VO2 Completely Inaccurate?

i00 said:

My VO2 Max is poor ... @ 33:09 for 5K...

Would you provide a little more information please? What is your VO2max? What is the relevance of the 33:09 for 5K?

i00 said:

I commonly do ~1hour runs that average over 15km/h!

You regularly run 15 km in one hour? At what intensity? Based on this statement it would not be unreasonable to think your 5km time would be well below 20mins.

VO2 Max is 35.  I find 15KM/h quite hard to maintain over 1 hour.

On my run yesterday Aerobic was 3.0 and Anaerobic is 0.3.. load of 74.  Is that what you mean?

33:09 is the time estimate under VO2 Max for me doing a 5KM run.

Hi. Here is the right article for you

discover.garmin.com/.../

Did you add correct weight, age, height? It's strange...

http://www.muscleoxygentraining.com/2019/05/firstbeat-vo2-estimation-valid-or-voodoo.html

http://www.muscleoxygentraining.com/2019/09/why-is-garmin-firstbeat-vo2-estimate.html

https://assets.firstbeat.com/firstbeat/uploads/2017/06/white_paper_VO2max_30.6.2017.pdf

The answer is that the Firstbeat algorithm does not have the level of precision that you probably expect. In a best case scenario in their own white paper it is +-5%. 

• If you do not have the correct max HR, it goes out to 7-9%
• If you are not using an EXG chest strap, they don’t even describe the error bars
• If you are running in heat or hills or rough terrain or altitude there is additional error that they don’t describe
• It doesn’t evaluate load across activity types at all and treats those as if they were separate people entirely
• The algorithm prefers threshold workouts and high intensity intervals of at least a few minutes length and punishes recovery, zone 2 efforts, and multiple workouts per day
• The algorithm requires accurate current pace as an input from GPS and that is problematic because it isn’t very accurate with the possible exception of multi-band in favorable conditions. 

In my experience, it especially seems to work poorly with medium to high volume polarized endurance event training. 

In practice I think the error bars should be around +-10% and for some people it is worse. Consider what +-10% means:  [45 = 50 = 55] or [54 = 60 = 66] or [63 = 70 = 77]. You can only make very rough comparisons between given measurements on a particular athlete or between athletes. I think it is unfortunate that it gives values at the 10s place. It would be better if the display emphasized the uncertainty more and a smeared out range of error bars. It’s also overly optimistic to use the oHR to drive this feature at all, IMO.

In my opinion the VO2max score is just for entertainment purposes and not to be taken very seriously. If it bothers you the. you can remove the glance from your watch and from GCM and “pause training” and you don’t have to look at it unless you go looking for it. (They will still calculate the metric.)

By proxy, other metrics like Performance Condition are also dubious. 

Yep that is all correct

I don't think that the heart rate is as accurate as my Fenix 6 ... the 7X seems to over read my actual heart rate in comparison ... but only sometimes.

That said, you think that doing regular runs FASTER than its estimate of what I am capable of, would recalibrate its estimate!

My 2p worth...

I agree almost whole heartedly with Brian above

Heart rate exactly matches my medically issued ECG and HRV monitor. It is within 5 bpm which is what can be expected in delays and time frames of calculation. That is with a chest strap or optical HR as long as you are not on a cardio option on the watch when it is up to 50% off with the optical HR. HRV is WAY off, way way off.

I have had my VO2 tested medically and found to be within 2 ml/kg/min of the Garmin estimate (and you must remember it is an estimate as it is using age, height, weight, breathing rate, active level and goodness knows what else to calculate it, rather than fitting a mask and actually tracking volume of oxygen breathed in vs volume of oxygen breathed out etc). This was surprising that it was so close.

Race prediction is laughable! Apparently I can do a sub 3 hour marathon and sub 17 min 5k. No matter how I try, I can't even hope to reach those times.

With the GPS being much more accurate on the 7X, the VO2 is calculated better than with previous watches (and I have done tests with 5X, 6X, 7X, GPSMAP 66sr and Ardusimple Professional Survey system with base station on a series of bike rides in and out of trees, open paths, roads and canals to compare; the 7x was the only Garmin device to closely follow the professional setup).

The built-in oHR from the wrist is perfectly fine for all day monitoring and walking. It is also ok if you just want to roughly track overall volume of intensity similar to what whoop and oura can do. It just isn’t good enough for heart rate training in my opinion. You need to ideally use an ECG HRM for that as the gold standard. 

I don’t think the Firstbeat algorithm uses HRV.It is necessarily a simplistic algorithm that can be calculated on extremely low powered devices. I think it derives a value from weight, percentage of max HR, pace and maybe a table of correction reference values by age. It doesn’t consider environmental conditions, terrain, altitude, or other activities. 

I have found the First Beat algorithm to be very accurate. Remember, GIGO (Garbage In - Garbage Out).

Now, how do I know that my VO2 Max estimate it accurate? I don't. However, I know from experience that the Race Time Predictor is very accurate in real world conditions. Of course, add in heat, humidity, sun, wind, hills, mental attitude, and other things, and your speed versus heart rate correlation might throw things off.

I think such a sinical outlook on this tool is unwarranted, given the data shown in the whitepaper, and also that it was 5 years ago. Surely the algorithm continues to be optimized since then.