They aren’t most of the time when that message is shown there is no downgrade. They map the ride against the standard and grade using machine learning, so if it feels there is enough of a difference you get the downgrade.
Both have the same 2 minute break? And you’ve confirmed that the AL is actually reduced, not just that the warning has been displayed. I’ve had the warning displayed and no change to the levels.
Yes.
Could you post the TR graphs and AL?
TR arrived at this based on data, i. e. it used machine learning and statistics, and during training their algorithm “came to the conclusion” that it does make a difference.
One way or another, you are not finishing the workout as prescribed, though. Back-pedaling, soft-pedaling or stopping are all modifications that indicate a higher-than-intended fatigue, although e. g. pauses may happen for other reasons. Examples from my training history include a pit stop in the bathroom or taking care of one of my kids who woke up much earlier than intended. In the end, the algorithms judge your workout on the work you did do compared to what was prescribed.
Perhaps TR is ahead of the science here, because its data pool is orders of magnitude larger than anything a study can cover.
Running in high school rests were standing, maybe some walking or light jogging. In college the rest interval was as important as the interval. It was basically a brisk endurance pace, rolling into the next interval without stopping until the workout was over.
I’m with @eddie on this one… they are not the same!
This might be a case where coaches know what works and physiologists are trying to find out why. (Not talking about 80w to zero, there’s no difference there) but I promise you that for mid-endurance events and up, extending and easing recovery is not happening “in the field.” It may produce a better next interval, but the goal result is race performance. This is where shorter recoveries are a benefit.
Maybe not apples to apples, but:
Active Recovery Induces Greater Endurance Adaptations When Performing Sprint Interval Training
Doing 4x10 sweet spot or threshold for example with 5 min recovery interval vs 8 min recovery intervals are 2 different workouts. Even though the hard work intervals are the same, the extra recovery periods make the workout easier. Therefore a lower score.
However, there should be an option to turn this off with a warning message that it will impact the data for AI to do its job accurately or at all. For those of us who pick our own workouts and have AI FTP off as well adaptive training off, it’s very annoying.
I can speak to my own experience.
I focus on ultras. My coaches have not cared how long my recovery between intervals was (outside of absurd examples), or what my exact power was during them (again within reason).
So my personal experience is the opposite of what you’ve said here.
While I am not an exercise physiologist, I have a large collection of academic training because I’m restless - biochem, medicine, some sport med, etc. I would suggest coming from that academic background that I (some random guy on the internet from your perspective) that is seems physiologically implausible that these sorts of tiny differences would lead to any sort of measurable difference in training outcomes.
But again, this is all gonna boil down to people’s opinions. Many people will disagree with what I’ve said here while many other will agree with it.
What I’d like to draw attention to is more that this isn’t really something we know with any sort of reasonable certainty, so I don’t think it makes sense to take a hard line on any side.
I can certainly see that with ultras, you are out there a long time! I also respect different perspectives and experiences. I think discussions like this is what helps grow the sport.
My experience was from college and an Olympic development team that focused on events from 800 to marathon. My comment regarding recovery was from the 5k-marathon group. Those days are long past but I’m still in the coaching world and am around some pretty fantastic coaches. Workouts and recovery haven’t changed much. But again those are elite athletes.
But we are talking bike racing… and I understand that changes things. I think it is kinda cool the different perspectives. It why I love coaching. There’s the science side and the art. And sometimes they clash.
If it helps. I’ve found (TrainerRoad session in Zwift) that during a recovery session first match the recovery watts and then press pause in the workout tab (Zwift companion), then back pedal, take a leak, pick up a dropped bottle whatever) then pedal and rematch the recovery watts before pressing resume. Works for me and no dropout. However, note the other posters comment about the recovery timing and purpose. I use it only when I really have too.
This has been an enlightening conversation. I didn’t realize how much opinions on rest intervals vary, or that the current standard seems to rely more on “coach intuition” than scientific consensus.
Given the lack of consensus, it seems two things should be standard:
-
For File Writers (Head units, Zwift, etc.): “Auto-pause” should be a standard option.
-
For File Readers (TrainerRoad, etc.): Platform behavior should remain consistent, regardless of whether the workout was recorded with auto-pause enabled. Specifically: whether there is a break in record message continuity, or consistent record messages with 0 values for power/cadence/speed/etc.
Regardless of any opinions of how athletes should treat rest intervals, the second bullet above was the initial point of this thread.
Thanks for getting this conversation going! 
I’m hoping that this can help get the ball rolling for that second bullet. 
Let’s not forget that the system we’re using is based on workouts/plans Coach Chad designed so I’d say his word carries some weight in this matter. Nate himself has said he wanted to “download Chad’s brain” on a few occasions on podcasts and here we are, though I’d say that’s a matter for debate and he(chad) probably would too.
In any case Chad discussed this very thing in a past podcast whether the rest intervals are just rest or part of the workout and he said it depends on the workout. He said in a lot of cases the rest intervals are just there to bridge between designated work intervals and get the workout to the desired length. He said depending on the workout, whether the rider does 5 minutes or 7 between certain sweet spot intervals or the like in his mind didn’t interfere with the desired outcome of the workout which was “get a lot of time at sweet spot and recover in between.” This can definitely be said for anaerobic workouts where taking plenty of time for the creatine phosphate stores to replenish takes time and it’s more about being reloaded and feeling fresh for the next all out interval so it could be 7 minutes or 15 minutes, in either case that’s the name of the game there, full recovery (also something Coach Chad mentioned in a previous podcast). He did however indicate there are times when the rest intervals were definitely part of the intention of the workout which would be obvious in the case of 1:1 or 2:1 work/rest. So yes, if you extend a break between hard intervals in say Taylor, Bluebell or dare I say Spencer/Kaiser that’s changing in the intention of the workout. But if you hop off the bike to refill your water between intervals sets in, let’s say, Wright Peak -7 it shouldn’t affect the desired outcome and by extension the WL. I use that example because last time I did that workout I recovered so hard in that 10 minutes my legs actually went cold. I’m also guessing if I did that workout again and decided to fast forward the rest interval down to 5 or 6 minutes I’d get punished with a lesser level as I’d be shortening the workout so riddle me why all that time spent at 40% is so important other than to get to 90 minutes and not 85. But figure how much more confusing the workout catalog would be if every single rest interval in every workout counted for something and had to be exact for the sake of the stimuli. They could no longer bucket workouts easily into lengths of 60, 75, 90 minute etc. The same could be said with extra shortened workouts. For example in doing the 30 or 45 minute version of bear creek isn’t providing adequate recovery but is clearly just there to make sure you get 2 interval sets in the crunched time period, sore legs, fish face breathing and tasting copper be damned. I’m guessing Coach Chad would say it’s that way to be 45 minutes and if you have the time to take longer between and turn the workout into a 47 minute go ahead.
For myself based on that understanding and the context of the workout the rule of thumb is outside the obvious cases mentioned above anything 4 minutes or shorter is part of the intention of the workout and therefore sacrosanct. 5 minutes and beyond in the majority of the cases is full recovery and free for fudgery which depending on the rider could be either shorter or longer. Between 4 and 5 minutes is debatable depending on the workout.
I feel with this in mind, though perhaps complicated, it would behoove TR to be able to tell internally whether a rest interval is to be strictly observed and when things can be more flexible as otherwise it just feels punitive. Maybe even add that information to the workout description. And to put in one last thing Coach Chad as well as other coaches have said is if a short break here or there means completing the workout versus crashing out and quitting it’s better to finish with 95% compliance than to quit and one should not be punished for that.
IDK. Seems simple to me. If I do anything to make a workout easier than it’s meant to be by taking a break or going easier at any point, then I’m not getting the intended training effect. Of course how deleterious that will be depends on the how much I’m shortcutting the intent. Going easier during an already easy recovery interval will have a small effect. Extending a recovery interval will have a bigger effect. etc. Personally, I simply adhere to the workout unless I just can’t. And if I don’t, I take TR’s AI evaluation of the effect at face value.
Will it?