Updated to Trainer Road AI this morning, new FTP detection increase of almost 6% to 315! Yay!
Then, predicted FTP in 28 days and 19 scheduled workouts, 310.
Wait, what?!?
Updated to Trainer Road AI this morning, new FTP detection increase of almost 6% to 315! Yay!
Then, predicted FTP in 28 days and 19 scheduled workouts, 310.
Wait, what?!?
I just checked out your account. You were on an “Aggressive” training approach. For many athletes, this doesn’t result in the biggest fitness improvement since it’s so challenging. Updating your approach to “Balanced” now gives you a 10 watt predicted increase.
100% aware from the beta thread that this is true and also I believe it.
What I find amusing and maybe can be discussed more once fully released is that I had a low gain or loss I forget when prediction first was released with it all balanced. I played around and actually agressive was a bigger gain. I’ve been a user for over a decade so it has lots of data for me personally and wondering how much this comes into play since I often times spend 3-4 months going hard and then time away. Not saying it has to be me, but this could be an interesting thing to do for a podcast explaining this type of changes to the system. Take some long time consistent users and show how them staying balanced might be good or take someone like myself who is way less consistent but can come back make 50 watt increase then vanish for a few months due to life. Ok a few times burn out, but usually life.
ETA: I set it to aggressive a day or two after its been what 2 weeks now? had some sickness in the house again so slacking and maybe balanced would be different but riding it out.
This is really interesting and maybe not super intuitive! I just checked mine because my latest AI FTP prediction only had me gaining 1 watt or essentially staying the same. I had my training approach set to “aggressive” thinking I am currently up for the challenge and very motivated to push a bit harder and get FTP up. When I set it down to balanced, it now shows a FTP prediction of 7 watts over the same period of time! I think that if this is going to be the case for most people, the TR platform should really make people aware of the implications of different training approaches and how it might affect FTP. Guaranteed, if it is actually going to be that big a difference in FTP change, more people will just opt for the “balanced” approach or whatever the best approach would be for them.
Could be worse. I lost 22 watts this morning (245W → 223W). But then I’m predicted to gain 6 watts 28d from now which would be excellent.
It wouldn’t surprise me at all if my 245W always was too high. I’ve been finding sweet spot work very taxing at that level.
If I can train easier and grow faster, that sounds like a win.
AI FTP Detection gave me +1.7% (4w woo) yesterday and then wanted to do it again today to enable the new prediction features - and then took half of those 4w away from me! I know I didn’t train today but sheeesh!
Is this so that users can … flag any issue they encounter?
So they don’t turn into Grumpy Roger?
hmm, seems counter-intuitive. Why would any competitve cyclist choose balanced over aggressive? I’m a time-trialist and fully expect and plan to train hard to win races. Or is TR being steered more towards the casual/leisure riders and fitness enthusiasts?
This isn’t true. The difference is that we’re setting you to around a level 3 threshold after an AI FTP detection to keep things consistent. Before, you could have been between a level 1-8 threshold, which is a pretty wide range of fitness.
Together with TrainerRoad AI we think the level 3 is the the right point to put you into your progression to get faster.
Harder workouts mean more recovery between workouts. You’re still training hard. For some people (like Jonathan), an aggressive approach can lead to a higher FTP. For myself, I get a lower FTP.
It’s different for different people, which is cool.
This is also just predictions, and there’s a range of where you could end up. So if you think you can handle more, you can try aggressive tactics and see if you can outperform the AI’s predictions. I don’t recommend that though.
I trust the process, and have for around a decade! TR’s database must be massive, and I’m excited to see if I can finally reach my goal of a sub-hour 40K TT.
I think that message needs to be more clear and maybe it will be in the podcast. I think some have taken it to be that whatever number gets you level 3 in Threshold is what AI FTP sets as FTP. You are saying here that AI FTP calculates a number and then the levels are adjusted down or up to Level 3.
I mentioned it in one of the threads but this could be a great podcast episode picking a few athletes across the spectrum and showing how it changes with some doing better on balanced and some on more aggressive settings. Maybe even someone who backing off a bit could do better. This could go a long way to showing people why this is a thing especially as lots of people are being onboarded.
I’m sure they will cover this. I agree that the onboarding could have been clearer on this. When I joined there were well over 1000 posts and I couldn’t get through all of them.
I think…that the focus has been on making sure it works on lots of different users on lots of different software settings and can still push a workout to Zwift for example.
Several staffers seem to have permanently manning the posts and responding really quickly.
I experienced the same thing this morning: I was prompted to trigger an AI FTP detection under the new system, which I did, and it came up with a +1W increase. That seemed reasonable, since I’d recently done an FTP detection under the prior regime, so it actually would’ve been kind-of suspect if a wildly different number had been generated. But then I took a look at my “AI FTP Prediction” and it showed a -2.2% change over the next 28 days. My plan is set to Aggressive, like others have mentioned, but that seems a little weird since, to date, I’ve been handling the program well and, according to TR’s prior FTP detections & PLs, steadily progressing, and haven’t had any FTP Detections that moved me backwards. That being said, I only started up with TR two and a-half months ago (albeit with a very extensive imported training history), and I did hard fail a key workout yesterday – only the second time I’ve failed a workout since starting the program, and the first was very early on in the program…but maybe being right on the heels of the fail is influencing things here short-term.
Thing is that I’m now unable to figure out how to adjust my training approach. In fact, I can’t even find it listed anywhere any more. I’ve looked under Career, Training Plans, Account, etc, on the iPhone and iPad apps, as well as on the www interface, and nothing. How does one adjust their training approach now, or even view it?
Side note: this is NOT a request for anyone TR-affiliated to go in and adjust this for me from the back end. I would simply like to know how to monitor and adjust the setting myself.
Thanks!
next to where it shows your current ftp w/kg on the calendar should maybe say balanced or aggressive.. click that
Money. Thanks for the quick and accurate reply!
Just an n=1 on the training approach thing, I had initially set mine to “demanding” and that gave a projected 246→250 FTP increase in 28 days. I set it back to “balanced” and now it shows a projected 246→271 FTP increase in 28 days! Going to keep it on the latter and see how it goes!
I am still confused about the new aiFTP then. FTP is a proxy for MLSS, typically a duration that an athlete can hold for ~40-70 minutes.
Shoving everyone to level 3 seems to completely ignore the TTE aspect of FTP?
If TR is taking someone at a high ftp “level” and giving them a large increase in ftp I don’t see a physiological reason for it?
Personally I can easily push my time at ftp to ~70m in the early season but doing an actual test my ftp has not gone up, just out. If I was to apply that large increase to ftp to myself and bring me down to a threshold level 3 my ftp would be rather inflated from what I can actually hold for ~40-70m.