Thanks.

I'm not convinced, though, that the inputs I've suggested are unscientific. The basic forms of measurement (RPE for swimming, Coggan's cycling measures, and time in zones for running correlated to time at specific mm lactate/l of blood) are scientific. And, the proposed measurement validity testing is also scientific. The entire model is unproven, but that's the point: try to fit a dynamic model with many, many users providing data overtime. Hey, your probably right in that it will be crap, but there's a demand for it, so why not play with it and use a beta app to validate and test the model, and accept constructive input on this forum for respecified nested models for comparison's sake?

it's my personal interest to see how many points of data people are willing to submit/input so that I can turn to sequential dynamic systems analysis (unique formula for each and every user - avoiding aggregatory regression), neutral network analysis (starting with end goal and identifying sequence patterns leading up to goal), and persistent homology (using Aysadi's simplistic program for data reduction and singular pathway/sequence detection). But yeah, this approach is pretty unscientific....

---

wovebike.com | Wove on instagram