Personal records show activity time instead of actual record time

I don't think that's a bug; it's accepting that, if the distance is very close to the benchmark distance, it's more likely that you were running an accurately measured course and the deviation from the distance is watch inaccuracy, not course inaccuracy.

Cuts both ways, too; if you run a measured 5k course and your watch reads 4.9km, you can still claim it as your 5k record and the total time will be used, not scaled to longer or rejected outright for the short measurement.

I do not agree.
As far as I know, it will not set a personal record if the GPS-measured distance is <5km.

Furthermore, you can update the activity to have the correct distance when running an accurately measured course.
If Garmin is providing you an overview of personal records within their eco-system, this should be accurate according to their collected data. (allowing you to update the collected data if it is inaccurate!)
Otherwise, they should put "±5km record".

Looking for an official statement from Garmin on this one whether they will take this bug into account.
(would be sorely disappointed in Garmin if they don't)

1770932 said:

do not agree.
As far as I know, it will not set a personal record if the GPS-measured distance is <5km.

Go into the activity in Connect, tap the cogwheel for settings, tap Set as PR. You're welcome

This doesn't change anything...
The problem is that the Personal Records is inaccurate.
Not that inaccurate activities aren't set as PRs...

1770932 said:

For example, my 5km record is showing the total activity time of my 5.06km activity, not the time it took me to run 5km.
For my 10km record, it is showing the time of my 10.03km activity instead of the time it has taken me to run 10km.

The difficulty would be deciding what to do there

Do we take from 0 - 10K

or 0.03 - 10.03K

or process through the whole file and find the fastest 10K?

Admittedly a toggle would be nice so you could choose.

But if it's a genuinely measured course - edit the distance. You're going to get approx 1% error with GPS, so I'd say probably the best thing in that case is where you're certain the distance is measured, edit the distamce.

Will work well with shorter distances, but with longer races its quite possible to have (say for example) a measured 50K and still run 50.2 (for example) just from the shuffling to overtake etc...

I agree that you will perform additional distance on a measured course.
But this is fine, if you put the exact distance in the record.
If you run 50.2km, your personal record should still cut the 0.2km if advertised for 50km.

It is still up to the user whether they want to keep the distance measured by GPS, or change it to whatever he expected the distance to be.

It is actually very simple:
- The Personal Record is set for a distance of 1km, 5km, 10km, half marathon, marathon. So, it should be the time registered for THIS distance, not a distance close to it.
(if your employer says your salary will be X, would you accept X minus a few %, it's close enough, right?!)
- If Garmin is not confident in the data they recorded, they should allow for correcting this inaccuracy at source level (the activity), which is possible. Compensating for a possible inaccuracy in multiple places will only increase the inaccuracy.

Regarding taking 0-10km or 0.3 - 10.3km, this argument is irrelevant. This should be identical to any other activity.
If my fastest 5km is in a 10km run, it will correctly take the time of the fastest 5km. This should not be different is the 5km is in a 5.06km activity. (or the 10km in a 10.3km activity)

1770932 said:

- If Garmin is not confident in the data they recorded, they should allow for correcting this inaccuracy at source level (the activity), which is possible. Compensating for a possible inaccuracy in multiple places will only increase the inaccuracy.

Regarding taking 0-10km or 0.3 - 10.3km, this argument is irrelevant. This should be identical to any other activity.
If my fastest 5km is in a 10km run, it will correctly take the time of the fastest 5km. This should not be different is the 5km is in a 5.06km activity. (or the 10km in a 10.3km activity)

1) Yep they do - for each activity you can trim the record and/or edit the distance recorded. I've done this for an ultra which went about 20K underground (10x5km loops with 2K of each loop underground) - came up to 48K because of this so edited it to 50K

2) Again, as McBadger said, the current system does allow for wiggle room from systematic errors. What you're saying is basically c above - process through the file. The issue here is

You run a measured race of 5K exactly at say 30 mins. 
The watch, due to associated error says it was 5.05k (again, errors unavoidable via GPS)

So you get your race results and there you are, runner xxyy at position z with a verified time of 30:00

However Garmin has gone through that 5.05K, and given your 5K time of 29:54 (because it's gone slightly over distance) for the same time. It's processed through that 5.05K activity and found the fastest 5K distance (by its reckoning) is (for example) from 00:05 to 29:59.

That's why, I guess, they prefer you to edit the distance so you can remove the discrepancy and still get a time that agrees with the official record. 

I agree a choice would be good, to allow for all eventualities...

Had this for my half marathon record. 

It wasn’t a race I just happened to set my fastest half as part of a longer run (14mi)

Only way I could get the PR to show correctly was to trim the activity to 13.11mi. 

For all other records that occurred inside longer runs the PR did get registered correctly only the half for some reason took the total time of the activity as the record so for me at least there seems to be an issue/bug around the half marathon PR registry unless Garmin have changed things and my other PRs would also suffer this fate of obtained now. 

Yeah, I think the problem shows worse for the longer runs as the GPS error is a percentage and so the wiggle distance gets larger the larger the race. For longer distances that error mounts up

Ɛpsilon said:

You run a measured race of 5K exactly at say 30 mins. 
The watch, due to associated error says it was 5.05k (again, errors unavoidable via GPS)

So you get your race results and there you are, runner xxyy at position z with a verified time of 30:00

However Garmin has gone through that 5.05K, and given your 5K time of 29:54 (because it's gone slightly over distance) for the same time. It's processed through that 5.05K activity and found the fastest 5K distance (by its reckoning) is (for example) from 00:05 to 29:59.

That's why, I guess, they prefer you to edit the distance so you can remove the discrepancy and still get a time that agrees with the official record. 

Yes, in this case, the user should update the length of their activity to match the outcome of the race.
However, in reality, he will probably have ran 5.05K.

It's still up to the user whether he wants to change this...

With this bug, for ALL users, it is impossible, or, I would have to accurately keep an eye on when I am actually hitting 10km, and update both the time and distance afterwards for my record to show correctly?

All of this 'wiggle room' is completely irrelevant in this bug...
The page is stating the Personal Record for X distance. It should show the measured time for that exact distance!
As mentioned before, if they are unable to do this, they should put "±5km record"!