Something has happened to HR load estimate


The data points are now sparse and all lying below the zero error line, the resting HR is old, the estimate is way off, and the R squared is poor. It had been .97 and is now .77. This is true for all the several riides I checked.



1 Like

I changed everything to use HRSS instead of “time in zones” that you had configured. It seems that I need to bring time in zones back. HRSS has the advantage of not needing training data.

1 Like

All the HR load models are now configurable per sport. If you were previously using “time in zones” this has been restored:

1 Like

Thanks! All models are also now loading the correct resting HR.


I’ve switched my “Other” sports to use “Time in HR Zones” - but when I go to old activities the HR page for that activity has the content and calculation for HRSS, rather than Time in zones:

Activities of this type (Backcountry Ski) used to show the Time in Zones stuff.

It falls back to HRSS if it cannot calculate time in zones. You need to have some activities with power and HR for that model to work. Before the big sport settings update it would have “worked” but it’s not certain that a model derived from cycling is applicable to Backcountry Ski.

Do the HRSS numbers look ok?

Ah that is a reasonable assumption.

I do actually prefer the estimates from the time in zone model though. The HRSS estimates are always pretty low. Obviously this is just finger in the air ‘feeling’, but it feels like the time in zone model applied to the skiing gives a more realistic figure.