I may be wrong here, but from some initial reading, and listening to early podcast EPs, it doesn’t seem as though heart rate is used much at all in calculating fitness trends on Xert.
If I’m doing a session that’s quite low intensity, I’m not likely to budge MPA, and certainly not generate a ‘breakthrough’, but I’m likely getting fitter.
For one, over time, I’ll likely be able to hold a higher wattage at the same heart rate - an indicator of increased fitness, and arguably a ‘breakthrough’.
Is it because heart rate can fluctuate day to day because of heat, dehydration and illness? If so, have you looked at ways at smoothing out these? E.g. resting heart rate test each morning?
Part of the reason I feel strongly about this is that Xert doesn’t seem to ‘reward’ LSD rides for example - which are incredibly valuable. And, IMO this promotes a culture of NO INTENSITY NO GAIN which I think is detrimental to the sport as a whole, and so often leads to burnout.
I don’t know about the Optimal Decay setting, but on No Decay (which is how my account is set) LSD rides result in increases in TP and LTP.
What kind of reward are you thinking of? Other than the training load matched fitness signature decay method (No Decay) I think the reward from LSD would come from your next breakthrough effort.
Agree, though it’s maybe less about the ‘reward’ and signature update, and more about potential insights from those low intensity rides. Having some useful analysis of power vs HR, decoupling or something else for the >80% of rides that are low intensity would fill a gap. (And yes I know other platforms do this, but a) why not bring it together in one place and b) given Xert’s origins from HR analytics I wonder if they can do something even better if they prioritized it)
Thanks for the heads up on no decay - I’ll switch to that for a while as I’m not going to touch any real intensity until early next year.
I’m not necessarily seeking the carrot here, but I do think that there’s benefit to rewarding the LSD ride in some way - which could look very much like a traditional breakthrough, albeit relating to heart rate as opposed to power.
A breakthrough here might be triggered by a continuous reduction in heart rate decoupling at your aerobic threshold. Or perhaps a reduction efficiency factor over time for LSD rides (which I believe is normalised power / average heart rate).
Haha you wrote this as I was typing - great minds…! Didn’t know Xert came from an HR background - I’d assume then that this has been considered at length, and that it’s probably very difficult to filter out all of the ‘noise’ in the data?! But I’d be very interested to know more.
I think it would be impractical to develop a secondary model to generate signature changes based on HR. Too much chatter and static to separate out the influencers.
You can probably come up with a dozen reasons why HR will vary on any particular day.
That’s why power is king in this regard.
If you want to confirm signature changes during Base or extended periods of low intensity you can always do that with a BT workout in slope mode. This one takes about 30 minutes to complete with total time above LTP less than 10 minutes – (3) Xert - Workout Designer (xertonline.com)
Yup, sure enough TP and LTP will rise over time with nothing other than LSD activities.
So you can confirm your signature during long periods of low intensity with an occasional BT workout (a warning appears when sig becomes stale) or use the No Decay method (TL matched) as @jw66 mentions.
Amazing discussion - I like the points raised here. As @JLC pointed out, some of the reason that HR takes a ‘backseat’ to Power is that the Raw HR data is less reliable than Power data, which (given proper calibration) is consistent from day to day, week to week, etc.
Another thing that I’d like to point out is that Xert isn’t trying to be the next software platform to provide 100 metrics and let you pick through & sort through what’s important. Instead, we’ve worked on presenting the main/essential metrics needed for users to track their training and their fitness (mainly TL and fitness signatures, respectively). We could build in much of the HR-decoupling, efficiency factors, etc., but that takes away from resources that we can utilize more effectively elsewhere.
Perhaps a ‘Pro’ subscription would be something to consider in the future, for deeper analysis of smO2 data, core temp, HR decoupling, HRV, etc. as those are all certainly interesting, but are still fairly niche analyses for a large portion of Xerters.
