More space for autoregulation/self adapting training

[Schroedingberg]

Is your feature request related to a problem? Please describe.

It seems I do have to decide pretty early on about rep ranges. That doesnt fit my style of training at all - because I don't care about rep ranges as long as I am between 5 and 30 reps. The exact number may vary by about 2, depending on the daily form.

Describe the solution you'd like

tl;dr : I want a lot of the configuration to be autoregulating

What I am used to from the original : I dont need to decide before I begin a meso what rep range I am going to use. I just have the meso that I created from a template and start my training.

I log my weight and reps for each exercise, which will tell the app what my 3RIR performance is.
It can then build the progression based on that. Only after this information is available it will actually start prescribing a more narrow rep range - but the great part is, that I dont have to think about that beforehand, I can just let my muscles decide.

Describe alternatives you've considered

In https://github.com/aaronmeinel/muscleapi I have prototyped that idea a bit (its mostly tests and a core model, not at all mature enough to use).

The modularity I am trying to go for over there might be an inspiration for this project (?).
I haven't been able to evaluate yet whether MyFit is that modular and easily extensible, since I have no experience with TypeScript projects - so if my implication that it's not seems arrogant, I can assure you it's not. I just have no clue :D

[WhyAsh5114]

Ah don't worry about it. The project is not super modular per se. It's flexible and can be modified easily.

For auto regulating rep ranges, we'll have to make assumptions. Like a barbell bench press, it would typically be 5-10 reps, a lower rep range as it's usually more of a strength exercise.

But some people like to do it hypertrophy style, 10-20 reps maybe? How can we make those assumptions reliably for a user? Maybe we can use their training history or the first few weeks to get an estimate?

This does sound like it would remove a lot of the laborious setup work, we just need to formalise the logic on this a little.

[Schroedingberg]

I don't think any complicated logic is required for that. The rp app just lets me enter whatever in the first workout and then holds me to the standard I've set there. The only assumption it makes initially is that I should stay in the 5-30 range (and asks for confirmation of I don't). After that first workout it requires me to stay within about +-2 reps of the target I've initially set only in terms of what predictions it makes. So, say I have done 10 squats at 100 kg, for the next workout it will prescribe at 10 reps 102.5 kg. I can change that weight now, e.g set it back to a 100 kg (maybe my equipment maxed out there or something). Then it will try to increase reps, if that's not too extreme of a change (that's actually a bit of a harder question: do we go to 11 in that case? That's a 10% increase!). If change the weight much more, at some point it will fall back to showing the target rir. (Actually, below the set card, it even shows the weight range it can make predictions in). In a nutshell: you basically don't need any rep range info beforehand. The inconvenience that might arise from removing it can be incrementally reduced via more intelligent progression algorithms and inline feedback

…

On Sat, Oct 25, 2025, 10:02 PM Yash Kolekar ***@***.***> wrote: *WhyAsh5114* left a comment (WhyAsh5114/MyFit#232) <#232 (comment)> Ah don't worry about it. The project is not super modular per se. It's flexible and can be modified easily. For auto regulating rep ranges, we'll have to make assumptions. Like a barbell bench press, it would typically be 5-10 reps, a lower rep range as it's usually more of a strength exercise. But some people like to do it hypertrophy style, 10-20 reps maybe? How can we make those assumptions reliably for a user? Maybe we can use their training history or the first few weeks to get an estimate? This does sound like it would remove a lot of the laborious setup work, we just need to formalise the logic on this a little. — Reply to this email directly, view it on GitHub <#232 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADMZFCMQM76RL3ZYDE5H5XT3ZPJMJAVCNFSM6AAAAACKGNNVVWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTINBXG4ZTINRQG4> . You are receiving this because you authored the thread.Message ID: ***@***.***>