This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

How VO2Max is estimated

I'm sometimes borrowing my FR235 to another runner, which, unsurprisingly, totally throws the VO2max estimate off. Unfortunately it stays off even when I delete such an activity.

Is there any number of runs or miles after which the device just "forgets" the event? Or some way to specify that a particular activity should not count as "Running" profile for VO2max estimation purposes, although it was generated with Running profile settings?
  • Is there any number of runs or miles after which the device just "forgets" the event?


    As far as I'm aware, the device does not retain ‘memory’ or previous VO[sub]2[/sub]max estimates except for the most recent one, so the next time the watch makes an estimate, the previous one will already have been forgotten. Garmin Connect, on the other hand, will retain persistently every data point on such that you have every uploaded (by connecting the device and syncing).

    Or some way to specify that a particular activity should not count as "Running" profile for VO2max estimation purposes, although it was generated with Running profile settings?


    By ensuring heart rate readings are not available to the FR235 during the timed activity, you'll prevent the device from estimating VO[sub]2[/sub]max. You don't need to change the Run profile settings for that; you just have to disable the toggle for wrist Heart Rate (which is a device setting and not an activity profile setting), and ensure that the device cannot connect to an external ANT+ HR sensor.
  • I am sure it keeps a few for trending rather than giving a new value from scratch each run.
    My guess would be 2-3. But I could be wrong.

    If you have a bad HR lock it may drop 1 or 2, never had it drop more than that which it would if purely based on that run.
    We discussed this recently where my 735 prompts for whether I want to accept lactate threshold, I think it really ought to do the same for vo2.
  • I can't find it documented anywhere but I believe the VO2max displayed on the watch is a rolling average of your last 12 runs.

    I tried to test this a couple of months back by deliberately allowing 'light leak' on one run in order to register crazily high HR readings. This had the effect of pushing the watches' VO2max figure down by 4 ml/kg min. It reverted to a more normal/correct reading 12 runs later which I assume means that this run had dropped out of a rolling average calculation.
  • I am sure it keeps a few for trending rather than giving a new value from scratch each run.


    I've never seen a VO[sub]2[/sub]max chart/graph on a Forerunner watch showing multiple data points from different days. As far as I can surmise, Firstbeat's algorithm does not need or use any previous estimates to calculate a current/new value.

    I know the FR230/235 Owner's Manual states,

    The estimate may seem inaccurate at first. The device requires a few runs to learn about your running performance.

    but I'd expect that has more to do with calibrating the user's running stride length, etc. to inform the calculations, as opposed to building a history of VO[sub]2[/sub]max estimates with which to tune the Firstbeat algorithm itself.
  • I'd read the FirstBeat article some time ago. They appear to be trying to detect (and disregard) running uphill/downhill, running on a soft ground, and some other factors. As the device has very limited inputs, and some runs are therefore entirely "disqualified" or severely reduced (objectively - but the device may not detect that, and it may happily pick certain values and use them to "recalibrate" itself), I can see how borrowing the watch can interfere with various parts of the algorithm in very complex ways.

    What I'm seeing is that whenever I borrow the watch even for a couple miles, the VO2max drops from something like 51 to around 45 and it takes weeks to climb back. I'll test the "12 runs" theory and report back. If the theory is exactly right, it would predict that the VO2max estimate will jump back, not climb back, after the 12th "standard" run. Other variations are possible if the aggregation of the last n runs happens over some inputs, rather than outputs of the estimation algorithm.
  • Might be worth contacting first beat directly or leave a message on their FB page. I had a doubt and left a message on their FB page, I got a reply a couple of weeks later, apologised for the lateness.
  • You could tell your friend to use the "Trail Running" activity instead of "Running." Or you could make a copy of the Trail Running one and call it something like "[Your Friend's Name] Running."

    Garmin doesn't calculate V02 max for Trail Running due to the hills and what not.
  • Wow, using "Trail Running" looks like the best way out so far (more flexible than disabling heart rate and/or GPS). But my FR235 only has Running, Bike, Other, and two indoor profiles and I basically use all the profiles already, with only Running setup suitably for outdoor running. Will try to configure Trail Running somehow.
  • Wow, using "Trail Running" looks like the best way out so far (more flexible than disabling heart rate and/or GPS). But my FR235 only has Running, Bike, Other, and two indoor profiles and I basically use all the profiles already, with only Running setup suitably for outdoor running. Will try to configure Trail Running somehow.


    Oh sorry I didn't realize the FR235 didn't have a built-in trail running mode. You could just make one from the "Other" probably.
  • You could also configure the Biking, Indoor BIking, or Other to have the same menu's as running with GPS on.... then when imported into Garmin...switch it to a run.