Lactate Threshold bug in GC

@GFM1011, one more suggestion.

Maybe your developers can simply adjust the Pace axis scale to something more "human" (there is no way anyone runs 1:10min/mile runs). If you adjust the pace axis scale to +-1min/mile around the median of all recorded LTHR values, then at least the "bad" values will be out of range and not be displayed until you fix them. This way we could still browse our pre- and post-bug measurements and still get an idea of progress or not.

Please fix that thing as it is unbelievably annoying for a feature that is only available on your $450 top of the line watch!

True, this is advertised as one of the major features of fenix 3 HR, but only works a few days since I got my watch in late April. After some update, it stops working and this is not acceptable for a product that costs about $700+.

@Joseph

I made an manual Lactate Threshold test an all is working welll.;)
I got a new value and the watch was upload to Garmin Connect and displaying correctly. :o:o

Only the old values are wrong.

Now I will try the automatically tracking.

Thanks

Hi all and here is a small update on the still messed up historical LTHR values:

I was independently contacted by Garmin's Social Support with a proposal on how to fix the erroneous past LTHR values. I am not copy-pasting the whole email here but the steps involved are:

"We wanted to reach out to you to offer an alternative solution in the meantime, so that the graph will be correct. The potential workaround would be export the activity files and then re-upload them to Garmin Connect. The reason why we did not offer this as a solution in the forums is because the team is working to automatically correct them for users. If you would like to export the activities and then re-upload them please do the following:

1) View the following link for instructions for exporting activity data: Export activities and wellness data files from Garmin Connect
2) Locate the activity or activities that contained a new Lactate Threshold value
3) Select the Export Original option from the above link
4) Extract and store the exported FIT files on your computer desktop for easy access
5) Once you confirm the exported files are stored on your computer's desktop- delete the activity or activities from Garmin Connect
6) You can then manually upload the FIT files to your account"

Now I immediately wrote back asking on how to perform step 2). Unfortunately, no one knows which exact activity "triggers" an LTHR calculation as the value is not stored in the FIT file (that's what they told me at least). So overall I don't feel comfortable "hunting down" activities, removing them, refresh the LTHR report (to see if it was the right one) and iterate until all bad values are gone. I agree that it could be a solution for someone who might have one or two bad LTHR values but I have a ton and I just don't have the time/patience to hunt them all.

I just wanted to share this info with everyone because it might be useful to someone. Also please make sure you do save your original FIT data because you don't want to permanently lose your runs. Lastly, as the original Garmin email states, this information was not share with everyone (until now) as they are still trying to come up with an automatic fix for everyone.

I will definitely be waiting for Garmin to fix this as I really don't want to lose any of my activities by my own mistake. Please share your experience if you go the above route.

Awesome, thanks so much for sharing (the response I got earlier today from Social is that there is no way to delete those entries - duh). I quite easily pinned down the dates for the corrupt entries using the 7 Day view in Reports, downloaded the corresponding .fit (Original) files, deleted the activities (then waited/refreshed like crazy until the lactate entries vanished from the chart - it took from seconds to minutes). Then I re-upped the .fit files, and waited some more until they reappeared in the charts with proper values. All good! :cool:

Ha! So it totally worked for you? If I may ask, how many LTHR entries did you have to find/fix?

Ha! So it totally worked for you? If I may ask, how many LTHR entries did you have to find/fix?

4 or 5. Seems only those I recorded in May/June were affected.

4 or 5. Seems only those I recorded in May/June were affected.

Thanks @ToffenDask! I think I am going to bite the bullet and do it!

Alright, I did it but unfortunately I have good and bad news.

The good news is that I had in total 6 wrong LTHR values spread out 10 activities. 2 days with wrong LTHR had 3 running activities each thus 10 total. I located each and everyone, downloaded the original and deleted it from GC, so far so good. I re-uploaded the six activities and here is how the LTHR values changed (date, LTHR before --> LTHR after)

Tue Jun 21, 164 1:56 --> 164 6:57
Tue Jun 7, 163 1:58 --> 163 7:03
Sat Jun 4, 161 2:03 --> 161 XXX
Tue May 24, 162 2:03 --> 162 7:23
Sun May 15, 164 1:57 --> 164 7:00
Tue May 12, 162 1:58 --> 162 7:03

Now for the bad news. As you see in the list there are is an XXX. Unfortunately the activity on Jun 4th post an LT-HR value of 161 but no corresponding LT-Pace! This is really weird and seems like a bug to me. I deleted and reuploaded the activity but again no LT-Pace was calculated.

See the attached graph to see how the data looks like:


Please keep an eye for your values if you go this route as clearly this is still a work in progress and this is exactly why I didn't want to do it myself...

I discovered one additional problem this time in the 12-month view of the LTHR:

As you can see on the right-hand Y-axis, the range of LT-Pace values is from 2:15 to 13:25. It seems to me that the Pace axis min/max values did not update after the re-uploading limiting the usefulness of the plot. As you can see it's very difficult to judge the variation of LT-Pace as it is so "compressed" and looks like a "flat-line".

Once again, I am very very disappointed with the way Garmin is handling this. It seems that the "data-scientists" behind these more "advanced" features don't really know how to handle calculation and visualization of data :-(

I deleted and re-uploaded all 11 activities between May 20th and July 7th containing LTHR values that had wrong LT values - all of them have valid HR and LT values now, no LT drop outs like KALAMARIOS mentioned two posts above. Thanks for posting the workaround.

My LT scale is also still messed after all uploads have been parsed and the graph values look fine again - the range is between 8:20 and 1:24 min/km, so the LT graph is almost completely flat. Min./Max. values stored in the background for all entries (which influence the scale range) are probably updated only if a new import contains values lower than the minimium or higher values than maximum. So please recalculate these min/max-values across all entries (within the scope of the announced LT value fix for all users without re-upload), but also for users where no LT value has to be adjusted, so that everybody who fixed the graph by re-upload also get their LT scales aligned correctly. Thanks!