Related
Options

Device didn't ask me to accept detected lactate threshold.

JoseR
8 months ago

Did Garmin remember the prompt to accept (or not) auto detected lactate threshold? I normally OK with Auto detection but I want to have control over it, last run it was updated but the option to accept it didn't pop up. Not sure if it is a new feature (it won't be very stupid) or a bug. 

Thanks!

  • 0 8 months ago

    What I meant is it will be very stupid from garmin to update LT without asking

  • 0 8 months ago

    https://forums.garmin.com/outdoor-recreation/outdoor-recreation/f/fenix-7-series/350860/garmin---fenix-7-epix-2-series---excel-changelog/1859008#1859008

    • Function Threshold Power estimates are applied automatically after training with power.
    • Lactate Threshold estimates are applied automatically after training with heart rate.
    • Removed legacy FTP test.
  • 0 8 months ago in reply to JoseR

    Sorry. I am the one that is late to the party. I am still on 18.16 and, boy, I am glad I didn't update.

    I am skeptical about the new approach to remove the LT guided test and rely on the auto-detection to get accurate threshold pace estimates.

    I understand this new development is easier and probably safer for the novice user, although I would be worried about accuracy if using the wrist HR to get results. I guess ease of use, out of the box way to get safe training recommendations and metrics were the guiding principle.

    For whomever cares about accuracy, using a chest strap is the only way to go.

    For lactate threshold pace, even with a chest strap case, relying on auto-detection  has not worked well for me at all. I always found that the guided test was closer to the all-out steady 30mn effort that threshold entails. The auto-detection ends up giving me a slower threshold pace (about 10-15s/mile slower). Slower is safer but sub-optimal for endurance training, in particular for 5k and 10k race training.

    For LTHR, I always saw good consistency between the guided test and the auto-detection.

    Since there is no way to input your lactate threshold pace manually, removing the guided test is worrisome to me.

    I guess the workaround is to enable auto-detection only for runs that are in fact threshold tests: let's say at least 20mn, but ideally 30mn all out steady efforts. Then, the auto-detection should be turned off.

    Unfortunately, suggested workouts tend to max out at 18mn for single block threshold runs, which is on the shorter side for real-world threshold detection.

    Not a good development in my opinion, but we'll see. Maybe there are new capabilities in the algorithm that will address my concerns.

  • 0 8 months ago in reply to Etupes25

    Yes, I guess I'll disable auto detection for the time being...

    Thanks!

  • 0 8 months ago in reply to Etupes25
    Etupes25 said:
    Not a good development in my opinion, but we'll see.

    TBH, although I agree that the self guided test is the best way to get relatively accurate LT, personally I am not ready doing it every other week or monthly to follow my progress. Hence having a consistent method to detect the LT is more important to me, than its accuracy or conformity with lab tests. So, as long as the auto-detection works in a consistent way, I am all right with it, even if it gives lower (or different) values than the guided test would do. Finally, I can always mimic the guided test, by doing a run with stepping up the effort by stages from low to all-out, and I assume the auto-detection gives the same result as the guided test would do.

  • 0 8 months ago in reply to trux

    I am not worried about the LTHR data point, because I am not aiming for a specific HR value when I run or race, but monitoring a zone/zones

    But for pace at lactate Threshold, this is a different story. I use this data point carefully to set a target pace for races (plenty of calculators to do so). So I need this to be as accurate as possible.

    When I see a discrepancy of 10-15s/mile between the test and random auto-detections, typically occuring after anaerobic intervals where pace varies a lot, I am worried. Maybe 10-15s doesn't sound like a lot for workout runs, but for race there it is a huge difference, at least for me.

    Although I am still tweaking +/-1ppt, my ratios are 109%, 101%, 95%, 91% of threshold pace for the 5k to marathon.

    The lactate threshold pace I get from a guided test with these ratios gives me PB very consistent race predictions with Galloway's Magic Mile, Garmin's, Stryd's and various online calculators based on it. So I grew used to trusting this number and discarding the auto-detection slower estimates.

    I personally like the test as workout, it is a good typical progression run, so I replace a tempo/threshold workout with a test from time to time.

    I will run a new test in the next days, and I have a 5k coming up in a months. So I guess I will get a new data set to confirm my past experience...

  • 0 8 months ago in reply to Etupes25

    I admit I did my last self-guided test almost 3 years ago, and since then I only use the auto-detection. Surprisingly (at least for me), the race predictor is pretty accurate. More than I could ever expect. I did no marathon yet, but the last half-marathon a month ago was within 20s of the predicted time (without me trying to match the prediction), and the 5k and 10k predictions are even closer to my PR's, which although they are a few months older, I think I could still repeat them again despite progressing age.

  • 0 8 months ago in reply to trux

    To be fair, I'm not against auto detection.  I wear a chest strap when I want to test LT in certain workouts or races. In May case OHR works pretty good for most of times so I shouldn't be a problem... Until it is. For instance, in winter my hands can get very cold and sometimes the HR sensor fails to catch accurate readings (at least in my case). I would like to have control about accepting the stimation in those cases