Lactate threshold detection is ridiculously incorrect

I thought that it could be related to stryd somehow (although it's quite strange), but nobody on stryd groups complained about it.

Anyways, I think you're right and it makes sense to try without stryd with GPS only. Although it's quite weird to switch from more precise technology to less precise to improve accuracy of algorithm. :) 

Same experience as OP here. Also using HRM chest strap and Stryd footpod. However, I saw the drop in auto-detected lactate threshold BPM and pace when switching from Fenix 5S to 6S back in September last year. I have been using the Stryd footpod since December - with no apparent effect on the auto-detected LT values.

Still, the pace and BMP values today seem quite off: Auto-detected LT pace is stated as 4:17 min/km - "true“ value based on experience and actual runs should be closer to 4:00. LT BPM is detected as 177, while it should be closer to 183 based on lab test results (HRM max is 199).

I actually don‘t really care that much, since I am mainly looking at VO2max as a personal training progress indicator (currently at 58). Race prediction time seems okay with the Fenix 6S (rather on the conservative side actually) - definitely way slower than with Fenix 5S (prediction times were too optimistic there). However, I believe the latter is due to an actual change in the underlying calculations between the Fenix generations. 

Thanks for sharing this information! Nice to know that I'm not alone. )))

I don't really care too, because I always know my thresholds much better than algorithms could estimate and use other techniques to predict races. But all in all this inaccuracy means that all other metrics are most probably quite unreliable too. For example, training load is HR-zones based. In it's turn, HR-zones are LT-based so training load is LT-bases and consequently inaccurate. If training load is inaccurate, then almost all these new fancy metrics are inaccurate too (recovery, load focus, etc).

I think it's a little bit too much of reliable / inaccurate stuff for a watch with this price tag. :)

I think it's a little bit too much of reliable / inaccurate stuff for a watch with this price tag. :)

I agree but would never admit it since I bought one (and owned a F3 for years) :-) Perhaps the correct would be to say that these kind of technology is too pricy regardless of brand/model...

I have another way of thinking than many of the users on this forum. Accuracy isn't as important as consistency. As long as the data I get is consistent I can use to to create my personal baseline and I can use it to analyze and compare one run from another. That is what I need to plan and train, not a 100% accurate value of my VO2 or LT.

When I look at 6-12 months graphs for VO2 and LT it is consistent without any larger exceptions (that's why I asked how your graph looks like) and reflects my shape at given times quite well.

I agree that in this case consistency is much more important than accuracy (although I would never buy a 500$ gun that is consistent, but not accurate :) ), but in my case it is not consistent too and the graph has nothing to do with my form.

Yesterday Garmin offered me to update LT pace to 4:51 after a long easy 20 km run @ 4:53 and HR ±145. No need to say this is one more physiological nonsense, but it looks to me that their model does not take HR into account at all and somehow based purely on pace/time pairs with some strict filter that for some reason filters out all my relevant data.

I discussed this issue yesterday with my friend who uses Fenix5+stryd and it looks like his data is much more accurate than mine on Fenix6+stryd. He has 4:02 LT detected and he runs 10k around 39-40 minutes, so it makes perfect sense. 

It seems to be the case, because LT is downgraded usually after lite training Ihad, but since the HR was also low, it should not Just Check my speed to estimate LT. That would be stupid. 

I can't say why it's so wrong for you but I can help saying how I configured mine and you can try.

I'm recording HR variability, 1sec recording, fenix 6 as my main watch, true up on.

Also I use hr reserve zones and change my rest hr everyday on the watch. I don't use LT as HR zones because it's a estimative and can change, that's why I only trust on he reserve zones.  

But I know my LT by feeling and it matches with watch the watch found. If I want to train at LT zones I just set the exercise to be at that range of HR (that I know it's my LT).

I'm having good results doing that