Following on from a thread over the weekend, having the 28 day counter (to the second!?!) is a bit of pain if you want to shift things at the start of a block for “life” reasons.
For example, I follow Low Volume, partly to give me flexibility for work/ family/ life. Coming out of the recovery week, I wanted to start the next block on the Saturday or Sunday rather than the Tuesday, but couldn’t update FTP, and get the replacement workout, until the 28 days was up.
Additionally, having it to the second means that eventually, we’ll be delayed starting the next block or not even able to move the workout earlier in the day.
I suggested 25 days, but maybe 21 days is better. And just to the day, not the hour, minute and second!
I believe a the end of a block you can insert non-riding days, ie. Adjust the start of the next block. If that is two far out then use train now to fill in the gap as you desire.
Hey there! The primary reason we keep that 28-day timer on is that changes in FTP take time and we don’t want athletes worrying about changing their FTP before it’s necessary to do so. This support article fully explains our reasoning.
If you only want to use AI FTP Detection to get a replacement workout for the Ramp Test, you can delete the Ramp Test and use a TrainNow workout, or you could search for a Productive workout in the TR Workout Library that matches the rest of your plan. If you don’t feel comfortable doing so yourself, you can reach out to TR Support via live chat on the TR website or by an email ticket, or you can DM/ping me here on the Forum and we’d be happy to get your plan adjusted.
I get the logic, but I don’t think it’s taking into account Life/ flexibility that I love about the TR calendar.
Whatever about the 28 days, I really don’t see the logic of tying it to the exact second.
If you’re insistent on the 28 days, you should allow the swap ramp test for workout option without the FTP detection - part of what I’m supporting is AT so I don’t have to guess what workout to do on my plan!
Don’t worry about it. 1 workout wont make any difference. Just pick a trainnow workout on the sat or sunday, update the FTP on the Tuesday and carry on
I feel like I mostly agree with this logic but what if you’re seeing the big gains from the start of structured workouts. As in when your gains are in the 15+ watt range it seems like your workouts could be done at way too easy of a level compared to your “real” ftp
You’ll still get gains by progressing to higher workout levels at your current FTP. If you’ve progressed really rapidly in the previous block and are bumping up against the top end of the progression levels and they’re still feeling pretty easy then you can always manually adjust FTP up a bit. But yes, a couple of days of doing workouts at old FTP instead of (anticipated) new FTP is really going to make no difference at all in the long run.
Am I wrong or the timer is effectively 29 days? I found that if, for example, I use the AI detection on a Monday, then the next time I can use it is on a Tuesday, because it will unlock only at midnight on Monday. Not super nice.
Unless you’re off the couch or coming back from sickness, are people really seeing massive jumps that warrant checking more frequently than every 28 days?
A counterpoint, though… if you’ve had 3 weeks completely off the bike due to vacation or illness, the current system doesn’t let you autodetect a new, lower FTP. You’re expected to pick up training at the same FTP but with lower progression levels. The lower progression levels definitely help somewhat, but I think 3 weeks inactivity really does warrant a new lower FTP to get you in the right zones…
That may be an edge case but Seems like that edge case would be important to handle. These would more likely be the people new to trainer road and you’d want the new people to stick around.
I’ve gone up ~50 watts since when I started in early Dec. (still low FTP but early gains can be big)
Personally speaking, 28 days seems a good fit (if you are working within the bounds of a relatively accurate FTP value).
Over the course of a 28 day cycle following the suggested workouts as prescribed via AT, with the majority of my sessions usually surveyed as hard or very hard (with some moderate ones sprinkled in), I find my FTP usually only goes up 3 to 5 watts, if it goes up at all.
To me, being able to check it any more frequently would likely yield no benefit… YMMV, of course.
This is an important edge case, especially when you have to take time off when your PLs are low. E.g. you leave a plan at Sweet spot 1.6, take a 3 week break and come back, SS 1.0 might still be too difficult.
Even if you’re PL’s are higher, it would be nice some sort of “ease me in” option after time off. There’s something that’s hugely demotivating to come back after time off and have TR give you an “achievable” or “productive” workout only to either fail it outright or barely be able to complete it.
Yeah. I don’t want to overstate it - if you have the experience to self-select a reduced FTP or go the @DingleTower route of doing an actual FTP test (gasp!), then that’s fine and you can work through it. And in some cases, powering forward with reduced PLs will be just fine.
But we’re talking about improving AI FTP detection here, so I think this case should at least be considered.
(Says the guy who just came back from a two week holiday)
Not off the couch, but new to cycling, so yes, every new detection is quite a bit different and the 28 days lockout has slowed down my training quite a bit because it took the algorithm several months to get to a level that feels kind of right.
And… no, I don’t intend to do an FTP ramp test because it feels too “ancient” to bother with it, and as far as I can understand from the forum, it has its own issues.
PL helped somewhat, but at the end of the day, if 75% FTP feels like a recovery ride, I should be able to ask the algorithm to look into it without waiting 4 weeks.
I think AI FTP is a great feature, but the 28 days lockout is a big downside. Once the algorithm gets it kind of right it will be less of a problem, and that is why users that already have a good idea of what their FTP is don’t see this as a problem.
I did that a couple of times, but for somebody who never trained with power until a few months ago, it took me a little bit to understand the “signs”.
Now I’m pretty comfortable, after reading several posts on how certain workouts should “feel”, but back in September when I started, the first AI FTP detection was 113W and it was not immediately evident to me that I should have bumped it immediately.
I also signed up for and learned to use intervals.icu fairly and that was an eye-opener.
My thesis remains, as of now AI FTP is not friendly for people ramping into the sport, that experience either newbie gains, have no idea of a ballpark FTP starting point, or both.
At the end of the day, technically the interval for AI FTP detection should be adaptive. Fixing it to a certain interval, I would think, is a stopgap measure while the technology improves.
My $.02 here based off the above 2 comments… if you aren’t comfortable with increasing your FTP manually, you may be doing yourself a disservice by NOT doing an “initial” ramp test at the outset of your training with TR.
yes, the ramp test may have its flaws (and I for one don’t like them and am thankful for AIFTP) if you came into TR in September having only done unstructured riding (my assumption), what FTP did you set your initial 10 workouts at to get that first AIFTP detection value?
Having done a ramp test FIRST may have gotten you MUCH closer to where you SHOULD be for your training. I suspect this would also be TR’s suggestion…
