Garmin Accuracy Again
It seems that just about every one of Garmin’s metrics has a detractor these days.
Altitude Acclimation
Raphael Fais bemoans altitude acclimatization. What does he know about it? Well, what does he know about it other than a PhD in the subject?
VO2max
As of today my VO2max shown in HRV4Training is 8 ml/kg/min lower than what Garmin says. They have the exact same data.
@MarcoAltini wrote the HRV4Training algorithms. Again, other than a PhD in VO2max what does he actually know about the subject?
FWIW: My VO2max data in Apple Health is 0.2 ml/kg/min lower than in HRV4T. IE they are the same if rounded to the nearest whole number.
%age Grade on a Slope with the Edge 1040
We’ve all stood on a slope and realised that it’s definitely not flat. I guess neither you nor I have a PhD in slopes but experience tells us that something isn’t flat.
When I last stopped on a freakingly steep (20%) bumpy, off-road track with the Edge 1040, it showed a zero gradient. Clearly the 9438 viewers of this topic on the Garmin Forums, share my real-world experience of slopes. Actually, they are more complaining about the lag of showing the slope data but I think the underlying problem might be that at least part of the slope algorithm comes from 3D GPS positioning. Hence when you stop it doesn’t work…and this is for the flagship 1040 right?
Resting HRV
OK, I’ve only got a business degree so I’m definitely not qualified to comment on statistical correlations to HRV readings. But I am definitely qualified to read. Try this…
The resting HRV data is wrong for me, which I outline in this post about the Garmin Morning Report – a great feature, shame about the data.
Other User’s Data
Here are some HRV data from the 955 from Marco Altini. In his case, the results look more in line with the other than mine, although not that I think his 955 data set starts to produce results on 1 August. I’ve collected my Garmin hrv data for longer.
Reader-Powered Content
This content is not sponsored. It’s mostly me behind the labour of love which is this site and I appreciate everyone who follows, subscribes or Buys Me A Coffee ❤️ Alternatively please buy the reviewed product from my partners. Thank you! FTC: Affiliate Disclosure: Links pay commission. As an Amazon Associate, I earn from qualifying purchases.
I broadly agree – Garmin metrics seem to be included as a box ticking exercise rather than have any validity that stands up to vigourous analysis. The new overnight HRV analysis is a total joke – even if we don’t discuss the wisdom of using an optical sensor to try to fake up HRV, Garmin’s sleep tracking is quite poor so HRV will be tracked when you are actually awake, meaning an artificially inflated baseline / average. No one seems to care or realise that the pretty HRV graphs are also pretty useless!
From the day I switched from Polar V800 to Garmin Fenix 3HR I realised that Garmin cared more for features than accuracy. I did not really learn my lesson and bought a 5X+. More features. Even worse accuracy. Now I have a 6X Pro Solar. Even more features. Accuracy? Well, quite good if you pair it with an external HRM, Stryd etc..
As for sleep and HRV stuff, I have an Oura Ring 3 and I find the agreement between them to be generally OK for sleep. Regarding overnight HRV, the data seems to indicate common trends, but the absolute numbers generated have quite a gap.
But the real question is – what device can I buy that is actually better in totality. Apple doesn’t even connect to Ant+ so they’re out of the game immediately. That leaves……?????
I think you’ve got it all nicely summed up there.
f7/epix IS better on the gnss front but for instant pace you are always going to need something like stryd. always. chest strap…same! as you know I’ve been banging those exact same drums for years. maybe only you listened to me 🙂
what is better in totality? well, that is a very good question. or at least it seems that way. but then you have to add the provisos for what or for who etc? If I asked you what is the best car, what would your answer be? When you are 21 you laugh at estates/stationwagons but then when you have kids or want to put a bike in the back Volvo suddenly seems way cooler.
Apple is the best smart watch (as you will see next month…it will be notably better with AW8)
Epix 2 is the best sports watch, all things considered. (not best at 500 day long adventures, i know, i know 😉 ) screen trumps super battery. but then if you have thin wrists you might get angry with me!
I hope ? it’s part of the redesign of Training Status 2.0 and Physio TrueUp across different devices. My main problem is with Race predictor. It’s way off since the installation of the first beta. On the other hand, HRV, MaxHR detection, LTHR and training load are at least coherent.
It’s curious what happens with Garmin watches, same model and software gives totally different results on each device.
yes it is.
sure there will be explanations about different amounts of historic data but then someone will pipe up and say “nope, i had the same data on both my devices”.
things like vo2max can be different on two devices whilst one is ‘learning’, the information i post above relates to significant periods of usage when learning should have easily been done over running and biking
In terms of accuracy on heart rate and HRV, Apple watch is the best. The best combo is using Apple watch to take a morning measurement by using the Mindfullness app and by completing heart rate and HRV-value MANUALLY into the HRV app by Altini.
https://mobile.twitter.com/altini_marco/status/1560165024701841409
Garmin 955 first look seems to be ok.
yes i saw that and have spoken with marko about the subject.
sample size n=2 !
Re grade/slope:
Considering that how many people would like to know the accurate instantaneous grade, I am still surprised that Garmin and many of its competitors still use the only-barometer based grade calculation, which is a dead-end street in terms of accuracy.
Garmin could and can finetune the smoothing algorithm, can play the lagging etc., but they do not exploit the value of accelerometers and gyroscopes, which are in their devices.
Without those calculating the instantaneous grade is im-pos-si-ble.
I am no entitled to depict the most accurate formula as regards grade calculation, but wind is wind, which will always mean a bias to barometer independently from the placement of baro holes and the smoothing algorithms, while accelerometers and gyroscope are exempt from any effects if the device is mounted in a stable way.
garmin must already have all (most) of that as inputs to their power calculation.
Let me add that maybe during harsh cornering the grade accuracy would be ruined for a couple of secs, but who the hell could check the grade number during a harsh cornering? 🙂
VO2 Max is most definitely overstated.
Comparing against Suunto and Runalyze calculations, it’s off by about 10 points. Fairly unusable in its current state.
If they have they are doing something wrong.
My Newton+ from Velocomp does show the instantaneous grade, while none of my Garmin or Bryton devices can show the same. Time lags are everywhere.