@splash …This (straight line on workouts with no intervals) is a plausible explanation, and as I think back, it seems to be consistent with what I’ve experienced.
Where did you come up with this hypothesis?
There is a whitepaper somewhere which explains how Garmin pick the VO2max value (Firstbeats??). I’m a bit hazy on the details now, I read it a while (years?) ago.
1 Like
Mine and my performance scores are always all over the place. I should just turn it off. Nothing like starting a race and seeing a -6 performance score 10 min in 
5 Likes
Same here, warming up seems to throw the performance reading right off
I’ve done some of my best power numbers after a terrible initial performance score 
2 Likes
FWIW I had my vo2 max measured in a lab 5x in my 20s, over a 3 month span. It varied 4 ml/kg-min. It’s very possible for vo2 max to change day to day or week to week. Getting within a few percent of your actual would be hard to do just because it will in fact vary a bit due to slight changes in physiology.
Garmin also estimates my vo2 max to be 60 (at 77 kg), which would be 64 @ 72 kg. 15 years ago my best vo2 max test in a lab was 68 @ 72 kg. So given that I’m older, seeing a small drop makes sense, and I would bet it’s pretty close to where I would test now as my power numbers are very similar.
Week to week, yes, at least if there are changes in physical activity. A couple of examples:
(In the latter study, VO2max increased by 9% after just 10 d of training.)
Day-to-day variability (which is about 2%) wouldn’t be considered a true change, however, just measurement error (which is primarily biological, not technological, in origin).
3 Likes
FWIW, there are multiple scientific studies evaluating the accuracy of Firstbeat’s algorithm (licensed by Garmin). It’s pretty good… generally w/in 5% or so, IIRC.
However, that isn’t necessarily all that surprising, since you can easily estimate cycling VO2max (VO2peak) to within 10% just from power alone (and to within a couple of percent with fancier analysis).
5 Likes
My point was that even lab measurements can vary a few percent, so for Garmin’s estimate to get within even 5% of an actual lab tested number that is going to have it’s own variability, it’s pretty good. It also appears to align with my own lab tested results, so for people wondering if it’s accurate, it at least appears to be accurate for me. I saw nobody else state that their own lab tested numbers were close to Garmins estimate, so I felt like sharing my own results may be helpful.
1 Like
would be nice if it’s really accurate… 
“indoor season” seems to be doing good things for my vo2max, more endurance is not good for the formula…
(from <60 to 68, not sure where the green/running vo2 max is coming from… I never recorded a run ever with the garmin…)
1 Like
Increase at the same weight?
I.e due to decreased weight or improved fitness
no change in weight.
(fitness seems to be the same to, but maybe a shift from endurance to intensity)
I see the exact same thing happen to my Garmin readings when I shift to harder intensity, specifically a lot of VO2. And vice versa, off-season and base season it drops off. And this is all without much change in FTP.
Last winter post VO2 block peak was ~65 for me. Low of ~55 earlier this fall. Back around 59 now but haven’t done much or any VO2 or Threshold in a while.
My weight will fluctuate ~5lbs throughout the year, but I’m usually to lazy to update it in Garmin.
i’ve followed my garmin vo2 estimates with some amusement lately, some days it gets to 68 and sometimes goes to 65, currently it’s 66
1 Like
One little caution…last year I had my Garmin VO2 max climbing, and climbing, and of course I was happy to believe it was because of my great training and hard work. But I started to get suspicious when I saw my VO2 max workouts were being listed as low aerobic….
Turns out that there is a garmin setting for automatically detect max heart rate that I had toggled on, and it had read an anomalous rate like 40 beats higher than my actual max. So it then decided when I was when I was doing an all out effort that I actually had lots of heart rate capacity left and it showed my VO2 max on a continual increase.
2 Likes
Just wanted to add to this thread an observation I made about the garmin VO2 max estimate. I recently got WKO5 and realized I haven’t done any max sprint efforts, so I went out and did a max 20 second sprint which immediately lowered my garmin VO2 max number by 10 points (my anaerobic capacity was definitely underestimated). My garmin and WKO5 VO2 numbers are now within a point of each other. I imagine garmin is using your anaerobic capacity combined with max aerobic efforts to calculate the number, so if you haven’t done a max anaerobic effort it’s probably not accurate.
Thanks!
The main problem with the Garmin algorithm is that thermal regulation is not controlled for. Just because of the temperature effect on HR, my Vo2max estimate swings 10-12 points.
Steve Neal said on a forum sometime ago that you’ve got to feed these type of algorithms the right type of information otherwise they are just junk
2 Likes
It’s too bad Garmin gives you zero instructions for their black box metrics.
I found this from First Beat whom Garmin own and use for their VO2max stuff. I’ve not read it yet though. I suspect you just substitute power for speed if you’ve got a power meter on your bike.
https://www.firstbeat.com/en/aerobic-fitness-level-vo₂max-estimation-firstbeat-white-paper-2/
I was interested to see how the First Beat estimation varied with error margin in value used for HRMax. Garmin has a HRMax value I manually entered 5yrs ago and haven’t changed since. That value was based on a TR ramp test (I haven’t done one since) and whilst in the right ball park unlikely to be spot on. If, say, it was a few beats under back then then it might be a few beats over now. At least I now have a better idea of how many pinches of salt to add when tracking my supposed V02max.
I could of course follow some deeply unpleasant protocol to try and measure my actual HRMax but I’m not convinced the pain wouod be worthwhile. My VO2mx numbers (according to Garmin) have been remarkably consistent over the least few years - varying by around 6 points and only changing gradually and in-line with my periodic changes in performance on the bike. On the one hand this might speak to the reliability of the model but on the other it doesn’t really tell me anything I didn’t already know about my fitness trend (and bearing in mind caveats in the accuracy of the actual number before and especially after I add my pinches of salt).
I still track it though - hard to completely ignore data.
1 Like