🎉 🎉 🎉 Introducing Adaptive Training! 🎉 🎉 🎉

So I’m wondering if the AI will penalize me for bombing the cool downs at the end.

2 Likes

I dont have the time to read through the 1366 posts so I went to TR home page and there I see a beta download and I thought everyone is in so I downloaded it. Now I have (2) TR loaded. Trying to find out if I have AT or not is not very easy. So Im looking through the 1366 posts and it seems like I dont have it. But I have a beta marked icon on my desktop? Should not be this difficult.

1 Like

You don’t have it. you have to sign up to be out in the waiting list which may take awhile. Be patient, I’m sure there will be way more threads when more people get it

1 Like

Just have to mention…

How cool is it for the CEO of a company to come on the forum regularly to answer questions, criticisms and expand on limitations/goals of the product. It’s not everywhere that happens, heck…maybe not anywhere.

42 Likes

Beta :+1:

2 Likes

It would be interesting to see if, for example, there are certain athletes where a LV plan plus 2 endurance rides a week results in higher FTP (or other relevant performance measure) than a MV plan. Maybe some masters athletes may fall in this category?

1 Like

How is that different to what Strava segments do? I don’t think I understand what you mean but I am intrigued because I’m frequent user of a limited number of routes

Isn’t that just a polarized plan?

Not necessarily. You have to look at the intensity of each of the three “hard” workouts that make up most LV plans. They can range from Anaerobic to VO2 Max (which are in POL Z3) down to SS and Threshold (that are below or at the edge of POL Z3) and sometimes regularly Tempo (POL Z2).

Following a typical POL plan will avoid the Tempo and SS, and may limit the Threshold workout too. These can get used for specificity training, but aren’t part of the stereotypes POL plan.

Doing a LV plan plus endurance rides can be POL on some cases, but also Pyramidal in other cases. It’s not cut and dry until you look at the specific plans of interest (Base vs Build vs Specialty).

4 Likes

Probably pyramidal.

2 Likes

Adaptive Training: The Right Workout. Every Time. - TrainerRoad

" Register for Updates and Priority Access to the Closed Beta" Seems pretty clear to me.

1 Like

Here’s hoping that trainerroad starts off with users across the spectrum of age, sex, fitness etc. I know many in my age group have struggled greatly with burnout on the previous programs.

My thinking is that if you took a course and a time/avg speed you could essentially covert the elevation changes into a series of intervals with power targets.

Really simplistically, my 3 hours Sunday bunch ride is a lot of sweetspot, with two long sustained threshold intervals, 2 VO2 max intervals (3-4 min), and ~6 short VO2 max intervals. There are recoveries between the climbs.

If you make the ride 3:30 it becomes a lot of endurance with a few tempo intervals.

I’mm sure this is more complex than it is to write about, but quite easy to imagine taking a FIT file and breaking the ride down to intervals with power targets. You could then compare your performance against prior rides… If your activity includes 4 x 4min at VO2Max then AT could give you credit for that progression level…

Obviously thinking on the fly :smiley:

:joy::joy::joy: apologies for the confusion- and yes being funny about three quarters of a percent difference between two power meters would be pretty anal :sweat_smile::sweat_smile:

3 Likes

@JesseFortson / @SeanHurley : Will the workout categories “Productive”, “Breakthrough” etc. lock and reflect you FTP and fitness at that specific time (when you look back in your plan), on will they always reflect your current fitness?

1 Like

I’m kind of laughing at reading about FTP again, as personally I’m now so much more interested to see my fine-grained relative fitness, to see the small steps in each category. FTP just became a whole lot less interesting to me, and that feel like a very good thing.

I do understand FTP is still a global gauge, and therefore very very telling. And that a FTP increase/decrease immediately corresponds with a total rescaling of all little bars. But the fine-grained energy system progression will be so much fun/motivating that FTP as a single number becomes less relevant, and I think that will be very good. You could even stabilize your FTP (which tends to happen at some point) and keep playing with the energy systems. The new approach has so much potential for fun and inspiration.

1 Like

:dizzy_face:

Could someone of the beta group tell me what happens when I shorten the breaks between sets, will it register as pass, super pass or fail?

The cynic in me thinks this was a bit of a stunt to hush the Dylan Thomas video

That would be amazing. A multi-year effort, including marketing campaign, getting the AT 1 min video with actors done, get the website set up, all timed precisely because one YT video comes out. Now that’s planning :stuck_out_tongue:

Actually, the scale of machine learning and adaptive training is so much beyond the small SS vs Pol discussion I am wiling to bet the opposite. Dylan’s video came out, and the TR team was just laughing. “You’re wondering about this kind of stuff? Wait till you see what we come out with!” The scale is supernova up close vs a sunny day.

Of course SS vs Pol is relevant, and with the excellent team, huge dataset and the machine learning I’m confident that TR will crush it on a level beyond the small population studies we have now. Good times, amazing times are coming :slight_smile:

I’m very curious about detecting / preventing too much fatigue, I suspect that a lot can be gained there. To recognize if someone should train hard 2 weeks, take 1 week break, or can even do 4 or 5 weeks before a break week. Recognizing that on individual level with machine learning would be great. Can’t wait to feed my HR and HRV data to the system.

12 Likes

It also raises an interesting question for me about how I might use that information to ‘maintain’ fitness when I have less time available, or how ML might learn what is required (for me) to maintain fitness in a particular energy system(s).