me next let's compare this performance that of other watches because at least looking at the results so far it seems quite good but another watch might actually perform better by the way for those of you that are new to the channel my name is Rob and I'm a postdoctoral scientist specializing in biological data analysis I should also note that Huawei sent me the watchfit 3 to test but it will go back to them after I'm done testing but okay back to the results and the testing and that overview is displayed right here now a correlation value I was talking about before is the metric I'll use for this which is displayed along the horizontal axis and we want that value to be as close to one as possible and on the vertical axis I order the watches from worst to best so the further to the right and the higher devices the better is its correlation with the reference device and here I marked the huawe watch fit 3 and red and as you can see it's amongst the top performers it has one of the highest correlations out there and most watches are performing more poorly but let's zoom in a bit so we can actually read these labels and here we have that zoomed in view with just the watches of a correlation of 0.
90 or higher so just a better performing watches and even amongst these the Huawei watch fit 3 is one of the better performers it's actually super close to the Huawei watch Fit 2 so I suspect they perform more or less similarly and it's in between the 41 and 46 mm versions of the Huawei watch GT4 so overall very similar in performance to other Huawei devices not better or worse I would say only the Apple watch is still clearly outperforming the Huawei watches they're really a lot better than the rest but I suspect the Huawei watch F 3 is good enough in performance for most people and if you're curious other watches that are close are for instance the pixel Watch 2 which is also a good performer on Android but with poor battery life and also the polar Varity sense a dedicated heart rate monitor so the Hua watch 3 is doing quite well compared to the competition ition overall the watch fit 3 is doing really well for the price at least for indoor cycling it's not quite as good as the Apple watch but it's super close and given the difference in battery life and price that's quite impressive honestly a cycling indoors is a relatively easy exercise for watch your track so let's make things much more difficult and let's see how the huawe watchfit 3 did for cycling outside this is a lot more difficult because of the increased bumpiness and increased tension on my arm let's have a look and as you can see for cycling outside we do have a slightly decreased perform performance compared to cycling indoors but honestly it still looks pretty good the correlation is now 0. 93 so not quite as high as the 0. 98 from before but still pretty good and we again see that in the higher heart rate range the device is really accurate so most points are on or close to the blue line and most deviation is here in the lower heart rate range this is actually opposite to what we see for many other devices where many other devices tend to be better in the lower heart rate range for cycling outside it appears to be the opposite for the Huawei watchfit 3 but let's take a look at the individual rides to see why this could be now as I said in total I did six different rides and this was the longest one by far so over 2 hours and you can see here that in the higher heart rate range it seems almost spot on only in the lower heart rate range do we see a bit more deviation so again this explains what we saw in the overview plot but because this ride took so long it did have a relatively big contribution to the overview plot we were looking at before and I do have to say the second ride I wanted to share was the worst one out of all six rides so here in the beginning it had quite a big deviation from the reference detecting a way too low heart rate later on it's actually a lot better but most rides are luckily a lot better this one still struggled a little bit as you can see right here but also here near the peaks of my heart rate but this R for instance in general looks a lot better though there's a bit of deviation right here and in this final example we see pretty decent agreement this is a lot better than what most watches do so it's not perfect but it might be good enough but to answer that question let's see how it performed compared to other watches out there and here we have a similar overview to before but not for cycling outside and again we want the watches to be as far to the top right as possible and as you can see the Huawei watch fit 3 in red is again amongst the top performers which is really good and again it's really close to the Huawei watch Fit 2 but we need to zoom in to R those labels and here we have that zoomed in view within this case just the watches of a correlation of 0.
7 or higher and as you can see even amongst these better performing watches the Huawei watch fit 3 is doing really well it is doing a tiny bit worse in this overview than the hu watch Fit 2 the we can in no way claim this difference is significant since they only have a limited set of data for the watchfit 3 still overall both the watchfit 2 and watchfit 3 are some of the top performers out there even seemingly a little bit better than some of the other bigger Huawei devices that could have to do with their weight or again it could be random chance but we actually see that the GT3 Pro is doing a bit better so I wouldn't be so sure of that I think again that many Huawei watches are performing more or less similarly the only one again outperforming them is the Apple watch so if you want the best hard rare tracking on the wrist the Apple watch really seems to be best at that but different Huawei watches are close second other good devices include for instance the similarly priced fibit charge 6 the pixel Watch 2 which is more expensive and has a poorer battery life and the coral Space 3 might also be good enough still overall the W watchfit 3 is doing really well so I'm not disappointed so far so also for cycling outside the watch fit 3 is doing really well it seems to be amongst the top performers though I would like to see some more data to be sure that what we're seeing is reproducible in the long r run especially given the relatively easy longer ride that was included next let's take a look at one of the hardest exercises for watch the track however before showing you those results I'm hoping that my diligent testing has earned a subscribe from you and it also really helps me to get access to devices sooner if you leave a like or a comment but enough self-promotion let's take a look at the performance of the huawe watch fit 3 for weightlifting weightlifting is one of those exercises that most watches struggle with because there's so much tension on my arm and the same is true to degree for the Hua watch three we can indeed see that there's more deviation from the Blue Line than before and especially many more points below the blue line right here but still overall compared to other watches this isn't as bad the correlation is a lot lower now at 0. 81 but again compared to other watches it could be a lot worse but here we really need to take a look at the individual weightlifting sessions to see how big the problem actually is and here we have the first weightlifting session I wanted to share with you with again the polar h10 in blue green and the Hua watch fit 3 in red and as you can see only some Peaks were fully captured by the Huawei watchfit 3 some were partially captured and some were not captured at all so it's a bit of a mix some parts of the training were really good whereas others weren't so good and we see something similar for this second weightlifting session it's far from perfect missing many of the peaks in my heart rate and that's also what we see for this final example I wanted to show you right here especially here in the beginning it struggled here in the middle it's a bit better and near the end there it's a bit of a mix but let's compare it to many of the other devices out there and here we have a similar overiew to before but now for weightlifting and in this case the watch fit 3 is in the upper middle class of watches but not amongst the actual top performers and the why watch Fit 2 is and I'm not sure why that is I'm not sure if this is just random chance with a better data collection for the watchfit 2 or if there's actually something better about it so I really need to collect more data the watch fit 3 is close to some other Huawei devices the GT4 in this case also the Huawei band 7 but some other Huawei devices are seemingly doing a bit better so the watchfit 2 we were looking at before but also the GT3 Pro but again I don't know if this is because of the type of exercises I did or what's going on here it could for instance be that I did more lower body exercises for some of these tests and I did actually no lower body exercise with the watchfit 3 and lower body exercises are generally easier for a watchfit track so I have to collect more data and see overall the heart rate tracking performance of the huawe watchfit 3 is quite good in my opinion it's not quite as good as the Apple watch but given the increased battery life it's doing really well actually we also potentially saw some slight differences in performance compared to the previous Huawei watchfit 2 but I don't dare to call those different significant based on the limited data set I collected so far and for those of you that are wondering unfortunately I couldn't do a test for running outside because of an injury still overall I'm quite happy with the performance of the yat fit 3 so far only for weightlifting I cannot recommend it but for the other sports so far it's doing quite well therefore overall I'd give the heart rate tracking performance four to maybe even four and a half out of five stars but that's considering the price but what about the Sleep stage tracking H introduced the new huawe creu sleep 4. 0 algorithm for their sleep tracking but is it any better than the previous generations let's take a look to test the Sleep stage tracking performance I'll compare the Huawei watchfit 3 against the zMAX EG headband which can actually measure my brain waves now this device also has its limitations especially when it comes to detecting awake time but they give us a rough impression of the Sleep stage tracking performance of the watchfit 3 and here are show an overview of the results for a total of four nights now on top are the sleep stages as recorded by the zMAX EG device and the left are the sleep stages as recorded by the watchfit 3 and each column here sums to 100% meaning that we can see what percentage of each of the sleep stages according to the zMAX was predicted as each sleep stage by the watchfit 3 and if they perfectly agree all values on the diagonal should be 100% now first of all we see that only about 36% of what was deep sleep according to the EG device was also detected as deep sleep by the W watch fit 3 so that's not very good more of it was actually predicted as being light sleep instead at about 45% now light sleep agreement is also not that great at about 45% and a lot of the light sleep detected by the EG device was actually predicted as being deep sleep or Ram sleep by the watch fit 3 RAM sleep agreement is similar at about 44% agreement and about the same amount was actually predicted as being light sleep instead at about 40% so overall not looking very good for the watch fit 3 now awake time is something we will definitely not be focusing on because the zMAX s detect a lot of awake moments some of which I think are not really awake moments or just short arousals and those are not interesting for us but just for completeness about 29% of what was awake time according to the EG device was also detected as a wake Time by The W watch fit 3 now looking at the Knights themselves will help us understand that even better and here you can see the first night I wanted to share with you on top we have the sleep stages according to the zMAX EG headb with along the horizontal axis the clock time and my sleep stag is on the vertical axis and on the bottom we have a similar plot but now for the huawe watchfit 3 and in purple right here I highlighted the deep sleep as recorded by the EG device and as you can see there's marginal agreement with the watch fit 3 some of the Deep Sleep has at least some agreement but later on the Deep Sleep detected by the watch fit 3 differs a lot it detects a lot of deep sleep right here later in the night and this is not what a typical night of sleep should look like you have most most deep sleep in the first half of your night and you have most R sleep in the second half of your night and for the watchfit 3 it appears to be opposite in this case so this is not looking that great now the general patterns might look a bit better for this second example night so with more deep sleep in the first half of the night and more R sleep in the second half but the overall agreement with the EG device is still not very good though arguably a bit better but overall this isn't looking very good and also the ram sleep agreement isn't very good so here I marked the ram sleep as detected by the EG device in red and as you can see this doesn't match very well with the watchfit 3 but some of the ramley patterns might slightly match so this first second and fourth R sleep segment show at least some overlap but the duration is different and this third one is completely missed and for this second example night it looks arguably even a bit worse so in the beginning there's basically no overlap later in the night it might be a bit better still generally I wouldn't rely on the ram sleep tracking of the watch fit 3 and again I said we wouldn't focus on awake moments but just to show you that the General pattern shows some agreement the longer awake moments sometimes agree between the EEG device and the watchfit 3 so I wouldn't really trust the EEG device here but if we just look at the longer awake moment so this is the longest one right here this was also detected by the watch fit 3 and looking at this second night we see por agreement but the moment I fell asleep and moment I woke up at least roughly match and for this third example night I wanted to show you we see that the two longest awake moments agree between the watch Fit 2 and the EG device and also the moment I fell asleep and woke up are pretty much much in line so though we cannot really judge the awake moments detected by the watch fit 3 this might actually be the best out of all the sleep stages so the Sleep stage tracking performance of the watch fit 3 just doesn't seem that great both the deep sleep and RAM sleep agreement are quite low and only the awake detection seems potentially decent enough so my first instinct is that you can use the watch fit 3 to detect your total time and bed and maybe your total sleep time but even with the new true sleep 4.
