More evidence in the use of alpha 1 (metric based on hrv data) to make sure you’re in the right zone. Any chance if getting tr to record r-r data to make use of this? (Not to replace the ai calculations but to improve them)
Relationship of Cycling Power and Non-Linear Heart Rate Variability from Everyday Workout Data: Potential for Intensity Zone Estimation and Monitoring
The reason why I want more straps to record better quality r-r data:
And yes, this idea but not this study was posted about in the past:
This is not just about zones and knowing the border of the zones. To me I see tr using the correlation between power and alpha 1 as just one more data point that influence the trends the ai picks up.
When alpha 1 is low relative to the power you put out your body is less able to put it power for whatever reason (fatigue, over training, lost shape, etc) and the reverse is also true. This doesn’t require using the data for zone demarcation
This is a direction I support, and actively sought out solutions a few years ago.
My Garmin 530 and 840 have been using and recording HRV during all rides. Garmin uses power, hr, and hrv to provide a real-time assessment (“Performance Condition”) of ability to perform relative to my average fitness level. Using ML.
For myself if Performance Condition is trending positive over a week or two, it’s a strong signal that fitness is increasing. One of several signals I’ll look at.
I’ve gone back and used free Alpha1 analysis tools on the Garmin recorded HRV data. To be honest it’s pretty noisy even on erg-like outside rides. But easy to see less noisy segments that a computer can remove (I think FirstBeat/Garmin have patents in this area).
Not a Garmin watch owner, and using Apple Watch with 15-min HRV to feed Athlytic app. Athlytic uses ML to trend sleep, exertion, HRV and other metrics for recovery and training recommendations. It’s pretty darn good at basic recommendations
just from getting Apple Watch and Health feeds of workouts and all-day HR / HRV and other biometrics (SpO2, respiration, wrist temp).
Again, basic recommendations that I may take into account if not feeling it on the bike after a proper warmup.
Garmin Edge plus Apple Watch plus Athlytic for $30/year works really well, but requires the user to make training decisions. I’m happy to self-coach, so for now I’m not looking for alternatives.
When TR gets to this point, or close to it, I’ll give TR another look. The new multi-sport import is an important first step.
Would be interesting if the data could be used to tweak a workout that you already started. Thinking erg workouts (very well defined and repeatable power outputs) the software should be able to predict what it expects your alpha 1 should be (an average of alpha 1 during certain parts of your workout maybe). This way it can tweak what the workout intensity is set to. This way you could get a slightly harder or easier workout which may help you progress better.
Sure, those who are good at listening to your body may be able to do better on their own but not everyone can do that.
Garmin gives you a real-time performance condition based on power, heart rate, and HRV.
And fwiw, imho it’s far better to learn to ride endurance by feel. If I feel truly awful, and look at performance condition on my bike computer, it’s between -5 and -10. But that only happens maybe once or twice a year. And there are times I’ll come home from a good workout and see -5 after the ride - in these situations I’ll review the last week or so to see if there are any trends.
I’ve used Alpha1 HRV on my Garmin for a year, and it’s only accurate during very steady efforts, and I really only find it useful during a longer ramp test where you are trying to test for zones. Virtually pointless in a group ride, with short intervals, on a ride with lots of undulation and inconsistent pedaling, stops, etc, as it is trying to look at how stochastic the HR is, and unsteady efforts can screw with the algorithm.
A better tool for tracking intensity zones on more typical day to day rides might be a NIRS device like Moxy monitor. It also takes a bit of black magic to interpret, but I find it to be interesting to use on interval days and group rides, testing, etc.
I still thinking an even more interesting and useful data point that many don’t talk about is respiration rate, which can be extracted fairly accurately from HRV data using the Alpha1 HRV software on your garmin (it appears more accurate than garmins own interpreted data)
That said, power and heart rate is good enough, and I don’t see much benefit to any of the other tools or measurements in the real world.
Getting in touch with your breathing is underrated. At least it was in my case. Props to Coach Tim Cusick and JoinBaseCamp for bringing that into my training. I’ve made some noticeable small gains by working on my breathing - without any questionable estimates being displayed on my Garmin.
So happy with those small but noticeable breathing gains I bought a book (The Breathing Cure by Patrick McKeown) and bought the (currently on sale) StrongFirst SECOND WIND Express online course (StrongFirst: Training Center)
I agree that its likely a better metric if you have it. But better metric in general? Many people already have HR straps and a HR strap that can collect good quality r-r timing data is under $100 making Alpha 1 usable by way more people. Since it is just a calculation this means the burden on the user is much less to collect this data. So for example, if TR wanted to use this data all the work is on the back end, users don’t need to do anything other than using a HR sensor that gives good data.
Lots of TR is long steady efforts in erg mode.
Power and heart rate are the low hanging fruit that have been around for many decades. No one is saying they aren’t the most important metric. The advantage of analysis through a computer is metrics that have a smaller influence on understanding your performance can be easily done this metric doesn’t have to always strongly influence the analysis as it may just go along with the other data. But the metric can be significantly more helpful when it shows something is off (different from what power and HR data suggest) to help better
I agree with those points. But for me the end goal is to get faster, and I do not believe that any of these devices or metrics outside of HR or power have made a measurable difference in my training or approach to training, or my endurance or power.
The one I didn’t mention is the use of a CGM, which IMO is on par with a power meter in terms of providing useful data. CGM is to diet and fueling as what a power meter is to training. I’d 100% recommend 2-3 months or wearing a CGM to test out foods, fueling, etc if you’re serious about getting your fueling perfected. If the pro ranks aren’t using it to dial their nutrition strategies during training I’d be very surprised.
If simple power and HR metrics is all you need to get faster, great. But thats not everyone. Not everyone recovers at the same rate or has the same stress, and lots of other possibilities of things that. There are times I could do slightly harder or maybe need things slightly easier. Sure, you can always tweak what adaptive training says to do, but what if adaptive training could be better? That would be better for more than just the lazy. Not everyone is good at listening to their body. But even if you could, how do you know how to optimally tweak it. What if you over shoot or undershoot the effort you should be targeting? You could either be leaving things on the table or require more recovery then optimal. Yes, adaptive training can make up for that if the future workout is too hard or too easy but if a correction can be done a bit earlier that would be better.
CGM may be great to tweak your workouts, but the number of people who have that data is tiny. The point of this feature request is that the number of users who can provide this data should be pretty large so could optimize the training of a large number of TR users and the users won’t have to do anything. (though TR should probably indicate what heart rate straps provide good data to enable this to work for them)
And I’ve published my stats and you know what, building recovery into your workouts IS optimizing IMHO. Fewer intervals, more often, dialed in less exhaustive HIIT, done 3 times a week.
Now all you young peeps are saying “WindWarrior you are old as the hills, not even close to being elite, and never averaged more than 8 hours/week for a year. WTF?”
And I’ll tell you it is like compounding interest, make small deposits, reliably average 4-5 days a week, intervals done to perfection each time, and fitness will surprisingly compound even if YOU ARE FIGHTING AGING DECLINES and getting old as the hills. Sure, it helps my training age is 9. And surprise surprise, without any killer vo2max work (ahem, Friel’s Fast After 50), after a couple of years I raised my absolute VO2max (estimated) and FTP to levels previously seen when I was all-the-time riding threshold and high-intensity, and taking recovering when needed. Power, heart rate, and the common sense to not chase marginal “optimizations” that may in fact hurt more than help.
I’m also very exited by all the insights that might come out of HRV data. However, I think there is one issue to solve that doesn’t look too easy for me: The quality of the HRV data depends quite a bit on the device. Polar H10 and Apple Watches eg. probably deliver HRV data that you can use very well for deriving insights, while I’m not so positive for other HR devices that measure on your whrist. Guess that’s a tough one to solve in the back end.
