Auto detected max heart rate way too high!

Thank you my friend…. This is exactly the informed opinion I was looking for. I fully agree and will follow your advice. Hopefully a 6 month break will re-align my data…

I feel your pain! I use LTHR and all my zones were changed. I changed them back but now I’m constantly checking them to see if they are still right!

You have good reason to question the high max HR the watch gives.

My Epix Gen 2 tells me mine is in the mid to upper 180s, with occasional hits into the 190s. Twenty-six years ago I had a lab-tested max HR of 185, registered during a VO2max test. I was 42 years old at the time. So I find it questionable that my max HR remains unchanged 26 years later.

I concur, been using chest strap reliably for about a decade, been comfortable with the numbers for Max HR and LTHR based 5K performance and years of training RPE data in Training Peaks. I am 54 and have a Max HR of 181. Use a Marq watch. Ran a 5K last week and posted a 197 max and a sustained 188 for half the race. Recovery HR was as usual and I felt great, so I decided I would not have a stroke. Chalked it up to a weak battery so I replaced it,  adjusted the numbers in Garmin and TP to what I feel they probably really were at from RPE and finish time. Did a short base run today in Zone 2 HR and Pace for 30 minutes and Garmin adjusted my Max HR up to 186. I never broke 140 for the whole run. This seems to be a recent thing in my personal experience, but the extra work it will cause to true up your training data and reset zones makes it not worth using the feature. I have turned it off, cant' trust the reliability. Ill watch for the updates automatically turning it back on.

Its definatley an issue that started for me since firmware 15.74. I did a Lab VO2 Max test last Friday at my local Universities Advanced Wellbeing Research Centre. Using a bike, in the 15 minutes of the test I reached a Max of 169.

Still waiting for full results but I turned auto detection off for Max HR and set my Max to 170. Now when I follow a suggested workout, using heart rate as the target, they are achivable & enjoyable again, and also match if I used Pace as the target.

On a plus side its woken me up! I've blindly been following Garmins data, diving head first into the Rabbit Hole, and ignoring my RPE (The Holy Grail in my opinion) From now on I'll be inputting my Max HR, LTHR & FTP from specific practical testing.

I will turn on auto detect at a later date, after a few more firmware updates, to see if its improved, but certiantly won't be relying on it anymore.

N34191 said:

Did a short base run today in Zone 2 HR and Pace for 30 minutes and Garmin adjusted my Max HR up to 186

silentvoyager said:

That was also much higher than the max HR observed during the run,

Yes! The algorithm can propose a new HR Max that is higher than the peak HR of the last activity.

silentvoyager said:

First, the optical HR sensor glitched

Since the new algo will use a range of historical data for its new prediction, and despite the qualification/filtering the algo will do, if you have bad HR data, it might eschew the filtering aand throw off the accuracy of the algo.

Because in reality HR Max will decrease with age, it is safe to assume the look-back period is limited and that the weight of past data and/or past predictions will decrease with time. I am guessing the look back period is around 6-9 months. This would allow HR Max predictions to follow  the yearly natural trend. Complete guess.

CRB said:

I will turn on auto detect at a later date,

I did this for most of 2023 and turned it back on in December or about. I was using the 5k PB peak HR + 5 bpm method during 2023. When I turned the feature back on it predicted an HR Max 1 bpm away from what I have. So I am happy with it now. The watch is proposing a new HR max that varies by +/-1bpm pretty often. I attribute these fluctuations to fatigue and the accuracy of the algo. Note that I got a cycling HR Max after doing a run.

CRB said:

From now on I'll be inputting my Max HR, LTHR & FTP from specific practical testing

I agree. The auto-detection of LTHR, LT Pace and FTP outside of a calibrated test is not very reliable on the watch. The build in tests concur with other platform (TP, WKO5) but the auto-detection is triggered during hard interval work and ends up underestimating all the values, so I always discard it. I keep it on to see what happens when the interval is about 20mn/30mn at maximal effort. 

I do find it crazy that you can't manually reset your Max HR. One piece of rouge data causing so many problems. I will be giving it a good stint before turning it back on, probably the rest of the year, if at all. And certaintly not before I've scoured the forum first.

If it worked properly, the feature would be great!

CRB said:

I do find it crazy that you can't manually reset your Max HR

That would be great. My experience suggests turning it off for a long while and using a manual value in the interim works the same. On previous FW, I was getting crazy high HR Max despite wearing a chest strap, but then I found runs and rides where the chest strap was throwing out of this world values during warm up. I looked up all activities with outlandisk HR Max (you can use Connect to do so), trimmed the ones I could or deleted them altogether.

That didn't solve the issue. THis is why I am coming up with the assumption of a look back period. I am assuming, just like other Training metrics, that correcting past FIT files/activities doesn't impact/correct the Firstbeat metrics.

"If the watch saw it, it won't forget it". Garmin suggests that even recording the same activity on 2 devices can distort the data, even if you discard one of the 2 activites before saving it. I am guessing that now that we have a primary device, the behavior is less of an issue.

Great info there and I guess it can explain what happened to me today, but I think something’s not right. 

7 mile run with 5 x 600m thrown in with a heart strap

Max heart rate during each interval

1 - 167

2 - 169

3 - 173(170 or higher the last 42 seconds)

4 - 173(170 or higher the last 57 seconds)

5 - 173(170 or higher the last 74 seconds)

After the run, New Max HR detected - 169

I had been using a manual setting(173) until about two weeks ago when I turned auto detect back on just to test it and today was the first time it gave me a new value