How do you define your PBs?

This could be a GPS issue or course measurement issue but I have found the majority of the courses I have run seem to come up short on the GPS (not that I've ran loads). Generally I'll use the PB from the course as my PB time and anything else as a strava time e.g. Ran Clontarf half and my GPS and strava clocked me 200m short where as I had covered the 21.3km strava segment.... So I go with the race time.

My official racing over the years includes the sum total of

1 x 5k race - last year at 19.5X which is close enough to my current level

1 x 10k race - 3 years ago (49.18 in 2019)

2 x half -3 and 5 years ago (1.56 in 2019)

1 x Marathon(not including virtual runs) - 3 years ago (3.59)

To put those official PBs results in perspective, three weeks ago I went out on a weekend 16 mile session, 13 miles of which were at 7.00/pace quite comfortably. After uploading the run strava advise me that the below were PBs :-) from this run alone.

1.31.38 HM

1.09.40 10 mile

43.04 10km

This is my normal so im inclined to use times I have recorded on watches while running a distance I am happy enough is very close to the GPS tracker. I have found that even if the GPS on some of my watchs in the past has been a slight bit dicky, strava usually returns better data on upload.

I really should race more. I reckong I could PB in every race distance(once I get rid of this fuclking covid)

Races only. You can have your TT numbers but I'd call them "TT PBs". For instance my virtual marathon TT time is faster than my best marathon time but it's not reliable.

Wouldn't discount parkrun for a 5k PB if you're confident the course has been properly measured.

You can have a shorter PB in a longer race (e.g. 10k PB in a 10-mile race, but again only if you have an official split time from that point in the race (e.g. from a timing mat at 10k - some races do this, not many though.)

You can have a net loss of 1m per km once the start and finish are not further than 50% of the race distance apart.

It's the IAAF rule for what profile counts for record purposes, 0.001% drop wouldn't be crazy or massively fast.

Whether an official race, or a personal run, my times tend to be quite consistent when captured in a similar timeframe.

That said, f I go out for a 5km run tonight and my Garmin tells me it's my new PB, then I'm considering it my new PB. It's only for me, it's not my identity or qualification or anything like that, and I'd caveat it if asked, but to me, it sets MY target to beat next time I'm chasing a goal.

Most of my personal runs are circuits, so there's no real elevation bias. Would official race distances also be based on an optimum racing line? Just know some of my runs I'd fall into the more congested middle.

That said, my PB for a 10 mile, half marathon and marathon are all official races. 5km and 10km are TT, given I'd cover those distances far more frequently.

Sessions don't count. Time trails don't count. Park runs don't count. And the Ballyshannon 5k doesn't count either 😁

Just make sure not to forget those extra 9 metres. I’ve seen it happen at a beer mile. 😁

1m per km IS .001%

Net elevation can be very misleading - look at Boston marathon course. Net drop 480 feet, point-to-point course so not record eligible. But still a brutal course with more distance climbing than descending.

On the whole though, if the rule is good enough for IAAF it’s good enough for me.

1m per km is 0.1%, surely?

As for TT PBs, maybe I'm different but there is no way I could ever PB in a TT, no matter if I would count it or not. That race day adrenaline that comes with putting a number on your shirt is powerful stuff, and no matter how hard I tried to push myself in training on some rare occasions, I was never getting anywhere near my race times.

Also, PBs are PERSONAL, and it's everyone's personal opinion that counts. If someone wants to claim a TT PB, and it is genuine, that's entirely their own business and I could not care less.

Just assumed (wrongly, obviously) that you might provide some reasoning to justify your opinion, seeing as Boards is a discussion forum. No matter. Are you confident that your Garmin/Strava PBs are accurately enough measured to mean anything? For most of us, that becomes critical as we get to the edge of our abilities.

As BB67 pointed out, that 1m/k is the IAAF rule if routes are eligible for records or not; if anyone wants to apply that to their own personal bests is up to them.

Automatically counting all races and discarding non-races for PBs isn't as obvious as it may seem, btw. I have run a 10k that was 250 meters short, I have run at least 2 marathons that were definitely short, and of course I've also run a few races that were too long, and by quite some margin at times!

Oh, and I definitely did not mean to put you on the spot for that math error!

If a mile is 5280 feet, then surely a 1% drop is 52.8 feet, or over 8 and a half feet for each of the 6 minutes?

0.1% = 5.28 feet.

This guys is why the metric system was invented 😂

Even in the Alphaflys 😋

Measured courses only count for me. In an ideal world, all PBs should be on a track as it's the most consistent environment in terms of distance and terrain. Unfortunately very few run track, including me so measured road races are the next best thing. I've ran what would have been big PB's in TT's last year but I wouldn't dare count them as PB's, moreso good training runs and positive race indicators. Each to their own but I think you're deluding yourself if you consider a run on a Garmin as PB. My PB's would look a whole lot different (and more impressive) if I solely referred to my Garmin. GPS is notoriously inaccurate and variable between two identical devices. Just look at some of the variations of a typical 5k (3.11 mile) race on Strava, you'll see anything ranging from 3.0 miles to 3.3 miles and the longer the TT, the more inaccurate you're getting. For that reason, the best indicator has to be an approved measured course and even that's never 100% accurate as we all know.