How training load estimation from HR works

This one is 130 now and the marathon is 216.

Much better! Thanks :slight_smile: I’ll keep you posted if I notice something off

1 Like

David.

Post the Push this has reduced the TSS for all non-cycling activites significantly. However, this still mucks up the ‘Time In Zone’ calculation for those with a broad body of non-cycling workouts (everything is base because time in zone is not weighted by stress e.t.c). Is there still a plan to allow us to only count cardio activities towards cardio stress calculation and times in zone?

1 Like

Yes there is. I hope to get on that soon. Planning to have a fitness factor and fatigue factor for each activity type so you can make strength work in the gym count towards fatigue but not fitness and so on. I hadn’t thought about leaving some of the out of “time in zones” but that make sense.

1 Like

I could do that if the fitness factor is zero … if its not improving fitness makes sense to ignore the HR data for “time in zones”.

1 Like

Aye - the reality is that whilst weightlifting (ignoring here Metcon or crossfit-style cardio workouts) causes heart rate rises, this is due to thoracic pressure rather than aerobic fitness pressure.

It’s hard to quantify any exercise that isn’t cardiovascular in nature using heart rate data. I’ve heard of including lifting etc in CTL - but the problem with this is that it impacts upon your form (aka your tiredness and capacity to conduct workouts). Whilst lifting will impact your muscular ability to complete workouts, it will have little to no bearing on how well prepared your cardiovascular system is for conducting difficult workouts, races e.t.c.

Probably best to allow all variable sport types to be given a custom factor but my default would be only cycling, runnning, crosstraining, ski touring, xc skiing e.t.c.

2 Likes

I have started implementing fitness and fatigue factors for each of the activity types supported by Strava (and fitness factor of 0.0 means leave out of time in zones as well). This will be configurable per athlete but I would like to have good default list. Note that this only apply to activities with training load derived from heart rate data.

I have all of these at fitness 1.0 and fatigue 1.0 i.e. full training load counts:

Ride, Run, Swim, Hike, Walk, AlpineSki, BackcountrySki Canoeing Crossfit, EBikeRide, Elliptical, Handcycle, IceSkate, InlineSkate, Kayaking, Kitesurf, NordicSki, RockClimbing, RollerSki, Rowing, Snowboard, Snowshoe, StairStepper, Surfing, Velomobile, VirtualRide, VirtualRun, Wheelchair, Windsurf

These only add to fatigue:

WeightTraining fitness 0.0 fatigue 1.0
Workout fitness 0.0 fatigue 1.0

These don’t add to fitness or fatigue:

StandUpPaddling fitness 0.0 fatigue 0.0
Yoga fitness 0.0 fatigue 0.0

Any suggestions or changes?

1 Like

Maybe the value of fatigue also depends of the duration? Something easy in math 30 minutes 0.5 Fatigue, 45 minutes 0.75 and 1 hour 1.0 fatigue

The factors are applied to the training load for the activity which is derived from the duration (among other things) so the duration is already accounted for.

@david - My HR is a fair bit higher for indoor sessions (I’m sure that’s pretty common). Obviously this doesn’t matter for the actual TSS score from those indoor sessions, because I always have power.

But could the indoor sessions be excluded from the model? They might already be, of course!

Hmm that would help a bit with time in zones and average HR models but it would also make the data sparser. I will have a look and see. Tx.

It would be sparser - but I do think more accurate (at least for me).

My outdoor rides estimated v actual TSS are pretty close, but last week’s indoor rides estimates were significantly higher.

I have the same issue, but it’s because my indoor FTP is really about 30 points lower, whether for biomechanical reasons or a difference in the power meters. I use the time-in-HR-zone estimate on my commuting and goofing around bike and would like to be able to exclude indoor data. It would also be nice to use only fairly recent data for the estimate because of seasonal fluctuations in fitness.

I have added this to the todo list.

Hi @david, does one of these options match the training load calculation performed by Training Peaks?
I can see quite a difference right now between the two platforms for outdoor activities, where I do not have a PM.

I haven’t seen anything published as to how TP does it and I am not a TP user myself so I am not sure. Maybe HRSS if your resting HR and max HR are correct?

I’ve found this article that says that “hrTSS will only be used if there is not enough data to calculate TSS, rTSS, or sTSS” so you are probably right.
The two metrics differ because the default for cycling here is Time in HR Zones and not HRSS.

What’s your logic for making StandUpPaddling fitness 0.0 fatigue 0.0?

I have only seen people doing it and didn’t look like much cardio … but open to suggestions from actual standup paddling people :slight_smile:

1 Like

I am using Ice Skating as my activity for speed skating. Would you assume that a factor > 0% for Fitness apply to this type of activity? What % would you then to apply? Attached is a view of my latest skate.

Thanks,