Just wanted to drop by and leave this video here for debate:
I mean, as of now, from what I’ve played around with, any change I do to my training plan that tries to include back more volume - as I was doing before - the new TR version is predicting me to have a lower FTP than if I just let it at a Balanced approach and Masters plan. I’m sure this new version is not 100% bug free, being a software engineer myself, but what do you guys think?
I might play around a bit more with the plans but, from what I’ve seen so far, I get even “afraid” almost to do any changes to it and loose the highest ftp prediction I was able to see from last changes… (and it takes so much time for each one that it’s actually very discouraging tbh).
Anyway, do let me know what you think of the video and it’s training approach please
I don’t think anyone would disagree, provided all the footnotes Chris mentions in the video. Volume builds fitness. Without being too harsh but pointing out the obvious, youth, little to no responsibilities, no regular job, and Dad taking care of everything related to training, helps. Glad Garrett is able to do this but that doesn’t prove the regular population can do away with structured training.
The top link only goes to the abstract. Is there access to the entire article somewhere?
And when I pull up the second article it’s ’restricted access’ and only has the abstract as well. Do you have access to the whole article or have you read it? It would interesting to know some of the actual numbers as the abstract is worded pretty generically so it’s tough to glean much from it
In that sense, there’s nothing new from Chris. This approach is completely in line with how he usually comments on races: a bunch of knuckleheads — everyone did something wrong and should have done something else, etc. And that’s fine, because you don’t get 100k subscribers by saying everything was correct and the riders did exactly what they were supposed to do. Chris has also said in past videos — probably two years ago — that he personally doesn’t do interval training. He’s also said that anything under 1.5 hours, even intensive sessions, doesn’t require gels. That’s fine too — it depends on your perspective.
However, all pro teams and lower-division teams do intervals. Most of the research focuses on which types of intervals are “better” and how to structure them. There’s significantly less research on how to avoid interval training altogether and still become a top rider
No matter which training plans we look at — whether used used by TR or sold on TP, or offered by other platforms — they all include intervals.
Without underestimating this young rider’s talent, which is clearly there for the reasons already mentioned, I’d say these look more like newbie gains — large gains, but still. And they said that they want to develop him into a GC or climbing contender, which is very different from being a sprinter, classics rider, track rider, or cyclocross specialist. I’d be curious to see how someone becomes a professional cyclocross rider without doing intervals.
So yes, this approach is interesting, certainly healthy, and clearly builds a massive aerobic base with strong prerequisites for becoming a professional. But we also know the peloton is long, and full of riders you rarely hear about — yet they’re still professionals.
My take based on many years of training (and anecdotally from comparing experiences with a lot of other cyclists) is that significantly increasing volume is very rarely going to lead to an FTP increase in a 1 month timeframe. Because in the short to medium term the extra fatigue that comes with that volume offsets any gains, and it takes the body a while to adjust. Building up to doing high volume productively is something that pays off gradually and over many months and years. And everybody is still going to have an individual number of hours beyond which it’s counterproductive to add more, whether that’s limited by lifestyle, recovery time, motivation, age, genetics, etc.
Obviously there will be exceptions - newbie gains if coming from a base of little to no cycling, talented athletes with a high natural capacity for workload. But if, say, you’d been training consistently at 6-7 hours/week for a few years, and then made some lifestyle changes that enabled you to increase that to 10 hours/week from the start of 2026 by adding extra endurance, my guess is it would take most people a few months to adapt, so you likely wouldn’t start to see an improvement over prior year progression until maybe March or even later. But that assuming you were incorporating that extra volume sensibly and getting enough sleep and recovery, you would at some point start to reap the benefits and have a higher ceiling than on the previous lower volume.
In short - the predicted AI FTP feature is awesome, but not all training changes pay dividends in the space of a month, and that certainly doesn’t mean they’re not worth doing for longer term improvement
Yeah… I agree with what you said and I believe that is the shortcoming / caveat of this new AI predictor. As it is currently predicting “only” (I figure it takes a lot of computational power, and it has to be available to millions of users constantly changing things in their prediction window) for one month, it may induce athletes on some short-sightedness not being able to see the eventual long term gains they could potentially be missing out on if they were to actually follow a mid to longer term approach to their training and goals. Tbh, I was low-key disappointed with it being only one month, although as I mentioned above I understand it is already VERY good, given the computational resources it probably requires. Nonetheless, this might be overlooked by most athletes and even the ones that know about this caveat, like me, might be tempted to just stick with the short term prediction to guide their long term goals, since even a short range compass is better than none.
Like I said above, I’ve been gradually increasing my volume over the last years.. Started structured training back in feb, 2022. Last semester I was already at around 10-12+ h/week, mostly comfortably. Mainly due to the old motto of “Volume is King” and also some back and forth chats with ChatGPT, as you can see bellow:
But the best AI predictions I could get last time I tried with TrainerRoad, got me at only ~6.5h/week.
For reference, here is my TSS and current calendar:
Note: I also got a decrease, like many other users, with this update: 266w → 251w FTP.
Note2: It would be nice if TR could just tell us what the best/optimized plan could be, instead of us going through it with a trial and error approach.
FWIW I don’t even think we should think of it as FTP. It’s a number for setting training power targets. E.g. My Build phase starts just before my next AI FTP update in a few weeks. First VO2 session is a 90 minute one scheduled 2 days before the AI FTP. As it happens I have a busy work day that day so showing an unusual degree of advanced planning I changed the duration to 60 minutes and the AI then picked a shorter VO2 workout with the same level. Which actually had pretty much the same intervals but then finished the workout immediately whereas the original one had 30 minutes of Z2 tagged on after the intervals. That change alone dropped my predicted “FTP” by 6W. I think we can all agree that extending a single VO2 session by 30 minutes isn’t actually going to improve FTP by 6W. But I guess what it would do is give the AI a greater degree of confidence that I have the fatigue resistance to knock out a bunch of VO2 intervals and then maintain some high quality Z2 riding (as opposed to slumping off the bike after the last set!) which then allows it to set harder workouts knowing I can handle them. If I did the shorter workout I might still have that capacity - I just haven’t yet proved it to the AI.
This clip mostly proves volume works when you’re young and can ride a ton with perfect support. For most of us, intervals are just a tool to get a strong stimulus when time is limited.
On the TR side, I wouldn’t chase the highest 1-month prediction. Extra Z2 often looks “worse” short term because it adds fatigue before the payoff shows up. I’d set a weekly volume you can recover from, keep one or two quality days, and let the system adapt instead of trying to game the number.
I don’t know if the full test is available online anywhere.
For the u23 study they found that race volume was the same but total volume among the group that went pro was higher. In terms of intensity time at medium and high intensity was the same between groups but time at low intensity was higher in the group that went pro.
To add:
Looked for anything that was easier to read the fulltext and there is this observational study on runners.
The full study is easily searchable online.
Nothing in any of this should be new/shocking. It’s just showing that volume is very strongly correlated with performance and that volume is easy endurance below lt1.
In regards to the video, yeah, if you do enough riding and there are hard efforts across a range of intensity you are probably going to get most of the way there compared to riding super structured.
This kid is living the dream, if the dream is to get fast and ride bikes. His progression is nothing short of phenomenal. You combine this kids’ genetics along with his dad’s wisdom and an ideal training lifestyle and this is what you’ll get. I think we can all agree that as he continues to improve and get closer to his potential, he’ll probably need some structured training to level up to the competition.
It’s interesting that Garrett waited this long to get into cycling when it was probably so accessible for him to do it earlier. I like that he hasn’t done any structured training yet because that’s the type of thing that can turn off some young athletes when the focus should be on learning to ride and have fun, while developing a love for the activity and then the sport.
It’ll be cool to see if he takes up racing and see how he does with his dads’ guidance. Good luck to him.
I agree with Horner in some aspects. I have seen very fast riders who do nothing but casual rides then adding couple group rides per week and be faster than a lot of people who do interval training. For every person who ride to specific power / feel and do a good job judging if they should push that day or not, there’s as many people who push when they shouldn’t, overtrain, or ride so casual that what they think Z2 riding is 3 hours coasting and chatting with friends while 1 hour of it is pedalling in Z2.
So overall I think we need both. Some people will do just fine without actual intervals. By the time you ride 5 hours in a group where 4 hours of it is actual Z2 and racing up 5-6 hills along the route at Threshold, VO2max etc, you get your traditional training anyway. Then other people absolutely need the structure and guidance. I love the structure because then all I do is look at today’s plan and follow it. Much like my wardrobe. When you have 7 same exact black tshirts, you don’t sit there and wonder what you should wear that day
I have a theory that benefit of intervals is they force me to stay in a zone longer than I would otherwise and beyond my comfort level, which then drives adaptation. I think appropriately challenging group rides and races can do the much same thing. I was never so fit as when I was doing 3-4 challenging group rides a week, riding the 8 to 14 miles to and from home to the start and end, and throwing in a race now and then.
He’s racing cat 3 this spring, I’m sure he’ll be ripping everyone to shreds on just raw power. He recently did a ~15’ effort at 6w/kg near the middle/end of a four hour ride. Not bad for a year of training haha.
I’ve done the same effort. I mean, it was at the end of a 4 hour car ride and my bike was on top of my car. But I’m sure I was cruising the neighborhood at 6w/kg….
Garrett sounds built for Grand Tour races. I’ll bet he rockets up the ranks here in the next 1-2 years.
