Comparing training effect between Software Version

The table below shows 6 easy runs, two from each of version 22, 21 and 20. The runs are between 39 and 42 minutes long on flat ground.

SW Version Average HR Training effect (aerobic) Training load
22 137 0.6 11
22 136 0.2 4
21 142 3.5 128
21 139 3.0 85
20 137 2.8 68
20 138 3.0 82

For reference I am 71 years old, with a VO2max of 44, and a lab tested (October) lactate threshold of 165. I run 6 days a week / 5 hours currently on a 80/20 half marathon plan.

The feedback from Training Peaks is that I am executing the runs correctly, with 80% in pace and heart rate zone 2. But Garmin seems to overestimate my hard runs and not recognise the easy ones. As you can see from the above table. This started with v22 of the SW. 

  • Stryd and Garmin are not partners, but compete with each other to some extent. I don't think Garmin has done anything intentionally to make its users abandon Stryd. On the other hand, there is a possibility that Stryd users will switch to another manufacturer, more loyal to Stryd, like Polar or now Coros, and I think AmazFit.

  • I agree it is not intentional. I assume that Garmin is making some changes to their physiological model to enable new features for the next generation of Forerunners.

    But I am a irritated with both companies. Garmin for braking stuff that goes to the core of what I want from my running watch and Stryd for being a bit arrogant  and not trying to do something to address the problem.

    I have long background as a consultant working with embedded software (in automotive) so I can understand the development approach, i.e. two-week sprints leading to quarterly releases and probably little traditional quality assurance. But they still get a lot of feedback during the beta cycles which they mostly seem to ignore.

    I would expect (hope?) Garmin to be more transparent and more responsive. Compared to the likes of Polar or Stryd, Garmin have much more resources, I wonder if this is management issue, chasing new features, new customers?

  • I wonder if this is management issue, chasing new features, new customers?

    I feel like it's something like that. It started about 2 years ago when they started with the regular quarterly update cycle. Ever since (especially in the 1st year) there were many quarters with added "features" nobody wanted or uses, and most bugs not fixed. I have a feeling that some new manager got into power and ever since the main goal is to add as many lines of new features added as possible to the quarterly changelog. It doesn't matter if it's useful or working at all or if it breaks other things that people do use. The main goal is to have the feature listed in the changelog.

  • When you look at the % in each HR zone in Connect for these runs , are they similar ,or have the HR zones somehow got altered in 22.xx .  Weirder things have been known to happen. if the max HR and HR zones got inadvertently changed with one of the 22.xx versions....

  • I have my zones manually set based on lab tested lactate threshold, so those are fixed in this six month period.

  • It seems Stryd is in denial on their Reddit forum. 

    Angus from Stryd killing any constructive discussion did it for me, I have cancelled my Stryd membership

  • Stryd is responsible for collecting data. Garmin is responsible for using and analyzing the data collected. There really isn't anything Stryd can do in this case. I would also try resetting Garmin to factory settings and reconnecting Stryd.

  • I understand that , I am asking if you look in connect under the activity , what does it show as the % in each zone for the 22.XX activities and the Prior activities.  People have reported that the Zones got altered in teh settings, If this happened and Garmin shows the zones strangely then this would affect the calculations by Garmin. So maybe just look out of interest as the % in zones on teh activities and compare.

  • If you refer to HR zones as I mentioned above zones are lactated based and the lactate threshold has been set manual based on lab tests.

    Up until two weeks ago I used Garmin default LT zones and they have not changed. Two weeks ago I switched to 8020 training plans because I could no longer trust the Garmin training plan or DSW. I then changed to using 8020 zones. I can see no substantial difference.

    I have been running versions 22.12, .14, .18, .21 and .24. The problem started with 22.12 and has persisted since then.

    With regard to power zones I wrote this on Reddit ...

    I did a 70 minute easy run today where I had set Garmin FTP to Stryd CP and Garmins power zones equal to Stryds. It did not solve the problem. So obviously there is more going on than pure time in power-zones.

    The 70 min run had average HR of 139, training effect 1,4 and training load 16.

    A similar run with v21 of the software from late last year had a training load of 108 and training effect of 3.3, so these values are clearly too low, although an exact comparison is not possible.

  • No what I am saying is , if you look in the activity on GC web interface , it shows % in each zone recorded with the Activity . Just take a look and see how they compare across the 6 activities you referred to. I'm trying to see if the setting in GC got messed up and that is what is affecting the TE and TL Garmin is calculating. My TE/TL still look logical across those FW versions so I though maybe looking at this may identify why its doing what it is. Of course I my be completely wrong!

    This is the table I am referring to (from one of my runs just to show where it is). Its at the base of the stats on the web interface