How does Garmin compute VO2Max?

Does this help?

https://www.firstbeat.com/en/consumer-feature/vo2max-fitness-level/

No, they say it is proprietary.  It seems to be more of a magic number than anything else. 

I took a few days off from running and my VO2Max went down 8 points (39 to 31).  ???  I ran a 5k time trial, namely 1 mile warmup and then 3 miles at the highest heart rate I've hit in several years.  It went down another point.  Next day I ran an easy 8 miles and my VO2Max went up a point.  I think part of it is a random number generator.  It doesn't seem to at all match reality.

What you're missing is that not only heart rate is taken into account, but pace as well. If you ran a hard 5 km, but the algorithm assessed that your pace was not high enough for the given heart rate then this could easily result in a drop of VO2max. The direct reason for worse performance could be high ambient temperature, inaccurate heart rate data, or you just being out of shape.

Andn easy run can result in a higher VO2max if the algorithm sees a lower than expected heart rate.

tmk2 said:

. The direct reason for worse performance could be high ambient temperature, inaccurate heart rate data, or you just being out of shape.

I think something else may be going on here. The OP only took a couple of days off and their VO2 Max dropped 9 points. I've taken a couple of weeks off due to injury and had mine drop 1 point. Even with temperature change, inaccurate HR or a change in fitness, 9 points after a few days off and a 5K doesn't seem right at all, wouldn't you agree? 

Forget about it, plain and simple. Short of getting a lab test, there are methods to derive VO2Max form a few factors, however, cheap consumer grade devices just don't deliver on the data front. Garbage in garbage out.

There was a recent review posted over at the 945 boards, that is pretty in depth: http://www.muscleoxygentraining.com/2019/05/firstbeat-vo2-estimation-valid-or-voodoo.html

Those abrupt changes usually stem from significant changes in HRmax.

There are many suggested VO2Max tests on the internet and none of them seem to line up very well.  I see no reason to think that the one by Firstbeat is any better than any of the others.  It seems to be just another "magic number" used to sell gear, and should probably just be ignored.

I've been tested on treadmills wearing the oxygen mask thing by "professionals" and one can't even believe what they come up with.  I had a second one done a couple of months after having had a first one done and the technician tried to tell me that one of the reasons for the difference I saw was that my max heart rate had increased.  Sigh....  Of course that is impossible. 

Commercial interests don't pay attention to the training of their staff, calibration of their gear, or how often filters are changed, etc.  I wouldn't believe any results unless they were done in a university lab setting, as part of some kind of study or some such.

If you don't have any faith in what Garmin or any other "commercial professional" test tells you, what's the point of your question?  If no test matters outside of a universery lab test, what difference does it make knowing how Garmin calculates VO2 Max? 

No way!  Not after a "couple of days off".  I don't care how significant the HR changes is, 9 points in this time period doesn't make sense. It takes weeks to establish a VO2 Max. One "bad" run after a couple days off won't drop it 9 points if it's being measured correctly. Take a couple of days off and then intentionally tank your next run. I guarantee your VO2 Max won't drop 2 points let alone 9. Something else has happened to drop it like this and it's not correct.