Thanks. I have since read up more and watched the youtube videos on Progression levels and adaptive training. I really love where they are going with this. I agree that some flexibility and using progression levels alone could be a good thing. However I do really like the idea of machine learning based recommendations. Ultimately I am sure it will take a while for these models to be optimized and they might be biased by their own recommendations ( though they said 50% do use custom workouts not based on their recommendations so it evens it out). I think I will keep it on for the first month and do some manual tweaking whenever I think necessary using the progression levels. I guess it can never hurt to get the recommendations in the background even if you change the workout every time. But wow a few months seem to have made a huge difference. Time to see how it works out in real life though.
But I am so excited at the potential of this. Imaging in the future it can use sleep, daily heart rate variability, work hours, nutrition, daily steps, heart rate at different intervals during your workout, even consider your club workout or perhaps even non-cycling workouts and use all of those to give reasonable recommendations. It may be a while away but you can tell thats their ultimate goal… would be an interesting time indeed.
3 Likes