Target Strain 2.0
planned (soon)
J
Jonathan
Target strain is right now mostly based on recovery (and previous attained strain in relation to recovery, right)? Bevel also includes TSB exertion, but right now does not do anything with it. I know that TSB and the recovery score should in theory converge, yet they often don't quite, at least for me. At the same time, when I'm having a good run and feel good, Bevel will often push me to extremely high strains, even though TSB would advise to slow down. And I personally feel like Bevel sometimes wants to push me to intensely, where I would actually need a rest day.
So I wonder if you it maybe would be good to just display the target strain in its current form and what the TSB model would recommend side by side, so that it is possible to choose which guidance feels right each day? Not sure if this is a good solution either, but in essence I would like the TSB model to be more prominent.
Also, why using TSB when Banister is experimentally better validated or Busso which accounts for training monotony?
I btw agree with your reasoning that exertion should be unlimited... but I do think it is confusing to speak of percent in that case. So maybe just a score? Or change to an exponential asymptotic function instead of what it is now (a log?).
R
Raven
I also think one problem with Strain is it is doing two things. It's a target but each activity also includes a score.
Here's a screenshot of two parts of Superset right after I finished morning workouts a few days ago. I reached 58% and my target was 60%-70% (I think based on the hash area). Now one might think the four activities add up to that 58%, but they don't directly: 10+8+7+27+12 = 64
The way I've been interpreted this is there's what I'll call "Strain Score Points (SSP)" which is different from my "Strain Percentage". In this case, it took 64 SSP to get to 58% (I started my activities shortly after coffee and don't think I had registered any strain prior).
The way things are now, I can make some sense of my workouts -- the rowing contributed more than Kickboxing, for example, and I was just outside the target (but as the day went on, ambient activity got me into range).
However, I have no idea what "100% strain" means. I also know the unlike Sleep and Recovery, where I did fine but could have done better -- 100% on these two is the goal, right? But 100% is almost never the goal for Strain.
Grey
Raven: Hmm that's a good point. Let me put some more thoughts into this and figure out what's the best UX to communicate this
R
Raven
Grey: My guess right now is “100% strain” is the amount I would be recommended
if
I had 100% recovery and 100% sleep? In other words, if I’m in perfect condition then I’d try to go for 100% strain, and I could go beyond, but that wouldn’t be effective.If that’s right or even close, then it sort of makes sense, except (for me) that’s a rare condition to happen.
So I think one of these makes more sense:
1) “100% strain” switches to mean maximum safe strain recommended for day. This should perhaps be accompanied by a “target SSP” using my nomenclature above (which I don’t love, just wanted an easy way to describe the number given to an activity, s o come up with better name, haha)
2) Just remove the percentage sign and use a target SSP number. You’d then have two values that use percentage: Recovery and Sleep, and two values that use a straight number: Strain and Stress. Stress is always a 0-100 scale while Strain is a 0-unlimited value.
With (2), you could have a 2x2 grid at the top instead of separating Stress?
Hope some of this scattered thoughts is useful. All the best.
M
Markus Eisen
Also the coaching texts should include those factors
R
Raven
Possibly its own topic, but one thing I've noticed with "whoop style" systems is that the usage of rings/circles differs from how the Apple Watch uses it for move, exercise, and stand. The idea of "closing your rings" is one the watch encourages. Apple does have awards for getting 200%, 300% and 400% of one's move goal, making it so one can have a system of doing 100% on a recovery day, and maybe 400% on a marathon day or the equivalent.
In comparison, one might not even want to close rings in the Strain area of Superset. With the Sleep category in Superset, closing to 100% is the ideal. I don't think the concept of 105% sleep is possible? Likewise with Recovery; one might rarely get to 100%, but it's always the goal, and mimics the watch's own rings.
It's be nice to have an idea of what "100% strain" actually means, and if Superset is recommending less than that (would it ever recommend more?), what the recommendation means in relation to the 100%. I think there's a natural idea to want to close the rings, or exceed that, so perhaps it's better to have "100%" mean "100% of today's recommendation, and a different number to represent the "absolute strain" -- is there a unit for strain? In cycling, I might have a kilojoule target I want for a session.
In this screenshot, I went beyond my recommended strain. Oops. I'd say the 57% isn't as important as being, what, 110%-120% over the day's target? If the lower part of today's target was "100%" (the goal to meet), and the upper limit "110%" (what's allowed at the higher end of the target range), then right now I have weird incentives; one can easily think they still have "43% to go" just to "close rings".
M
Markus Eisen
Raven: there is a coach in tab which explicitly tells you what to do and not close your ring to 100 🤷♂️
R
Raven
Markus Eisen: I’m on the free version; coaching is restricted to pro — i can see the target within the hatched area though. My comments were meant to explain what an Apple Watch user unfamiliar with whoop might expect — i.e. close all your rings.
M
Markus Eisen
Raven: ok the coaching tells you what you should do with that data
R
Raven
Markus Eisen: does coaching explain what “100%” strain represents and how it differs from 99% or 101%? Can you get 101% sleep or recovery?
Grey
Raven: Thanks for the feedback. I like the idea of treating 100% as a target for the day. However, it will make it harder to compare across different days, but I see what you mean. I'll take this into consideration as I build out Target Strain 2.0
Dylan
Grey: I personally don’t think 100% should be the target. 100% strain sounds like the maximum strain my body can take; this isn’t actually true and people will be scared to “max out” their strain all the time.
R
Raven
Dylan: You can go beyond 100% now though and 100% currently also does not represent the maximum strain. It’s unclear what 100% currently means.
Currently, if Superset tells me my target is 70%-80%, for example, that’s not “70%-80% of what I could ever possibly do,” and if one goes out for a long run/cycle session, comes back to Superset and sees “103%” all I know for sure is I’m “23% over the recommended target” but I don’t know what the “3% over 100%” means.
R
Raven
Note that Grey has stated the goal here is this: “ Make target Strain easier to understand by splitting into recommended minimum Strain and maximum safe Strain.”
What the best way to represent “maximum safe Strain”?
Dylan
Raven: what if instead of a percentage this became a number? That way it can be infinite number, but still is in a range for targets?
R
Raven
Dylan: Yeah, that might be better — see my other post I just put up in this thread where I show how my activities “strain scores” differ from “strain percentage”.
Rory Plewes
Dylan I agree, If your recovery is 50% it seems counter intuitive to go to 100% strain, even if that 100% is actually just the target strain goal (60-70%).
Especially if you then looked at month view and you hit your strain goal every day it would show 100% every day, but you then wouldn't be able to see days that you went hard, and even if you went above 100%, a 150% strain on one day isnt proportional to 150% on another day.
Grey
planned (soon)
K
Kevin Meijers
I don’t think it’s a great idea to make 100% the goal each day as it’s harder to compare the goal in context to other days which also has a lot of value. So something that would deliver the best of both worlds would be great imho. 🫡