Versa 4's heart rate monitiring is bad

[Rjpnak]

Sounds like Fitbit made a change after the Versa 2 and they all have problems.  It's sad because the Fitbit app is on point.   

I had a  Versa 2 for years then the button began to stick.   Heart rate would go janky above 160 bpm but I could deal with it.   Everything else was spot on.  Why dont they go back to Versa 2's HR monitoring?   

Versa 4's HR is so laggy and inaccurate.   I don't care about smart watch features and notifications.   I only want a fitness tracker.   To keep the current coarse with the new products having the same problems, is silly and stubborn.  The reviews on the Sense 2 are similar.   Why??!!  Someone made the call and they're sticking to it!

[Guy_]

Hi @Rjpnak - Starting with the Sense and Versa 3 there was a known issue with the new technology heart rate sensor which can produce inaccurate heart rate during exercise. Unfortunately the Sense 2 and Versa 4 didn't revert back to the good Versa 2 [and prior models] sensor.

You can still get the Versa 2 model which has some features not found on the Sense 2 and Versa 4.

[Biceboca]

I looked at the Sense 2 discussion which was nice because there were as lots of graphed data comparing to a chest-band HR monitor, which is expected to be very accurate. I was actually impressed with the data in most (though not all) cases. People are upset if there is not minute-by-minute high precision, and perhaps some athletes need that info for interval training or something. But the average beats per minute for workouts was usually nearly the same as the chest-band monitor, I mean sometimes like one or two beats off, which for my use means the watches are darn good. That's what I care about, an average over a stretch of exercise, and I don't mind a few high or low spikes. The newer tech might have fixed some issue, otherwise it does not make sense to use it over the old one if the old one looks better when compared to a chest-band. Maybe the new sensors are more gentle on skin, or more sensitive on darker skin, or more energy-efficient. There must be some upside. And from the data, my impression is that most of the time HR monitoring is still very good.

[Triletics]

@Biceboca Similar average from wrong data is nothing more than a coincidence. I don't quite know what's the use for an average HR (maybe except estimating final calories). What the average HR tells you (this is a serious question, I really don't know)?

[Biceboca]

It is not really a coincidence. It is an accurate average from the data the watch collects. It is not a random number that by coincidence matches the chest-band tracker exactly or nearly so. And this means that a few spikes does not significantly affect the average. The watch vibrates when you hit particular zones of heart rate (the number of vibrations depends on the zone) which I think is a nice feature. My guess is, the zone detection is from an average from a minute or two, so that spikes won't affect it. Same for minutes in the various zones shown on the fitbit app on your phone (I'm talking about the particular HR zones are customized to an individual). What you really care about is hitting various intensities for a certain number of minutes. That is, knowing the intensity of your workout, and the number of minutes at that intensity. For me this is useful info mostly so that I know how different types of workouts or activities compare. For example, I was surprised that walking and mowing the lawn give much better cardio workout for me than riding my bike. (I commute by bike and ride a lot, but because of this insight, I now choose to walk instead of ride sometimes.) Probably there are additional uses for averages but this is what I could quickly come up with. I don't think most people would benefit from knowing minute-by-minute precise heart rate. Actually, I think the tracker measures every 5 seconds, and some of those readings are off, but over a couple of minutes it is probably accurate. And for practical purposes, for most people that's good enough.

The thing I actually care the most about is resting rate, and also % below resting heart rate during sleep. And fitbit is very accurate for that. It gets spikes during high intensity exercise which is what bothers people, but even at high intensity, the average rate is accurate, which I think is what matters the most. That way you know whether you got a good cardio workout.

Most likely there is more info on this topic in the health-related fitbit discussion boards (I did not look there, but if you go to Community, and scroll down away from the product help boards, there are other topics including health discussion -- maybe the Get Moving board has more info on this topic). For a pro athlete, it might make sense to get a chest band for highest precision, but my guess is, for the vast majority of people, the watches do a good enough job.

[Triletics]

@Biceboca cycling is in fact much better cardio but Fitbit isn't able to obtain accurate HR, not even close. I'm not talking a few beats off or some spikes but totally incoherent results comparing to the chest strap. Cycling is probably the worst case scenario (thousands rides, I track it with chest strap and Fitbit, not even one remotely accurate and yes, averages lower than if I walked when looking at Fitbit data). You may do yourself a favour going back onto bike and trust less your Fitbit 🙂 In such case (which I observed a lot with raw data not eyeballing charts or watch screen) it falls into a coincidence. I never recorded a cycling ride using Sense/Sense 2 (same sensor as Versa 4) which had acceptable HR (not accurate, just acceptable). This happens often when it comes to other activities, too. Running, cycling, weight lifting, calisthenics, rowing, climbing (swimming can be kind of compared with some trickery but that I won't count as officially Fitbit doesn't support HR tracking when swimming). All of that I do with chest strap as well as wearing watches and Fitbit doesn't have to be beat-perfect but at least close in terms of not only numbers but trends. I admit I don't track walking (although for that, when tested Sense/Sense 2 showed consistently slightly inflated HR but I don't care about that) or mowing a lawn. I track only actual sports and in that field Fitbit isn't doing even a good job. I say, the worst case scenario is cycling, and best case scenario is running at steady pace (although sometimes, when watch locks onto my cadence I have consistent 180 to 200bpm running at easy pace).

When tracking activity, HR is measured every second and that is used by users during those activities. If you do zone 2 run or ride you want to be in that zone most of time. Averaging wouldn't do because averge may come up in the correct zone yet the time in that zone may be minimun. Statistically, human and dog have on average 3 legs. This illustrates the problem. Median might be a lot better but even better using normalized values. Averages are useless without knowing standard deviation and when analysing Fitbit data, the stddev is affected by inaccurate readings (spikes and drops will affect stddev). If users wouldn't use HR sensor during exercise they wouldn't be complaining, yet they do.

What is a good cardio workout? Workout serves a purpose. I know I can do 1000 jumping jacks, elevate HR but this isn't a "good cardio workout". It all depends on purpose. Word "cardio" is like a lockpick tool for many people. Average HR doesn't tell me whether I did "good cardio". When I run at base effort (easy, steady) for 3/4 of run and then I add a few strides in the end then it will inflate average only slightly but anaerobic effect of strides isn't negligible either. Average (from real data of my last Thursday run) of 146bpm from 3/4 of 136bpm and last bit of 176bpm. Average only 10bpm more yet the effect is vastly different from a similar run for the same average HR but entirely at steady pace (for example, there would be no muscle fatigue and no anaerobic effect). My Fitbit decide to drop my HR under 100bpm during strides and never recovered until the end so data comes from chest strap. Was that a good cardio? Maybe, I don't know. I know only that the average is no use to me as it blurs the actual effect of workout and tells me nothing. I know it may be enough when mowing the lawn but users buy those watches with hope of using them during actual exercising and it's not about an absolute accuracy but getting at least acceptable results out of it.

Just curious, what is your reference when you say that Versa 4 is accurate? Have you done own comparisions with chest straps? I do understand that the level accuracy may match your needs and that's absolutely fine but there are many users who need a little bit more accuracy for various reasons 🙂