Heart Rate as another input for TP

In the future, could HR data gathered alongside power data be used to better establish the fitness signature? It seems that if you had enough data from an athlete and all the data from other athletes you could start to establish correlations in HR and power. Seems like it would be really helpful in seeing changes to signature in workouts that might not otherwise have breakthrough efforts. Something that can be incorporated, or am I missing something?

YES! HR data has so many treasures buried inside it. Did you know that MPA itself came from analyzing HR data? http://baronbiosys.com/?p=977 There are so many ways in which we can use HR to help assess the fitness of the athlete and their performance during activities. From accounting for gear and cadence choices, to the effects of heat acclimation, to correlating recovery loads with HRV, to etc. etc. Our HR models and the opportunities to apply them a far-reaching.

Now, if only we had the resources to pursue them all at the moment. :slight_smile:


to bad you could not link up with the Elite HRV people to add HR variablity to your quiver of tools.

since the last response to this in 2016? has there been any change in the use of heart rate for determining metrics? One might think that heart rate recovery or decoupling might be something that would add to the way Xert could more appropriately deal with individual differences. So now in late 2020 is Xert doing anything in the algorithm with heart rate data?

I’m curious to see if there’s anything closer to release here as well. I suppose you’ve got a lot of environmental variables with HR, but adding even some additional context to training load or MPA would be cool.

One of the Xert gurus will chip in I guess, but in the meantime: if you’re not using it yet and provided you have a Strava account, go to intervals.icu for this sort of in depth analysis.

The ‘dependency’ on Strava will be removed in the near future and it already has a link with Garmin, to upload the workouts you’ve created…

We’ve released HR Derived Metrics which establishes patterns in your HR, power and cadence data to determine your Xert metrics from HR data. We have new features coming out soon that will use this algorithm to a greater extent.

In terms of using HR recovery and decoupling, there are just too many variables to draw any meaningful conclusions on a broad scale. HR data is truly a mess when you look through all various user data that gets uploaded. Developing software that can weed through this mess and extract meaningful information isn’t easy but is key to developing algorithms that will work. HRDM works because it doesn’t rely on individual data points and has some correction built in. Looking at HR recovery would be extremely difficult with the quality of data available.

HRV is very interesting and we have some unique applications in our roadmap.

Core (www.corebodytemp.com) is also a promising tech that we think we can leverage both in terms of creating methods to analyse it meaningfully for athletes as well as establishing models that can be applied more broadly, particularly as it relates to HR and central fatigue.

Unfortunately, as you hopefully can appreciate, our eyes are way bigger than our stomach when it comes to uses of data and developing new features. As we continue to attract more and more customers, we can also invest more and more into new innovations.


Hi Armando, one thing I truly appreciate with Xert is the engagement you have with the customer. I appreciate the many options one can have to enhance the way that your algorithms behave. My thought is to not look at heart rate data at a fine detail level but maybe at a higher level as a means to more individualize the depletion and recovery metrics/ constants for MPA? Also if there was a means of aligning ones decoupling effect on fatigue during a ride again this might allow some level of nuance to change the way MPA recovers over time during a longer ride? These are just thoughts and I appreciate when one looks at a lot of data it might be hard to deal with effectively since there are multifactor issues that impact these things. Again thnaks for taking the time to respond.


Hi I think that it would be useful especially for endurance rides that the HR dial in the remote player indicated your HR zones in color code, similarly to the power dial. I guess it would not be difficult to implement asking the rider his/her HR zones in the settings, at least optionally.

We don’t use Zones at all at the moment and training is exclusively power based. We have a lot of ideas for HR but it won’t likely be using Zones. Training in zones and tracking time in zones isn’t how we believe training and improvement works. For example, after a hard effort your HR could be elevated for quite some time, even if you’re pedaling lightly. That doesn’t mean you’re not riding in Zone 1 polarized for example.

1 Like

Thanks Armando. You are right. Today for example while doing 90%TP intervals my HR started to increase in the 3rd one to reach the middle of HR zone 4 whereas the previous 2 ones were Pretty flat in the HR sweetspot zone. Strangely, RPE was not significantly higher in the third one. By the way, the power target in the player decreased by 10 Watt for unknown reason…

While HR cannot be as you say a good guide for training I think that this type of information is useful to be analyzed afterwards.

Check that sweat didn’t land on the - button. Happens…

Analyzing HR data automatically for 10k/100k’s of users will be a big challenge. If all we need to do is show and put the data in Zones, that’s easy. But to do what we want to do, we’ll need quality data. There are so many HR data issues to to filter, remove, replace with most monitors on the market.

We probably won’t offer real meaningful HR data analysis (DFAa1, for example) until there are quality HRM’s on the market that are widely available and we can limit the analysis to users with these units only.

1 Like

Concerning the power target reduction I have been suggested in another post to use the trainer mode in the android player as Auto 100%. It was set to 97% for unknown reason. I will try tomorrow. Anyway I am not using any trainer, just a spinning bike with favero Assioma pedals, and I change manually the power following the prescribed target…a little bit over to tell the truth as we all cyclists do or at least try :sweat_smile:

Hi Armando, it would be cool if XERT were to use HRV as as part of the planning aspect though. Apps like HRV4training do a decent job, I think, and even they recommend certain straps/monitors as the best such as Polar H9-H10 as the gold standard it seems. But just looking at rMSSD over time the trend could help make which type of training effort is recommended somewhat more meaningful?