Performance condition as a proportion of VO2max

I think a few things are happening here (just educated guesses):

- when the whole number value of your VO2 Max changes, it’s impossible to know by what percentage it changed, as the internal value is stored with 2 decimal places and rounded for display (*). e.g. if the visible value changed from 54 to 55, the internal value could’ve gone from 54.49 to 54.50 (+0.018%) , or 54.01 to 54.99 (+1.8%), just to name a couple of possibilities.

(* The training status glance’s VO2 Max page does graph unrounded values on the chart, but the numerical value is still rounded. So you can kind of eyeball the chart to guess at the unrounded values and to see how they changed, or even take a screenshot and count pixels, or you could use runalyze to see the unrounded value to 2 decimal places.)

- I think performance condition refers to the difference between your current VO2 Max (for the activity) vs your average VO2 Max which is displayed in Connect (**)

- (**) I think the VO2 Max that the watch displays and saves to FIT files is the average VO2 Max (a rolling average over some period of time / number of activities.) I don’t think Garmin ever stores or displays the true per-activity VO2 Max (in isolation)

So let’s say your average VO2 Max is 54.49, and you have a performance condition of +5 after an activity. My conjecture is that the per-activity VO2 Max would be (roughly) 54.49 * 1.05 = 57.2145. But your new VO2 Max (displayed in Connect and stored in the FIT file) would *not* be 57.21, but a new rolling average would be calculated based on the existing 54.49 average and the new per-activity 57.21 value, and the new average would be stored in the FIT file and displayed in Connect. I don’t even think the supposed per-activity VO2 Max value (57.21) would even be stored anywhere. It seems that what’s stored in FIT files (and displayed in Connect) is always a rolling average (meant to represent your current fitness after an activity, as opposed to exactly how well you did during any given activity.)

All of this is based on using https://www.runalyze.com to view the VO2 Max stored in FIT files to two decimal places (“VO2 Max by file”). I noticed that this value never changes too drastically, even if the VO2 Max calculated by runalyze itself (based on hr and pace, similar to Garmin’s algorithm) differs wildly between activities (e.g. if you run a shakeout followed by a road race, your runalyze VO2 Max for those two activities might differ by 10 points, but the Garmin VO2 Max will never move by that much between two activities.

And anecdotally, my VO2 Max by file number always increases by a much lower percentage than my performance condition would indicate. e.g. If I had a +5 performance condition after a run, my FIT file / Connect VO2 Max would probably go up by something like .5% - 1%. (This supports my educated guesses above.)

I recognize your described behaviour, so you are probably not widely off the truth. Although I experience the whole number changing not at .50, but before that, at .30. Seems strange, but last run , my whole number went from 53 to 54, and the RUnalyze VO2max by file went from 52.98 to 52.35.

I agree with what Flowstate wrote, but I'd add that it's not only performance condition at the end of run that affects VO2max change. The watch analyses the performance condition graph for the whole activity and fits that to expected behaviour based on VO2max. That is then used to calculate the VO2max change. (Sorry, can't give you a link, but this was from some Garmin/FirstBeat document.) So, for a short easy run it's expected that performance condition at the end is still high. And even if it's negative at the end of a hard run, the VO2max estimate may still raise.

DaKo said:

Although I experience the whole number changing not at .50, but before that, at .30. Seems strange, but last run , my whole number went from 53 to 54, and the RUnalyze VO2max by file went from 52.98 to 52.35.

Yeah I've noticed something similar. I'm not sure if the rounding algorithm changed recently or not. I could've sworn in the past that the whole number changed at x.50.

When I first noticed this, I wanted to report it as a bug, but it seems pointless. (We're not "supposed" to know or care about the unrounded number, right. I guess the bug could still be reported without even mentioning runalyze, on the basis of the unrounded vo2max graph in the performance status glance.)

Bitti said:

So, for a short easy run it's expected that performance condition at the end is still high. And even if it's negative at the end of a hard run, the VO2max estimate may still raise.

Yeah this is a good point. Personally I've found the best indicators that an activity will increase VO2 Max is that the performance condition starts out high (or at 0) and is still fairly high (+4 or +5) in the middle of a hard workout.