Agreed, with life’s pressures and job demands - I might have the time but then look at the workout and say “no way”.
Yeah see, this is the problem. When I said the ramp test does not work, I didn’t mean it didn’t work in a mathematical sense over an entire population - I meant that it does not work as a method to train by.
It can under estimate, over estimate, or it could be dead on. But you have no way of knowing - and it could change for you over time, based on training (like it did for me), or it could stay the same!
An analogy is probably the best way I can think of to illustrate what I’m trying to get at.
You could almost certainly derive an estimate for shoe size based on height, which would on average, over the entire population, give the correct result. The estimates would be off in plenty of cases, some not too bad, some where the shoe would be too tight and would cause pain, and if you persist over the long-term thinking that’s how a shoe should fit, could do even more harm. In some cases the shoe would be much too loose, which would be inefficient and slow you down. So why on earth would you use such an estimate when you can just get a ruler out an measure your foot size?
You could use the shoe-size estimator as a first guess, and then try the shoes on and go up or down a size depending on how it feels. The trouble is, what if it’s your first time wearing shoes, and you don’t have anyone around to tell you how it should feel? The analogy here being that for many of us starting with TR, this is our first time following structured training according to power - i.e. our first time wearing shoes!
Now, the problems are further compounded if you only have to wear the shoes for short trips. You can squeeze your shoes on, pop down to the shops, get home, and take them straight back off, still thinking that the pain is “normal”. The analogy here, is that the TR plans only have very short intervals at or near threshold, allowing many to overcome them through either anaerobic capacity or sheer force of will.
So measuring your feet (aka a field / TTE / KM / whatever you want to call it test) is one solution. Or you could mitigate the problem if you were forced to confront your FTP through say working up to 3x30 minutes with 5 minute rests at 100% FTP over the course of however many weeks (of course, tailoring to your TTE and working to expand that would be better, but lets just take a 60 minute TTE as a “goal” here for a stock plan and leave it at that). And of course, once you have enough experience with either of these (field tests and/or properly long FTP intervals), you could probably use the ramp test to ballpark things and then you’d be able to feel if it was wrong or not during your intervals as you’d have experience riding at your actual FTP. But honestly, at that point, maybe you’re better off just going by feel? Or just… avoiding all doubt, at the same time as getting in a solid workout, and doing the field test…
TTE (Time to Exhaustion) @ FTP can occur anywhere between 40-80 minutes (maybe slightly less, like 35, if you’re very fresh or just had a big bump in FTP). It is totally possible to ride for 45 minutes above FTP, and it is exactly as painful as you just described. Lower your target by 10W, and you can probably ride at a level that lets you go a bit longer, but instead of your legs screaming at you they’re, well, “discomfortable” is a great word for it I think.
In my experience, there’s three phases you go through, first you’re like “this is super easy, I should just dial it up” - but don’t. Then it’s like “ugh this isn’t fun, how long can I even hold this for”, and finally “okay I guess I can hold this for a fair while, I don’t really want to, I’d rather go easier, but I’m not watching ever single second tick down and/or praying for the sweet release of death so I guess I’m ok” - these three phases, for me usually happen in the first 10-20 minutes. Then eventually, after somewhere between 40-80 minutes, you sorta just run out of gas - your legs don’t explode, they just can’t hold the same watts anymore, regardless of what your brain tells them to do.
I think most people go through something similar, have a read of the Kolie Moore FTP test thread if you’re interested. But there’s also some people who, for whatever reason, will still push through and try and get as close to their ramp test FTP as they can rather than try and “feel out” their FTP - and yes, the test would be miserable for them.
In conclusion, TR developed the ramp test for a specific reason - people weren’t doing the 8 or 20 minute tests that they already had - probably because they’re long and painful. Not only that, but pacing either of those would be extremely difficult for a beginner, and I imagine not an exact science even if you’re experienced? The ramp test negates all of this, you only need a rough estimate of FTP to start with, no pacing is required, and the pain although more intense promises to be shorter lived. This makes it more attractive all around than their existing alternatives, so it makes sense they’ve made it the default. And they say they did a lot of tests, and set the FTP = 75% MAP based on the ability to complete planned workouts after it. Sure, that makes sense too - but, and this goes back to my original point, and the whole reason I risked re-opening the ramp test debate, with how short the intervals in the TR plans are, I don’t believe this was a sufficient test.
It’s a difficult problem to solve, to be honest. You do need some mental preparedness for a field test (think it’s been covered already but field test does not imply outdoors, just using that term since I think that’s the original term for such a test, happy to be corrected here). It’s also not trivial to convey the instructions for how to pace one, and some people are going to push themselves too hard no matter what, which will of course lead to just as bad an outcome as an overinflated FTP from the ramp test. And if you are completely fresh off-the-couch then you need either a fair bit of patience in the test and/or might need to take it more than once to really sort things out - with some recovery in between since truly going to TTE will mean you’re unlikely to be able to hit the same TTE the next day without any rest intervals. It probably works best with 1-on-1 coaching, but I also think KM has put out enough info for free to figure it out yourself, if you’re willing to go through it all.
The other alternative to a field test is to use modelling software like WKO5, which uses multiple “max-efforts” along your power curve to estimate your FTP. That doesn’t negate the need for testing, in fact it requires more of it, and you need to keep the model up to date with such efforts as you go. But it is an alternative that does pretty damn well without requiring you to ride at your FTP until exhaustion. Such tests are useful for giving you supra-threshold to target as well, if that’s something you’re interested in. I don’t think TR will go down the multiple-duration tests path though, because if that was their plan, they would like have implemented something along those lines already.
I think they’re more likely to keep the ramp test, and try and implement some “smarts” to suggest if your FTP is too high or too low. In fact, I think this is the “Thing #1” that Nate keeps mentioning. If you recall, they revealed a while ago that Amber was working on some specific functionality for female athletes, to help them track the impact of their menstural cycle on their performance, and perhaps adjust their workout targets accordingly. They said this was close to release quite a while ago as well. I think, and this is all just speculation so I could be way off base here, but I think that that functionality would have some overlap with detecting over-estimated FTPs, and perhaps the reason it hasn’t been released is due to it being folded into the development of “Thing #1”.
Assuming that my guess is somewhat on-target, and that they’re working on some sort of ML algo to complement the ramp test, I really wish they’d just acknowledge the issues with it now. I can see why they wouldn’t though, because it could undermine peoples confidence. But the thing is, the intertwined problems of the structure of the TR plans, which are targeted at “bang for your buck / minimal effective dose” type training, AND the over-estimated FTP problem, is already causing a loss of confidence - just look at the powder keg that is this thread, the multiple other threads popping up here and on reddit asking if they should just abandon their TR plan thanks to this (frankly, underwhelming) DJ video.
By:
- Adding plans that guide people towards filling extra training time with more Z2 rather than more intensity days, rather than just suggesting it in the notes and/or on the podcast, and
- Adding options that include longer intensity on the threshold days, for those that want a “moderate to large dose” rather than just a “minimal effective dose” , and finally
- Acknowledging the ramp test problems, and providing a guided “field test” workout as a suggested alternative for those who have at least a little experience under their belt.
- they would not only be adding lasting value to the TR product in the long run, they’d buy themselves time to roll out and refine “Thing #1”, assuming it is what I think it is.
Thanks for coming to my TED talk 
Also @OreoCookie I’d just like to apologise for my the rancorous manner of my posting last night. Not an excuse - but I was way more wiped out than I realised from a long, tough event* earlier that day 
I’m gonna leave it up without editing as a lesson to myself not to post potential flame-war inducing stuff when I’m feeling like that.
(* I live in a country/state with zero COVID transmission, before anyone @'s me about that lol.)
Yep, agree with you both - IMO, Training Intensity Distribution is an artefact of basic principles (recovery, required training stimulus, etc) and constraints such as time available, rather than a goal in and of itself. i.e. the TID for a given block of time is an output of the overall plan, not an input to it.
So I definitely thing it’s legitimate to question the applicability of Polarized training as a prescriptive tool to people who are time-constrained vs those who are not (which includes, but is not limited to, pro-athletes).
What I was asking though, was if anyone had any research that pointed to pro-athletes having a different physiology than the rest of us, rather than just a “better” version of what we’ve all got. If there’s no evidence to suggest the former, then I don’t think we should be reasoning along the lines of “we can’t learn anything from the training of pros because their bodies work different to us”. Which was how I was reading a few responses on here as trending towards.
Not trying to derail but, since his names come up, Kolie’s commented on Reddit re Dylan’s vid and wasn’t speaking that highly of TR. Outside of it being somewhat adjacent to the theme of this whole thread I bring it up mostly because in it he also says “death to the ramp test”
Damn dude, nailed it. Most of my issues with the ramp test rolled into a neat little analogy, however you even missed the fact that in this analogy you’re still growing so your feet size is changing relative to your height all the time.
Good addition! I was trying to think of a way to incorporate something along those lines, best I could come up with at the time was switching shoe brands (ie changing training focus, like from endurance to anaerobic). Your idea is clearer and more accurate though.
I get it. Just search for MV+/HV- and you will see several thread, in which people make the same argument — yours, truly, included. But with a little bit more experience, I am no longer sure whether it isn’t better to instead start from a lower-volume plan, making those workouts mandatory and pad with other workouts as time allows and you see fit. Balancing granularity with offering overwhelming choices is not easy from a product design perspective.
TR has a big hump ahead of it, and that is truly interactive workouts. Ideally, TR should take your past training history into account when suggesting a new plan. It should be more flexible and recognize when you are fatigued, and perhaps suggest to pre-pone a recovery week. Implementing this seems like a very hard problem.
I can’t speak as to what the differences are, but I am very sure that pros have some important differences. Probably one very important ingredient is the ability to recover from a massive amount of training (without injuring themselves or burning out). The second one is psychological resilience. I remember a few soccer players who were heralded as geniuses, and they were, while they were on point. But some of them simply couldn’t handle the pressure and crumbled under it rather than flourish. Noodling for hours on end each week in Z2 takes some resilience, too, I think.
I, for one, would be very cautious to base my training off of the training plan of a pro, even if it were scaled down appropriately. That doesn’t mean we can’t learn from what they do, quite the contrary, just that we should be cognizant of the differences between us.
It’s cool, we all have bad days. And perhaps I should also learn to step back from the keyboard every once in a while. In any case, no hard feelings. 
There is a crucial difference in your analogy: there is a simple way to measure shoe size, whereas for FTP all protocols I am aware of measure something slightly differently and infer FTP from that. And we did not even touch on the “proper” definition of FTP, where there is legitimate disagreement. Your (almost) “hour power” is useful in certain circumstances, even though it combines mental and physiological aspects. Using lactate threshold 2 likewise makes sense, but basically puts sole emphasis on physiology. Both have merit, and I think a lot of debate would disappear in a cloud of smoke if we simply used distinct names rather than pretend all are your FTP.
A TTE measures something subtly different than a ramp test, and both are IMHO valid for their respective purposes. I just don’t think that a TTE gives you an inherently better number than a ramp test, they are just different and have different caveats. A TTE places more emphasis on mental toughness, which is probably more relevant if you train for very long, sustained efforts. In the real world, the only instances where you put down power consistently for ~1 hour are TTs and very long, sustained climbs. (I live close to the mountains and only 2 climbs would qualify, which are not even entirely distinct routes.) But even with climbs you have more variability where you have small kickers, where you have to put down the power to keep momentum and can relax a little when the gradient declines a little. I don’t see it as very relevant when you are a crit racer. So a TTE may add information, which is not actionable to you.
I totally agree that a TTE (which implies also a 20-minute old-school FTP test) adds another dimension and may help inform your training. If you know what to do with it, more data is better — provided the data is accurate. But as you write below, if you roll out a testing protocol at scale to people who aren’t already well-versed with training, it isn’t suitable IMHO. More testing also means more room for error. And you need to know what to do with the information. With a TTE you have the result of your FTP test and your TTE. What do you do now? Do you subtract 5 %, 10 % or 15 % from your 20-minute power when you perform your TTR. Clearly, there is a distribution, and aiming for the safe side rather than the average will mean you are giving more people the “wrong” number.
Precisely, and because you need to know the FTP you are shooting for in advance. This is a huge limitation for beginners. Over time I learnt to predict my FTP (as measured by the ramp test) accurately within 2 Watts or so, but that took me a while. I reckon if I changed the testing protocol, I could do the same with some experience. I think there is good reason why ramp tests have become so popular. There are smaller issues that Ric Stern, the creator of the ramp test raises (e.g. that TR and Zwift do not implement his testing protocol for whatever reason). I’m not sure whether these ultimately change the outcome. Probably not by much, I reckon it is more important that you have a consistent, simple-to-use, repeatable testing protocol.
Simplicity and repeatability are huge virtues. After all, what is the value of an incorrectly administered TTE? Less than nil. That’s why I don’t think a TTE is suitable to be the default “FTP” test on TR’s scale to users who probably don’t have the prerequisite knowledge.
I think that ties in with another point: how do you expose the complexity to users? How much rope do you leave users to hang themselves with? Do you stick to one product for all users or do you eventually release different tiers? Do they want to be Photos or become Lightroom?
A product aimed at beginners should likely have more automation, whereas experienced users would likely want to have less. I think many users with some experience would prefer to have bare bones plans that they customize rather than a plan for every occasion that in the end might not fit 100 % after all. Do you offer performance tests that have no actionable consequences on TR’s side (e. g. an analog of 4D power or perhaps TTE)? Would you block HV plans to new users? What if the new user is a pro, presumably there should be a way to override TR, right?
The decision of what kind of decisions you put in the hand of the user is a really tricky one, because you are trading off things. If you offer bare bones plans and aim to educate your customers, you are losing people who either haven’t or don’t want to educate themselves. (The latter is totally fair, there are many things in life I decided not to learn about. Time is finite.) Or are they offering more choice? How do they make sure that customers make the right choice? It is really easy to lead them astray here. Just like most here, when I started TR, I thought “MV only has so few hours of training per week?!? I’m riding way more now!” How much emphasis should you place on what you think is the best path forward, i. e. how opinionated should TR be when it comes to training plans? That’s a huge one, also with no good answer. Presumably, the default training plans should be “good” training plans, although what is “good” also depends on who you ask.
Personally, though, while I think it’d be nice to have more diverse and more adaptable training plans, I think TR has much more immediate problems:
- The race/ride analysis tools are very poor. That’s why I kept TP around, I was able to focus in on e. g. a crucial climb in a race and see how my data changed over each lap. I let my subscription lapse because, well, there weren’t any races last year that I participated in.
- Basic training calendar functionality is missing. E. g. I’d like to mark when I was sick or when I was on a business trip so that 2, 3 years from now I know why I missed 2 weeks of training. TP gave me the option to log data like sleep, mood, weight and the like. That’d be nice to have. If I were a woman, I’d probably want a cycle tracker and that my training schedule is aligned with my menstrual cycle. (This would also suggest that it is better to always have a 4-week training schedule, which would require reworked SSB1+2 plans.)
- The training calendar is not synched, i. e. if I make changes, they do not appear in my (ordinary ≠ TR) calendar app.
- Plan Builder needs to become much more flexible and not break when making too many modifications. Typically, I get sick 1–2 a year. I used to go on many business trips, too.
- Plan Builder needs to be smarter and take your past workout history as well as feedback into account. It could ask you how last season’s training was, whether you have the same/less/more time to commit to training, etc.
The big hurdle to overcome for TR is to figure out a way to make its training plan properly dynamic, i. e. your training plan should take past workouts into account. Once you have the plumbing and logic in place, you can also accommodate people whose FTP is over- or underestimated with the ramp test. The tricky thing, though, is to figure out what your ML algorithm (or however they implement it) should optimize for. If you optimize for the wrong thing, you get into very dangerous territory. But once it is there, it opens the door for so many things. You could take a person’s power profile into account (e. g. where your FTP-as-scaling-factor-for-workouts lies with respect to your maximum aerobic power). Then testing for more characteristic numbers would make sense, since then these numbers would be actionable.
There are some very thoughtful and considered ideas above from @chidlow and @OreoCookie
I think that encouraging more people to do the 45-60m TTE test would help alleviate some of the problem, as most people are much less likely to ‘over test’ on that.
My (limited) understanding is that most pros are going to be several standard deviations away from most of us in terms of genetic gifts (off the couch fitness, response to training, recovery) but they aren’t governed by different physiological rules. Yes, they will start in a better place, respond better to training, and recover from training more quickly, but what they are doing (in terms of training and response) is not categorically different; it’s the same but better, basically.
I also think there are environmental factors to consider. The vast majority of pro sportsmen were not only very good as children but had the opportunity to develop throughout their teenage years and early adulthood. They are thus in a completely different place, as adults, to most TR users. Part of their success (probably a huge one) is a probably 10 year period in their lives where they could focus very largely on their sport without the ‘distractions’ (hopefully it’s clear how I’m using that word before anyone picks me up on it) of work and family. I am quite sure that the substantial majority of users on this board would make huge improvements (though I don’t tthink they’d make the TdF) if they had no concerns outside of cycling and wanted solely to dedicate themselves to improvement.
Now as we all know, that isn’t realistic (or even wanted) by most people, which perhaps limits the applicability of ‘pro’ training for Joe Average. But that is because of time and life constraints, IMO, not because of innate physiological differences. If you have lots of time, volume is the way forward; that has been shown time and again across almost all endurance sports. If you don’t, you need to consider other ways.
I forgot to quote …
“I think the amount of 5watts/kg on this forum, as well as TR athletes such as Nate and Jonathan disproves that you cannot get further than Cat3 …”
I’d like a poll of all 5w/kg+ riders (no let’s make it 4,5w/kg but 1h best instead of “ftp”)
How many hours LIT do they do per week (never mind splitting up phases, make it across the year) in order to create/maintain the foundation for intensity fun. (I’m nowhere near this level btw, but I feel that I seriously need to up my LIT (!) volume to get past 4w/kg (for true 1h power or say at least 45min power, NOT some mathematical guessing with 20min or even step tests))
expected outcome: Without a serious amount of volume (and of course it has got to be lit, otherwise you couldn’t take it) you go nowhere near your 5w/kg. I think we need total numbers to avoid confusion: To me a serious amount is 7h+ Lit, rest intensity (high tempo, ss and above)
However this whole discussion is quite off topic and doesn’t directly match the main aspect of DJ’s criticism which is the lack of recovery from intensity sessions in TR plans
I would be willing to guess the amount of LISS required to get to the destination greatly exceeds the amount required to remain there… It would be interesting to see how many people went from <3.5w/kg to >4.5w/kg (as measured on a true TTE) using a primarily LV training model.
LV defined as in a TR LV plan without tacking on endurance to sessions?
Not specifically TR, but a plan which is less than (say) 6 hours a week and focuses mainly/very largely on shorter (<90m) sessions without any long endurance rides
Edit: what I suppose I am imagining/‘thought-experimenting’ with is the idea that building the base takes time, but maintaining it is easier, so I can see a situation where someone who is already ‘there’ can maintain or stay very close to with much less volume, if that makes sense.
That goes without saying, but it still comes with a caveat. Let me try an analogy
I play the piano way better then I ride my bike. I upped my time “on the instrument” progressively as I made more and more progress. Time at intensity (learning new stuff, hard techniques etc…) rose but as did time at LIT (really, I mean REALLY nailing basic timing, repeating stuff I had already down).
Right now I spend 80% of my time doubling down on basics, improving efficiency in my fingers, repeating stuff and maybe 20% or less on adding more bells and whistles
Oh wait… am I a polarized piano player 
I peaked at 4.5 W/kg (= 321 W FTP at about 72.0 kg) last year, just following a bog standard mid-volume rolling road race plan. This year, I was on a MV+ plan, and within six weeks of SSB1 I went from 295 W (about 4 W/kg) to 323 W (at 75 kg, so at a lower 4.3 W/kg). Then I got sick, and I had to restart. My next ramp test is next week, but judging by how the workouts are going and how my outdoor ride yesterday went, I reckon I’ll be somewhere around 320 W. So judging from my own experience, you can definitely get above 4 W/kg however you want to define it on possibly low-volume training alone.
During on weeks, I reckon I had 1–2 low-intensity workouts. This year it is more, it is about 3, although two of them are short (30–45 minutes after my usual weekend workouts). I think the results speak for themselves, although I do not know whether this is because of the upped volume or the added low-intensity workouts. Last season I tried to add a low-intensity workout on Fridays (and nothing else), and that destroyed me. It was such a pity, I found someone for morning group workouts, but my body couldn’t handle it. Now I am handling much more volume (TSS), but I still maintain two days off the bike. Apart from that, I commute by bike, 2x4 km, at a mellow pace, because I have my daughter in tow.
I don’t have 1-hour power numbers to share, because we have things like traffic lights and hills, so there is no stretch where I’d be able to keep the power on for that long. The last time I did was in fall of 2019 for a hill climb TT, where unfortunately my Bluetooth connection crapped out and I don’t have any power data. My FTP then was around 309 W as tested if memory serves. It was the equivalent of a Cat 1/2/3 start, and I placed around the 50 % mark of Cat 1-equivalent, in the upper third of Cat 3-equivalent. Even though that suggests I might hang in Cat 2-equivalent or so, I don’t have much racing experience, so my racing skills are practically zero. I trained a few times with the Cat-1 guys, and let’s just say they are a lot faster than I am in normal conditions. For a hill climb TT I could rely on my own pacing skills, though, which are comparatively good, and I didn’t need to pay much attention to other riders.
And indoors I just follow TR’s workouts, so also no 1-hour power numbers here. But at least if IF is an indication, I can easily hit 0.9 for 1:00–1:30 hours. (I have the (workout) receipts to back that up, indoors and outdoors.)
Just totally as an aside, if you ever do want to test c.1hr power, find a flat course on Zwift or just set up a custom workout and ride in resistance rather than erg mode. Outdoors, about 7km from my house, there’s a road where I can do an out and back. It’s a pretty flat route with minimal traffic and no traffic lights (just one roundabout to turn round on), but it would probably only be about 40-45 mins total; I’d need to loop again to hit an hour. I may try it one day, though in the summer, probably mid-morning on a sunny day with no wind 
So a “bog standard LV plan” with extra 30 min low endurance/pedalling economy every single day
So added 2.5h every week, 20h during a 8 week block or you could also say an additional short training camp every 8 weeks (yes I know… it’s not continuous 2.5h, not speaking of 2 10h back to back days, but still…)