Tell me I'm wrong about Adaptive Training

TrainerRoad Software
25

I don’t mean to be harsh, but saying over and over that “My workouts were too easy but I didn’t do anything to correct that and I don’t hold myself responsible for that” is an odd angle.

When I first started doing AT, my workouts were too easy. The TR folks told me to be patient and eventually the system would figure it out, but I didn’t want to wait, so I just did several “Stretch” workouts until my levels felt correct. Since then, AT has worked almost flawlessly. I believe in the system and the vision, but it’s early days. It’s going to take time and will never be perfect, so I adjusted.

26

Exactly this. And before reading the 4-5 replies after this post I was gonna say that the issue is that people are expecting too much from AT but as it was mentionned, I feel like TR over hyped AT and now people got unrealistic expectations.

27

I hold myself completely responsible! I just think it’s a mistake that has been fed by the structure of AT.

I was thinking about it wrongly. There is a reason I was thinking about it that way. I don’t know what good it is blaming anyone really - I was just asking a question.

28

I firmly agree it’s been oversold a bit in the print media. If you listen to the podcasts (and of course not everyone does), they frequently point out that it’s in beta and those are the long term goals. But if all you read are the press releases, I absolutely understand the disappointment.

I have one knee that is crap. I can do short hard intervals, but I’ve learned that I underperform ramp tests to protect the knee. The same with with the super long 16 or 20 min intervals in the polarized plans. My lungs and heart can take it. My knee can’t. It actually took the Pol plans for me to realize that was my weakness. Now I’ve learned to back those long intervals down or change the workout, but the reality is I’m guessing when it comes to FTP anymore. I can’t wait for the day tests go away!

Having said that, I accept that it’s still basically beta and it will be a long time before it’s “perfect”. I love the levels, AT, and TrainNow though, and I’m willing to wait.

30

Do your preferred purveyors of coaching and software not have their own forum(s) you can utilize? I’ve never seen someone spend so much effort to degrade a product they do not utilize.

31

Because people overhyped AT (I am not saying TR but users) and decided it’s something extremely amazing and now it change everything along with basic physiology. AT is good idea and pretty basic algorithm that do for you something you have to do yourself before. And it works… or not. When I looked what TR wants me to do in the base phase I am always amazed how bad the plans become because of AT. Why? Because I can do 2x40@90% or 3x30@90%. So TR decided I have to do this 4 time a week for 5 weeks straight because AT decided my sst is lvl 10. It’s contradition to every training research and fining but hey, AT said so.

The issue is that AT works for many people, because moves them through simple progressive overload. But it not take into account a plethora of variables nor common sense. Is it good solution for majority of the users? Yes, it’s definitely step forward for TR. Is it solution for training? Not really, because you still need to know the basics and question algorithm and the plans from time to time.

32

If you really want to avoid doing a test, and from either SS workouts or maybe something better like over/unders you think your FTP is set too low, just increase it manually a bit.
Even with AT, having an appropriately set FTP is important, but AT does give you more margin for having it wrong than before.

33

I think the same but I know marketing types prefer to overhype …they somehow try to explain away the underdelivery.

34

There’s no functionality in AT right now to ‘test less’. That’s the long term goal but they need to be clearer with users that you still need to re-test periodically. Levels can’t do it alone for the reasons you say here.

35

They could test less by modeling ftp but they would have to insert max efforts into workouts to feed the model.

36

That sounds strange, at least it does on my plan. I am using a LV plan and AT and I still have a ramp test scheduled every 5 weeks?

37

The only problem that AT solves at the moment is that it makes every workout “doable” (with a range of easily doable, to hard). If you look at the forum, all the threads about “vo2max is too easy” and “over-unders are too hard” have dissappeared. Before AT, you had to make those adjustments yourself out of you own judgement of your ability, but many people didn’t.

AT doesn’t solve any of the other training questions - how much should you train, what should you train for a given goal (plan builder is supposed to do that), and also it doesn’t know if your ftp is set correctly (though that might be the next thing it will be able to do, given that other software can fairly easily say “new ftp detected”).

38

It seems that people are treating AT a bit like drivers treat Sat-Nav, i.e. turn brain off and no personal thinking/input!

Nate stated in one of the podcasts, possibly the one that introduced AT, that the ultimate aim was to remove the need for FTP tests and people have misinterpreted that to mean they don’t need to do a test now. That’s been compounded by others focussing on that and ignoring those pointing out the truth.

Advertising overhyping things? Who’d have thunk it? As my dad used to say: ignore everything you hear and half of what you see and you’ll be somewhere near the truth

39

You don’t need to do an assessment unless you have no idea / or no interest in guiding your training.

AT does not need FTP to make you fitter. It uses every workout you do as an assessment. If you have low PL numbers, then you are not really doing the workouts in the energy system that is claimed. Or, maybe more precisely, it is on the fringes of the energy systems. Same for if you have high PL numbers.

If low PL number workouts are hard in the areas you care about, I would manually bump FTP input down. If high PL number workouts are moderate, I would manually bump up a couple percent. If they are easy, I’d do that and pick some stretch workouts following to recalibrate the system.

In a rate of rapid fitness gains and especially in the absence of knowledge of your abilities, ramp is a great idea. Once you get in that shallow rate of fitness increase, error in the ramp test starts to overlap with fitness increases and can be counter-productive. It’s very probable to get a static ramp score or even slightly decreased ramp score after a productive training block. Objectively that’s not gonna matter much unless you’re sitting at the max range of PL for energy systems you care about. But mentally that’s a hard pill to swallow unless you understand the test is limited. And once you understand that, well, it doesn’t make as much sense to do it. Similar things can be said if you over score your actual FTP, but AT kinda fixes that by telling us we have a great FTP but giving us workouts that we can manage even though we don’t.

Related to all that, finding threshold level 5 workouts “hard” ( I.e., not very hard or all out) means one is in that range where your FTP input is fairly close to accurate. For a person hyper focused on FTP, that may not be true - that person very well have an FTP that they can hold for longer than 40 min, or they just have a really precise assessment for it because they ride there a lot. But for the average Joe, it’s likely close enough. For me, that means I’m level 10+ in everything lower than threshold. So most production in fitness comes from really long SS intervals, plus threshold and VO2 labeled workouts.

If I raised my FTP input setting, most of my fitness gains (that I wish to target) would come from a wider sweep of the sweetspot catalog of workouts, which will be hitting that threshold end of the zone more, and threshold, which will be trending towards VO2 more. The VO2 labeled workouts will be the easier versions of the short on/offs that are mentally boosting, but I am a bit skeptical do much for increasing my fitness.

In either scenario I’m still targeting the same energy systems, they’re just being pulled from differently tagged workout bins. There are no hard boundaries on any of these things and in the end, the only thing that matters is overloading your system with proper recovery. And AT is a nicely designed approach for systematic progression towards that if you don’t worry too much about how the energy systems are labeled.

40

I found your post interesting. I am putting this question out there more than asking it to you, but it relates to your point.

If I have a large enough amount of HR and power data, would it be possible to plot the HR at a given sweet spot interval and estimate changes in FTP?

My reasoning is that if you are close enough to FTP, but not over it, no algorithm will estimate a higher FTP, so you’ll always be stuck there unless you test. But if there are a number of sweet spot intervals you completed in past workouts, close enough in time to the FTP tests to conclude these are really sweet spot for you, there will be data that could be analyzed to be able to say your FTP improved, without having a higher watt interval completed.

Say you tested your FTP at 300, and there is a 5 minute sweet spot interval at 270w that you complete at 160 bpm. Maybe you have half a dozen such workouts, in the past, with other FTP values, but 90% of FTP always falls in the 155/160 bpm for you. If three months later you complete that same sub-threshold interval at 150 , it means you will surely get a higher FTP if you test, right?

42

Yes, intervals.icu has some nice HR variables. Trainerroad seems not to take HR into account as much. I know that it is slightly unreliable data but if you can do away with some of the noise, it should inform you that something was way too easy and propose a reassessment of power zones until HR data looks as it used to. Adaptive training should do this for you automatically, IMO

43

Fitness has likely increased in this case. But an assessment may or may not reveal that increase for a variety of reasons. So, no, it is not a given you get a higher Test score, even if you have a higher FTP. One of the reasons one should weight a body of data higher than any singular test - at least for average folks without expensive testing equipment and well controlled objective test protocols.

44

I’m not sure that HR is reliable enough. Every modeling software uses max efforts to draw your power duration curve. Max efforts are tests. They can be done informally during workouts and you don’t need that many of them.

45

I’m still getting a ramp test scheduled at the beginning of each block.

Sounds to me like one of yours has somehow got deleted from the plan.

Probably worth checking you have ramp test at the start of each block going forward.

If not something is amiss and you might want to contact support

47

Same. I’m fact, I’ve had a few cases where I’ve used Plan Builder and ended up with ramp tests 2 weeks apart. :rofl: