Why does Garmin VO2max numbers differ so much from Runalyze VO2max?

Well, the VO2max is simply estimated differently by Garmin/Firstbeat and Runalyze. :) First, both require you to have entered a correct max hr value (Runalyze doesn't seem to complain if you don't, but will of course give inaccurate values). Second, I expect both of them to base the VO2max estimate on some time window in the past, not just your most recent activity (which is good). Runalyze docs say the VO2max estimate is based on the last 30 days (this can be configured). I haven't found out any exact info on Garmin's method, but considering how VO2max reaches the correct value gradually when you start using a Garmin watch, there's probably some past info used in Garmin's method as well.

According to Firstbeat's white paper (https://assets.firstbeat.com/firstbeat/uploads/2017/06/white_paper_VO2max_30.6.2017.pdf), the Garmin method takes heart rate variability into account as well when estimating VO2max. Runalyze doesn't seem to, so I personally trust Garmin's VO2max estimate more than Runalyze's.

This seems to be pretty common, I've always noticed that Garmin has my VO2Max 4 to 5 points higher than Runalyze.

Garmin has me as super fit and in the top 1% for my age - which is as senior runner, would be nice if true. Looking at my predicted race times vs my best times, I think the Runalyze metric is more accurate for me. I treat Garmin VO2Max as a trend to track, rather than an absolute value.

PS: I used to have an Amazfit Stratos which also used to have VO2Max via Firstbeat and Garmin and Amazfit both gave me the same  VO2Max score, so it's not just Garmin, rather I think Firstbeat over-estimates a bit. Or at least it does for me.

PS2: If you use the Elevate plugin for Strava, this can give you a Running Index (Polar's version of VO2Max), it's handy to track that vs. other VO2Max estimates.

5135968 said:

Глядя на мои прогнозируемые времена в гонках по сравнению с моими лучшими результатами, я думаю, что метрика Runalyze для меня более точна.

For me, too, the Runalyze distance predictions on VO2 are much more accurate than Garmin's

So the Garmin VO2max is measured over a longer period of time whereas Runalyzers VO2max is on a per-run basis.

Bitti Garmin bought Firstbeat last year so both FB and Garmin methods are the same

Based on Runalyze's docs, the Runalyze method uses runs from the last 30 days by default (that's configurable). Or have I understood something wrong? From https://runalyze.com/help/article/vo2max : "Your current VO2max shape is based on the average of the last 30 days."

(And Garmin has used Firstbeat's algorithms for a long time, so Garmin buying Firstbeat didn't change anything in that respect.)

Bitti said:

According to Firstbeat's white paper

It's old algorithms support.garmin.com/.../

The link you give doesn't seem to say what you claim (best performance of 30 days). It says that IF last run is older than 30 days, Garmin will start using "less accurate" VO2max estimates based on other activity types (like walking) also. The link says "will only show updated readings with additional Run Activities with GPS for the next 30 days", so doesn't sound like "best of 30 days".

Honestly, I think everyone is so keen to outlink each other and win an internet argument people are losing focus of the actual question in the title!

I don't think the differential is anything to do with averaging, or taking the 'best' out of 30 days etc etc I don't think that would account for the large differences you can often see between the two metrics. Certainly in my case, I do not record walks at all so that wouldn't impact my large observed differential, and for most folks their walking HR is going to be too low to be used for VO2 scores IMHO. If Garmin are doing this, I think it's the wrong thing to do. I'm assuming most folks walk below 50% max HR here, maybe that assumption is wrong on my part. I would actually think walks if anything would trend into the low VO2Max side... 

Thank you all for your thoughts. 

As the season has just started for me here (Moscow, snow and ice having recently gone, etc) I can see both VO2max's are now beginning to converge. 

I blame Putin!