Hi, sorry if this is slightly off-topic and equally if it is a stupid question but I literally just logged here with a very similar query.
I’m not following a plan at the minute but am just riding outdoors, doing a couple of Zwift races a week and then hitting “train now” on Trainerroad the rest of the time.
My question is, if I use automatic FTP detection does it take my Zwift races into account? I assume yes but also can’t shake a nagging doubt that all those lung-busting crit races aren’t actually going into the pot, if you see what I mean? I wonder if I need to associate them with a workout?
Secondly, if I was following a plan (which I doubtless will be again sooner or later)…does adaptive training take the Zwigft races into account in its calculations? Or again, do I need to manually associate them with a workout?
Yes, assuming you have the required 10 inside TR workouts.
What AI FTP Detection Needs to Detect Your FTP
In order for AI FTP Detection to have enough data to accurately detect your FTP, you need a minimum of 10 completed TrainerRoad indoor workouts. If you haven’t completed enough workouts, don’t stress—you’ll see how many more you need when you click the button to use the feature. Once you’ve completed these 10 workouts, AI FTP Detection will analyze and incorporate all indoor and outdoor rides with power data, whether or not they’re TrainerRoad workouts.
No, that is the missing piece of Workout Levels V2 (unstructured rides/races) that we are all waiting to get.
This is generally not recommended, especially per TR reps. Unless you actually perform a TR workout to match that workout, pairing other outside workouts that are “close” has the potential to cause more harm than good.
Without debating the words, a number of people noted the same or similar when AT was launched. TR’s adaptive algorithm appears to be based on progressing structured workout difficulty, which is very different approach from those that evaluate your progress and adapt training based on goals and actual results achieved.
@PhilSJones Outside rides are factored into AI FTP Detection, and work continues for scoring unstructured outside rides. Insights and data gleaned along the way from this process have helped us improve how AT manages your training.
@Chop_StickWorkout Alternates can be a good tool to use in the kind of situation you’re describing. We advise athletes to stay in touch with how they’re feeling while training, and Workout Alternates can often help athletes find training sessions they can sub into a plan if, for example, they’re feeling fatigued after a big weekend ride.
Can you elaborate on your feedback about VO2 max training? VO2 max workouts will vary depending on your Progression Levels and what plan you’re following. To generalize our entire approach to VO2 max work as “out of date” doesn’t seem to fairly describe how AT integrates training your VO2 max.
As for AI FTP Detection, knowing your FTP is an important factor in optimizing your training, and AI FTP Detection makes this piece of the training puzzle much easier to stay up to date. Not everyone has years of training under their belt to know where their FTP might be, and even if you do, this feature is in place to make sure you stay locked in at the right workout difficulty.
On the subject of Android app slowness, we have a team of engineers working to address this problem at the moment. If you haven’t already, try out the TR Beta Program to get the latest performance improvements as soon as they’re released. Keep us in the loop on how things are working on your device as we’ve been actively making updates to address this issue.
I thought exactly the same back in February but decided, very reluctantly, to cancel. I’d had enough of paying, waiting for outside rides being taken into account. Also they’d made me feel like a bit of a freeloader only paying the reduced rate. If and when they get it sorted I may reconsider though depending on the pricing.
I’m just thinking out loud here. Would appreciate others to chime in and change my mind (or not).
If outside rides and races would influence my training PL wouldn’t that potentially skew them and make them unsustainably hard? For example if you train for your A race and go trough base, build, specialty and properly recover before the race, taper and then really smash it and lay down some power PRs, your progression levels will be “off the charts” due to some power PRs. FTP doesn’t get bumped in Specialty so once you hit the end of your plan your FTP will (potentially) increase and the PLs knocked down a bit (as far as I understand). And now you train on that higher level - hopefully sustainable.
Do you want to continue to train on that high level that your A Race has potentially set? The upcoming workouts will now be much harder than before the race because they now take your race performance into account. AI FTP already considers unstructured outside rides and bumps your FTP if applicable. FTP is still the metric, which workout levels are based on (to my understanding) and set the range in how hard workouts are. But maybe you want to relax after your A-race and detrain a bit.
I don’t really see a need to have unstructured outside rides affect my PL - on the contrary. Just because I may do a 4h Z2 on a sunday morning doesn’t mean I have the time to do that on lunch break during the day. And short endurance rides can only go so hard in 90min. Same with threshold, VO2 and anaerobic workouts. I don’t have the time to taper or fully recover before these workouts. They shouldn’t be so hard like a race or A-event imho.
I personally don’t think that makes a whole lot of sense to have unstructured rides count towards PLs. They count towards FTP, which is good enough imho. I’d rather have PLs be workout specific and FTP the gauge in which range workouts fall and fine tune that with the PLs rather than have races and fast group rides skew those PLs and throw off my future training. Hope that was somewhat comprehensive. Not english native, so sorry if its just a blurb. Curious to read your thoughts.
I’m sure this is exactly why it’s taken over a year and half for them to release this, it takes a lot of contextual nuance to apply PLs. Like you say, a one off heroic effort shouldn’t reflect the level you can train at, but it should inform the model in some way.
Yes, I get that @ZackeryWeimer Its counting outside rides in PLs that have been promised for ages.
Anyway, given that, how do my recent 24hour and 12hour Time Trial racess affect my AI FTP. I suspect they are an edge case
As others have said, (@jn92) perhaps an A race should not be FULLY reflected in changes to the PLs. @Nate_Pearson@Jonathan what is your thinking on this? Maybe this should be optional and our choice.
(PS With the the 24hr my A race, and the 12hr 3 weeks later, I could only put the 12hr in as a B race.)
Even if PLs accurately reflected your A race I’m not sure there’d be an issue. E.g. Let’s assume after a 12 and 24 hour TT (which blows my mind by the way!) that your Endurance and Tempo PLs were at 10.0. That’s fine - after those events I seriously doubt you’d be jumping straight into another block where you’re targeting improvements in those energy systems. I assume you’d likely have a block of recovery where you’re either not doing structured rides or if you do pick a TR ride it will be an Achievable one. And then quite likely after that you’d look to target a different energy system e.g. A block of VO2 to raise your FTP/ceiling. Similar for any other A race - e.g. If you’ve peaked for something punchy and achieve very high VO2 or Anaerobic PLs, then you’d want to recover and then go back to some base training before building again, so by the time you get back to trying to raise your high intensity PLs they will likely have degraded to a level that enables you to do that.
@jn92@NateP@PhilSJones@cartsman Good news! We’ve already thought of how those kinds of “hero efforts” might affect your training.
Our system will intelligently detect races/rides that are way outside the bounds of where your current PLs are and exclude them so as not to derail your plan. We also have the ability to manually exclude these kinds of outlier efforts from PLs and AI FTP Detections if we need to override things.
They obviously don’t trust the results in all cases enough to go live. I wish they would exclude the cases they are worried about with a message like “PL cannot be calculated for this workout” then release to opt in status.
Is it possible to at least release to Early Access the AI/ML assessment of an outside ride, even if it doesn’t feed directly into adaptations? This would go a long way in showing progress and help users assess the training value of unstructured rides.
Additionally, I assume this applies (or will apply) to structured rides where intervals are added or skipped, rather than just a pass/fail.
It would really be great to have a Repeat Interval button. Now you have to extend the cooldown and adjust the intensity, which is a bit cumbersome in some cases when the cooldown is a ramp.
The analysis is just based on your survey response, qualitative. There isn’t any detailed quantative analysis at the moment. At the extreme, you could tell the system you’ve passed with flying colours, when in fact you struggled and came nowhere near close thus misleading the system.
Are you talking about for an outside workout? If so, I go into my Garmin and use the workout menu to repeat, pause, or skip intervals. I can’t remember how you do this on Wahoo, but I’m almost certain you can.
