Related
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

Garmin running power ftp

ceejay
over 1 year ago

I'm doing a 10K race in a few days and decided to use Garmin running power rather than pace. Today I did a test run based on 90-95% of FTP, which should be around my 10k race pace. The resulting pace was far less than required to reach the watch's 10K estimate.

Assuming the FTP was too low, I did a 3-minute/9-minute FTP test and manually updated the running FTP on the watch. I reran the workout on the same course and got a pace in the ballpark to meet the 10K race estimate.

I've been using the watch for a while and have running FTP set to auto-detect, so I wonder why the estimate was so far out - 32 watts less than my FTP test.

How is Garmin actually estimating the FTP?

Thanks

Top Replies

All Replies

  • 0 over 1 year ago

    The FTP auto detect has even more problems with accuracy than lactate threshold auto detect, because the detection cannot properly account for uphill/downhill and all other disturbances and sources of error that normal running contains. It might work for you, or it might give inaccurate results, and there's no way for you to know (and the same goes for maxHR auto detect).

    I personally do the guided lactate threshold test once in a while on a running track. That minimizes disturbances and unknowns. Of course it might still be off, but that's the best one can easily do.

  • 0 over 1 year ago in reply to Bitti

    Thanks for that. I'll stick with my manual test for now, it seems more accurate than Garmin's.

  • 0 over 1 year ago

    I don’t know the answer to your question but I’ll say that my last 10k was at 121% of my FTP (435 W and 360 W respectively) - at least according to my FR955.

    I never use running power so I don’t know if that’s normal or not. It just seemed odd to me considering you were planning on running at 95% FTP for your 10k. 

  • 0 over 1 year ago in reply to jambongris

    I read that 10K pace would be between 95% and 105% of FTP.  Though I've trained for it, It's my first race in several years, so I chose a more conservative target.

  • +1 over 1 year ago
    ceejay said:
    How is Garmin actually estimating the FTP

    I didn't find any documentation of the method used to choose the FTP value. I doubt they use a power-durartion model (like 95% of your best 20 power) because the builtin threshold test is not long enough.

    I think they use deflection points in the relationships between HR, HRV, pace and power and assimilate the FTP with the lactate threshold. Purists would argue that the technical definitionof FTP is not linked to the lactate threshold but actually is the maximum power you can sustaing for a known duration (60mn).

    For more information of these deflection points, check out this article: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3438148/

    From my experience, the built-in test is reliable and accurate, but the auto-detection is not. You might get a new FTP detected during an anaerobic workout. When it happens to me, the FTP detected seems to be lower. Vice versa, I tend to get slightly higher FTP values when I actually run/bike a target duration at maximum steady threshold effort.

    For the purpose of training, the built-in test is very useful to set your HR and power zones. I am very happy with it.

    ceejay said:
    I read that 10K pace would be between 95% and 105% of FTP. 

    This is correct. 10k effort is typically at threshold, while 5k is slightly above (around 4% higher .

    jambongris said:
    I don’t know the answer to your question but I’ll say that my last 10k was at 121% of my FTP (435 W and 360 W respectively) - at least according to my FR955.

    121% of FTP seems high indeed for a 10k. I would suggest your FTP value is too low and should be very close to your 10k average power. You might want to manually enter that value and see if the power zones match your efforts better. You could also try the built in test. Note that the buil-in test is influenced by the HR Max value that you use on the watch, because this value is a driving factor in the relatiopnships models between HR/HRV, pace and power that the watch is calculating and maintaining.

  • 0 over 1 year ago in reply to Etupes25

    Thanks for the detailed answer. The race is tomorrow and I'm happy with the FTP from the manual test for now, but I'd like to get the auto-detection working in the future. 

    When you mention "built-in test", are you referring to the guided lactate threshold test, or something else?

    Thanks

  • 0 over 1 year ago in reply to Etupes25

    Are you referring to the guided LTHR test? If so, I haven’t used it in a long time, I just rely on the auto-detected values. I’ve got a marathon this weekend but I’ll give the test another shot once I’m recovered. 

    I’m not really bothered my power zones since I don’t use them but I am curious to see if the guided test gives me different results than the auto-detect function.

    Im confident that my MaxHR value in Connect is correct. 

  • 0 over 1 year ago in reply to jambongris
    jambongris said:
    Are you referring to the guided LTHR test?

    Yes, the guided LTHR test. 

    Also, as I was researching this, I found that article that suggests that auto-detection is using a power-duration rule for the FTP estimate, whereas the built-in test has to use the expected deflections in the HR/HRV vs pace or power.

    This could explain why some anaerobic workout get me an FTP estimate with 20mn average power degraded by recovery periods.

    It is also possible that this article is only referring to cycling, an activity for which we know that the device is maintaining a power duration model.

    support.garmin.com/.../

  • 0 over 1 year ago in reply to Etupes25

    Yes, I think the linked article is about cycling. Running FTP isn't recorded on GC web but can be found on GC mobile under performance stats->running lactate threshold. I'll do the test and see what estimate it comes up with and whether it is anywhere close to my manual test. It would be good if Garmin made it clear that the LTHR test is instrumental in getting an accurate running FTP. Also, it is not convenient for users without HR straps.

  • 0 over 1 year ago in reply to ceejay

    I think there are two factors to consider between the guided LTHR test and the automatic LTHR detection. One is whether they both use the same algorithm (only Garmin knows), the other is the "guided" part, i.e. the test guides you to run in a way that provides best data for the algorithm (and preferably choose a route with no up/downhills etc.). And forces you to use a HR strap for most accurate HR variability.

    For me, when I still had auto detection on, it often offered values that were close to what the guided test gave me. And sometimes it gave values that were off. So auto detect is better than nothing, but understandably more unreliable than the guided test. That's why Garmin gives you both.