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

ETA/Live Segment Algorithm

I am considering purchasing an Edge 530. I currently have a Sigma Rox 11 GPS computer. I just tried the Strava Live Segments feature and I am curious how Garmin's algorithm compares.

It seems like my current computer uses a very simple algorithm to calculate whether you are ahead or behind your PR (or whatever you are tracking). It seems to simply calculate your average speed, and compare it to the average speed of your PR, then calculates a time delta. This is not a good solution because if the segment is a climb and the grade varies, your speed will vary throughout the segment. So if the segment starts out easy, it will show you way ahead of your PR, but when the road gets steeper, your speed drops, and your time delta drops. It's the same when navigating a route. It gives an ETA of when you will finish, but it's usually wildly off, because my routes tend to end with long descents, and so my average speed increases drastically at the end.

Is this how Garmin does it? Or are they smarter? I wish it would calculate live segment times based on where you were and at what time on your PR. Does that make sense?


Thanks!
Alan

  • Hi Alan, it's a coincidence that you post about this now as I was just about to do the same. I am a long-time Garmin user (sufferer) from the 105 to the 520 and just for my last two rides the 530.

    My 520 used to be OK with Strava live segments and was a good motivation for training as it was fairly accurate. But it seems something changed, I think Strava changed? and now it doesn't work properly with Garmin, but this just a guess, anyway it went crap with my 520. 

    Time in front or behind will jump from 30sec to 1min and back to 30sec very quickly, sometimes even more, it makes live segments no use at all :( 

    I hoped the 530 would be better but it's just the same :( 

    The problem is much worse than the averaging of up/down segments that you said about as most of my segments are long climbs with no descent and the time still has random changes. 

    I have tried turning off the Auto Select goal feature but no difference, I hope someone here has some ideas?

    If it's not fixed soon I might just give up and cancel my Strava Premium as I only pay for Live and it's broken with Garmin.

    There is a long thread about it here it is not really a new issue I'm surprised there is not more discussion about it.
    https://forums.garmin.com/sports-fitness/cycling/f/edge-1030/143209/strava-live-segments-issues-comparative-times-are-not-accurate

    Probably best to just go back to doing what I did 15 years ago and put a chalk mark on the road and use a stopwatch with that and write it on paper. 

  • I have done some more checks and the 530 is definitely much worse at Live Segments than my 520 sadly.
    I'm using beta 3.55 and GPS 2.42 I will try beta 3.57 soon but won't be able to use it on the same segments for awhile as I will be away on holidays for a few weeks.
    I assume the "improved livetrack reliability" of 3.57 has nothing to do with Strava live segments anyway as that is livetrack for the whole ride which is different to Strava?

    Riding this Segment https://www.strava.com/segments/4548789 with my 520 it shows me how far in front or behind my PB I am and it's easy to read and fairly reliable.
    Riding the segment with my 530 it jumps back and forward between KOM, PB, rival time, recent best I have NFI what's going on or what the plus and minus times are referring to and it doesn't matter anyway because the times are all far from accurate. I have auto Select Effort turned off, but it does it anyway.

    Then when I get three-quarters of the way through the segment the 530 tells me "off course" and drops the segment, did it three times in a row at the same spot on the road, no tree cover, sunny day. My 520 never drops the segment and I tested it again today and it worked fine and was easier to read and understand. 

    Considering the other threads here about the 530 poor GPS accuracy when MTBing and now mine losing the segment on the road I'm starting to wonder if Garmin have used cheaper inferior GPS hardware in the 530 ?




  • Thanks for the info! It looks like I will be staying with my current setup for a while. The live segments actually aren't that bad, I guess.

    Alan

  • I’ve noticed the time behind jumping around all over the place on the 530 as well. Was always solid on my 510.

    I’ve also read that Garmin have started using lower power Sony GPS chips that have had issues in new devices (unconfirmed if in the 530 but would make sense) that may explain some of these GPS related issues.

  • Wow.....I waited for them to release the recent huge bug fix before purchasing. But foolishly it looks like I have still been tricked into buying a broken product

    The lack of official replies about the GPS chip is a real concern. I expect the 530 can't be  fixed by a software update.

    I wonder if I can return mine to get a functional GPS chip installed? 

  • I bought a 530 full well knowing that I was likely to be a Beta tester, software problems can be fixed, hardware problems (if that’s what this may be) are an altogether different proposition - there’s no proof of this yet, but I know other brands from Polar, Suunto etc that have used the Sony chip have had a lot of issues with it.

  • Hi, Garmin Support, any chance you can address this question? Are Strava Live Segment virtual partners based solely on average pace or based on actual "ghost" riders that represent the target riders actual speed variation across a ride (e.g. faster on flatter parts of the segment, slower on steeper parts)?

    Thanks!