If you made the data available publicly we may be more inclined to participate.

Would knowing things like the lengths of arms/legs and individual lengths of your bone lengths relative to the joints matter? Like lower and upper leg lengths. Upper and lower arm lengths.

The entire project will be shared as a bike fit calculator (we have previously shared over 50 spreadsheets and calculators) but if you mean the raw data, the sheet currently has 3,200 cells and I am not sure it will be that helpful to anyone not obsessed by bike fit, but thanks for considering helping.

Do you have any way of validating if the fits are good or not? Seems like if you get a bunch of bad responses, you're not going to get a good model

Yes quite true, the more biometric data the better but if there are lots of qq, then the less people want to contribute (either now or in the final model)....so we are aiming for a compromise. You can actually enter arm span and footsize (optionally) in this version. If someone has typical body proportions then we can do a lot with height alone, but if someone deviates (for example tall but short arm reach) then yes, we need to model that too.

there is a thread on ST where a 50-100 or so people posted their fit numbers

the thread was title something along the lines of "post your Y/X"

that should help you fine it

---

Brian Stover USAT LII
Accelerate3 Coaching
Insta

Quite true!....our first method is validation against typical (mean) sizes (the conventional middle of the road bike fits if you like), our next method is validation against subgroups (eg casual road vs keen aero road), our third method is to validate the model against people with unusual body sizes (ie biometric outliers), our fourth method is to re-run the model in an entirely new sample attending a professional fitter who has agreed to work with us.

However there are still some challenges: people who enter seriously unusual data; people who like a seriously unusual bike fit by personal preference; people who are in a very niche group where frankly we don't have the data. This is basically why I am asking for TT/tri data.....our current sample is too small (compared to 209 who have returned road bike data).

Our aim of an automated dynamic bike fit calculator will never replace an expert bike fit, its not designed to, its designed to give an initial fit as a guide for the vast majority of cyclists who never get a quality bike fit, or simply as an initial heads up.

bw

alex

Thank you for the head's up.....I found it based on your recommendation:

https://forum.slowtwitch.com/...our%20y%20x#p5976483