Related
Options

Discrepancy in 5K PB, Garmin vs. Strava

slsmpl
17 days ago

I hit my first sub-30 5K yesterday. My watch showed 5.00 km in 29:49, and I was absolutely hyped. I then slowed to a walk to catch my breath, ended the activity at 5.04 km in 30:13, and Garmin logged my 5K PB as 30:13. How does that make sense?

I imported the same activity into Strava, and it recognized my 5K best as 29:49. What went wrong? Now I’m wondering if my previous 5K splits, those I recorded during 10K runs, were accurate, or if they were off by a similar margin. A half-minute discrepancy seems huge.

Has anyone else seen this?

  • 0 17 days ago in reply to kochmans

    Thanks! I did try searching before opening a thread, but I probably used the wrong keywords as I misunderstood where the problem originates from.

  • +1 17 days ago

    Just to summarize the issue, if you run slightly further than the PB distance, Garmin basically treats your activity like it was a race, and for the PB, it uses the time/pace for the entire activity, instead of using the time/pace for the exact distance. This is because when you run a race, it's very unlikely that you'll record (or even run) the exact race distance, so what really matters is the time and pace for the activity (assuming you started and stopped the timer exactly at the start and finish line).

    Like if you run a 5k race, you may run slightly more than 5k just because you didn't hit the tangents perfectly. And your watch may record even slightly more than that, just because of GPS error. But despite both of those things, for the purposes of your official distance and pace in the race results, you only ran 5k. And that's kind of what Garmin is trying to reflect here.

    If you ran a 5K race where the watch recorded 5.10 km (for example), and you set a PB at that race with a total activity time of 28:30 (for example), if Garmin took the fastest 5 km section of that activity, it would obtain a PB that is actually a bit faster than reality.

    On the flip side, if you run a lot further than the PB distance in question, Garmin does the same thing as Strava and just uses the time/pace for the exact distance. So if you set a 5k PB during a 10k run (for example), both Garmin and Strava will use the fastest section of exactly 5k in that activity for the PB.

    Honestly, I think what Garmin does makes a lot of sense if you understand why they do it, but I also get why people don't like it (it's not transparent, intuitive or consistent).

    On the flip side, all of my Strava "best efforts" that came from races are slightly too fast, precisely because it doesn't do what Garmin does. Strava's method of always using the exact distance for PBs is very straightforward, but it actually gives bad results when those PBs come from races.

  • 0 16 days ago in reply to FlowState

    I get what Garmin’s trying to do, and it’s a clever workaround for race day GPS issues. I probably wouldn't notice or care if it was just a regular run, but I really wanted my first sub30 ever to be true. So I trimmed the activity, just this once, and now it's 5.00 at 29:49. Hopefully I'll repeat this feat soon and make it count without any "cheating". :D Thanks!

  • 0 16 days ago in reply to slsmpl

    Yeah, I get why this behaviour is confusing or annoying for users. It's one of those cases where doing the "right thing" (in terms of races) is confusing or unintuitive for users. Strava arguably does the "wrong thing" (when it comes to races), but everyone likes Strava's approach.

    I will point out that Garmin's approach accounts for more than GPS issues on race day. When you run a 5K race (for example), it's very unlikely that you run exactly 5K, unless you run on a track. Weaving around other runners and failing to perfectly run the tangents can add to the actual distance you run during a race.

    Besides, GPS error is something inherent in the system, it's not an intermittent issue that can be fully resolved or that only pops up some of the time. (That's why Garmin, Apple, etc. introduced track mode, right?)

    When I run races with my Garmin, I'm used to seeing about 1-2% more distance on my watch compared to the official race distance. I have only one 10K race where my Garmin recorded 10.0 km, and I think it's probably a freak occurrence (probably all the errors cancelled out just this one time).