I understand that Training Load represents the exponentially-weighted moving average of activity XSS values. Does that mean that if I want to achieve a particular Training Load of x, that I need to perform activities that generate an average XSS of x per day for a certain amount of time?

(And follow-up question – what is that “certain amount of time”? Is it possible to find out more about the equation that drives the exponentially weighted average?)

Yes, 7 days is true of the training XSS deficit in XATA, and the deficit is vs your training load + ramp rate

The TL itself is essentially the average XSS per day over a longer period, as described by @RobOnWheels . The period of the moving average depends on the energy system, and you can see that in your account profile -> Training Load constants. It’s longer for low intensity (60 days default) and shorter for higher intensities (22 days default) recognising that it takes longer to build base fitness than high intensity…

So for low intensity I think the formula would be 59/60 * previous low TL + 1/60 * low XSS for the day… similar logic for the high intensity components but with 22 instead of 60 in the denominator… then add them up…

Not only does LOW Training load take longer to build up (and decay) because the time constant is longer, but also LOW training load typically makes up the vast majority of your total training load. This is why we emphasize time and again that a proper base is an important part of your training cycle. Since high/peak strain contribute relatively quickly to increases in your high/peak TL’s, then you can add in high/peak strain later into the build/peak phases and still arrive at your TED properly prepared for the event.

Makes sense - thank you all. Because of your explanations, I finally have an intuition for why the time constant is different (it’s higher) for LOW vs HIGH/PEAK TL.

And mathematically, the answer to my original “follow-up question” seems to be: it would require 60 days generating an average XSS of x to achieve a TL of x.