Agree! I only know a little about machine learning, but this doesn’t seem like a reasonable application of actual machine learning.
I would suspect that they are doing a regression analysis, not machine learning.
2 Likes
But who wants to do a ramp test every month?! I’d rather just bump my intensity up and test less frequently.
2 Likes
I don’t input 100% of my physical activity. I don’t have a power meter or bike computer for outdoor rides. I don’t wear a fitness tracker of any sort. I also do things like rock climbing; no idea how that could make its way into the model.
It’s nice to say that there could be lots of data like Apple Watch and Whoop and outdoor ride/run tracking, but most people have some subset of these gadgets. Or, suppose I tell the app that I do have the gadgets, but I forget to wear one during a run? The model has to be able to do reasonable things when we give it bad or incomplete data. It’s not easy. I’m not saying it’s impossible, but given the size of the market and the cost of machine learning specialists, I don’t expect magic.
I said earlier that there are simpler ways to achieve some amount of personalization without re-testing and without risking unexplainable results. I still think that’s the reasonable way to go. Maybe that’s actually what TR is doing, except they attached the magic buzzword.
2 Likes
I would very much use this feature. When I’m in the midst of mostly running, I like to fill in some cross-training. Adaptive training fits the bill for me. I don’t want to have to be my own curator for deciding which workout makes sense for me today. I’ve used it with TR and it works well. TR also now receives my running results and includes that in load calculations for projecting an adaptive workout for that day. I really don’t care to test every few weeks. Just bump my intensity!
I am nearly 100% sure that running doesn’t factor into TrainerRoad’s adaptive training or TrainNow recommendations. Runs that are imported don’t even have a TSS value (I just checked my run from this morning). The only way they are “adapted” is if you don’t check the box that you completed it. As a triathlete that has used TR and am a current subscriber, their system is a mess in my opinion.
That said, I would love a TrainNow style workout selector in SYSTM.
1 Like
With SUF (sorry SYSTM!) I’ve taken to adapting my own training i.e. manually adjusting power targets to suit my form. Repeating the Full Frontal test on a monthly or even quarterly basis is not something I want to fit into my routine for the sake of tweaking my profile. For me that’s an annual nightmare at most!
I do think there would be some merit in simple adaptation of the fixed plans based on performance feedback as you go along.
At the moment I’m using a new UK based training app called “Pillar” and they are working toward adaptive training plans based on user feedback and multiple events (similar to TR). They allow users to input their planned regular weekly rides too (e.g. club rides/races, Zwift racing etc) so the training plan can be adapted around those additional sessions. I still use SYSTM for individual training sessions (simply because I enjoy the vids), but I’m now using Pillar to build my long term training plan. I would like to see SYSTM developing more long term training plans adapted around the user’s riding and event schedule.
3 Likes
@Peteski I suspect that is the plan. From following the TR forum it seems like their solution is not quite fully baked so perhaps Wahoo is taking its time to get it right. I am not familiar with Pillar but will check it out. I don’t actually mind the periodic testing but my results seem to be affected by seasons. When it is warmer and more humid in my pain cave I don’t do as well and during the winter when I can let a bunch of frigid air into the cave I smash it. So I have been relying more on race results and practice races during the summer to guide me.
3 Likes
One of the comments in the thread mentions a lack of data to create a useful model. Surely for those that are now also using RGT, there’s potentially both HR and Power data as well as data pertaining to how well you have performed against other users (assuming they are using real age/weight values). Is it worth restarting the discussion?
1 Like