Thanks for the replies xertedbrain and ridgerider2.
If XSS is the primary variable that is going into the XATA recommendations and fitness gauges, then I’m comforted but still the results metrics and recommendations seem difficult to follow and I’m admittedly having a hard time seeing my way to effectively using this as my workout coaching and training plan tool.
I do almost all of my workouts outdoors. Even though they are almost all structured, there’s no way to get around some stoppage time on every ride and sometimes the stoppage time can be considerable even during build and peak phases. For example, today I went and rode to a practice race and racked up 100 XSS in an hour. Then I had to stop and take a work call that unexpectedly lasted over an hour before I rode the last 20 minutes home at a leisurely pace. Total ride time was about 1:30, but elapsed time was almost 3 hours. Yes, I could’ve stopped the computer and saved the ride right after the practice race, then recorded a separate ride when I went home. But does that seem like an odd way to trick the XATA to anyone else? I’m still not getting a good handle on the utility of elapsed time being a valuable metric on any of the displays instead of ride time and it does seem to be influencing some things. Right this very moment I can look at the ATP tile on my training page and it now says I need to target 12 hours/week to achieve the slow IR setting with FF at -5. So far this week I’ve logged 3 rides (one per day) for a total of 4:22 moving time and 302 XSS, with today being a hard race ride and yesterday being VO2 max intervals. A typical XSS/week for me ranges from 500-700. But at this very moment the XATA says I should get out and log another 115 XSS over 1.5hrs of riding and I know when I check it tomorrow it will recommend an even more strenuous ride. Basically, from observing the XATA for months now while I’ve had a human coach, I’ve decided that the XATA is trying to kill me. Sure I can ignore it, but isn’t the point of having it to serve as a guide for what you should be considering next?
I’m no pro, but I train a lot and enjoy putting in the work. I think what XERT has going here is very interesting and the art/science of picking workouts has intrigued me ever since I’ve had a coach. And as a guy who works in data for a living I’ve long thought that an algorithmic approach to a training plan and picking workouts must be possible and even optimal. But I’m having a hard time making sense of what I see here. Some of that is because I need to become a better user of XERT, but I also think there’s some room for system improvement and I’m hoping that me writing here isn’t perceived as complaining but as an attempt at providing feedback that might help improve the system.
Thanks Scott. I may need more clarity on things to understand what the issue is you’re seeing.
XATA is fundamentally about XSS and Focus. It calculates your XSS deficit to maintain a set improvement rate. This is so inherently simple yet so perfectly suited for training. Secondly, it determines what focus the XSS should have. Again, so simple yet perfectly suited. The remainder - estimated hours, workouts, workout XSS are all just ways to put the deficit and focus into actionable workouts but that’s just a convenience. Ultimately if you keep the needle between 11-1 and hit your Focus, you’ll be all set to improve at the improvement rate you set towards your athlete type goal. There are no ifs ands or buts. If things are too much, adjust FF or back off IR. Again so simple yet perfectly suited.
Elapsed vs. actual time has no influence in the advice but will need to be something that you will need to manage because everyone is different. Time management for training is an entirely other dimension that has little to do with training but more to do with scheduling. We haven’t attempted to solve the scheduling problems because frankly they are very difficult and the utility is questionable when 90% of athletes have varying schedules all the time. Giving you an XSS deficit and Focus frees you up to do what you can in any given day. That’s always been the recommendation: if you have 2 hours to train, train for 2 hours. If you have 4 hours to train separated by a 1 hour break, then do that. You’ll need to use filters and plan things out if that’s what you’re wanting to achieve. Review XSS deficits beforehand and choose workouts that meet your needs both in terms of training value and time you need. We have many athletes now using XATA to get to 5 star Training Status and beyond. Some may find the training too much once they get to 5 stars and leave improvement rate at something like Aggressive-1. (We should probably add a warning for that choice.) but at 2 and 3 stars, you should be able to easily close the XSS deficits at improvement rates lower than Moderate-2. How long you ride on any given day will not affect your ability to do that.
Sorry for long delay in reply here. I’m still really struggling to make sense of this. Maybe the elapsed time thing is ultimately acting as a red herring for some other issue, but at the core of what I’m getting at here is that the expected training time and corresponding stress (or stress and corresponding time?) just seem to be very high. As racing season is basically over I have been toying with IR settings between Maintenance and Taper to see what the XATA spits out. But the XSS expectations each day, and the resulting workout recommendations, are a little shocking to me. I’m no stranger to working hard. As of a couple months ago I was at 4 star training status. I’m truly not trying to brag (I get ego beatdowns regularly in races), I bring this up to prove that I know what it is to work at my limits, and sometimes over, for extended periods of time. Yet I’m finding what is getting generated out of the XATA every day to be hard to understand. Like right now I’m coming off 3 solid consecutive days of training right after a race Sunday that yielded an XERT breakthrough, and I know my body very much needs either a day off or a light recovery spin. But XATA’s top recommendation today is “We Got This - 90”, which I am not afraid to admit is just far too hard for me today. The next several suggested workouts are also too much stress. I am concerned that if I followed the XATA religiously I would pretty quickly burn out. I’ll spend some more time watching/reading tutorials and picking through the forum here. But if anyone has an idea of where I am running afoul here and how I can better manage this, I’m very eager to find out how to become a better user of xert.
Pre-base? That’s going to give you a much less focused training suggestions. Once you get to Base, you’ll likely see more targeted and less variation from the focus that’s used by XATA. It’s the reason why the Endurance focus suggested by XATA is resulting in workouts like We Got This and Iron Man. You can experiment by moving your target event date to something that’s just under 4 months away (so you’d be in a proper base phase) to see how things will change.
But this does bring up a good point with how things work during pre-base and we may want to improve the advice so that you’re not offered workouts that are too challenging even if focus isn’t critical. We’ll have someone look into that.
Thanks. I will play around with the target date some more.
Might there be a better way to understand the weekly XSS that the XATA is trying to get me to? The time goal is for me part of the confusion. The XATA seems to want me to crack out 110-140 XSS per day. The phrasing used in the dynamic text seems to suggest its starting with a weekly time goal of X (based on historical elapsed time), so X/7 is my daily time needed and then the XSS and resulting workouts appear to be some function of that X/7. If I follow it strictly it ends up being 770-980 XSS in a week which for me would be the kind of weekly stress at the very extreme end of a heavy build phase. Sorry for all the questions. I just think I must be missing some key concept(s) here in learning how to manage the recommendations of the XATA. Possible I’m just a hopeless case.
I left this reply sitting from awhile back and forgot to send it so here goes –
I assume you have resolved the elapsed time issue by not relying on auto-pause to determine every activity’s start/stop times. 
If you were swimming laps in a pool that might include short breaks in the pool as part of the workout. If you leave the pool, towel off, and hangout at the cabana for an hour then swim more laps later, that’s two workouts, not one.
If you were to run up a mountain at a ski resort in the summer, turn around and run back down, that’s one workout. Different strain efforts, but one activity nonetheless.
Next time you decide to eat lunch and hangout at the summit lodge for an hour before the downhill return. That’s two distinct workouts.
The problem with cycling is momentum, gravity, and darn auto-pause. 
The deeper you dive into the logic of auto-pause and moving time the less sense it makes to track as a training metric. All short breaks are rests-in-between same as coasting downhill (0 watts, 0 cadence) or stopping at a red light. Any long breaks off the saddle and away from the bike are subject to stop/save and start another activity. Makes sense, no?
Perhaps one day head units will be smart enough to figure that out. Nix GPS movement as the only factor. How many minutes with no power/cadence activity constitutes auto stop and save logic? Jump back on the bike within that time frame and the recording continues as one activity. Jump on the bike beyond the limit and the initial activity is saved (without the end pause) and a new activity starts as cadence/power data is detected.
Those short breaks are legitimate RIBs. Moving time is irrelevant. What matters for training purposes is the elapsed time and strain accumulated during each activity.
I get why so many ask for moving time since they’ve grown accustomed to seeing it elsewhere, but the metric is really a false narrative. Auto-pause based strictly on GPS movement blurs the distinction between valid rest intervals/stops/breaks during activities.
Moving on…
As @xertedbrain mentions Pre-base (and Post-event) doesn’t follow any progression logic. XATA will generate a variety of workouts around your selected focus duration (athlete type).
See last line here: Program Phases – Xert (baronbiosys.com)
In this scenario I suggest you consider an alternate workout depending on how you feel including selecting from the full list of twenty (Load More).
Yes, “endurance” in Xert isn’t the same as old-school sub-LTP minutes/hours. Xert’s hybrid 3 zone and high, low, peak strain model is quite different than a traditional 5/7 zone model.
Open up that workout in Workout Designer and note the Rating, Description and especially the XSS/KJ split. It’s a “moderately difficult endurance” workout including what I would call the gentler side of VO2 max intervals. If I was in Pre-base though and not feeling up to it, I would definitely be looking at the alternates list.
Sweet Spot, Threshold and Polarized Training … By the Numbers – Xert (baronbiosys.com)
Focus duration and power targets for Endurance - Support - Xert Community Forum (xertonline.com)
You can kick TL down to a lower level at any time my changing IR to Off-Season. Also consider how the pacer needle in XATA works. It is a rolling 7 day average. If you suddenly decline this week and purposely let it lapse into the red, a week later the needle will reflect your new pattern/hours. TL in turn affects your stars status count and corresponding difficulty level for suggested workouts. This is most evident during a phased progression.
XATA - what does it say? - General - Xert Community Forum (xertonline.com)
I’ll assume your “2 years of non-stop structured training” must have included some downtime. 
Thanks @ridgerider2 and @xertedbrain. A lot to digest in that last post. I think clearly part of my challenge is that my mindset is that of a pretty traditional phased program and the vocabulary that goes along with it. I think also the fact that I do most of my workouts outdoors in a fairly urban/suburban sprawl geography where stops are inevitable us also adding to my confusion. Xert definitely seems more tailored to indoor training but I am convinced it can work well for an outdoor regimen as well since other people seem to have no trouble doing so. I just have to gain a better understanding of some of the logic in the system, which you guys have done a great job of helping me with and I sincerely appreciate the time and effort.
I do definitely take rest periods in my training. My favorite weeks!. @ridgerider2’s note about the status needle and what happens if I take a week off triggered a lightbulb for me. Back to the drawing board for a few weeks to try some different approaches.
I actually feel xert is suited well to BOTH indoors and outdoors training. Probably even more so to outdoors.
Just my take after using xert for the last 12 months.
In the hands of a knowledgeable user, Xert can guide you through indoor and outdoor training, optimizing outcomes based on your availability and target focus/event. Only the top end coaches know how to prescribe outdoor unstructured rides and activities as part of your training. Many frown upon it and insist you follow their structured plan as much as possible.
Top end coaches are able to match what Xert does. A smaller subset of them can out perform it. A knowledgeable Xert user can guide their own training, gaining a level of self-coaching that will be hard to surpass. The best coaches are the ones that can bring all that Xert has to bear developing a comprehensive training program that combines indoor, outdoor, fun rides, skills training, psychology, race preparation and race events, optimizing results in the process based upon the athlete’s feedback. It’s a continuous feedback process just as it is in Xert. It has to be in order to be optimized.
Couldn’t agree more. Just please add the feature that will tell me how many hours I’ve actually ridden the bike and not sat in cafes.
Quit arguing about the validity of the feature and just add moving time to the weekly summary table. It doesn’t even have to be selected by default.
There is better way to measure “active” time and apply it to averages. Moving time is just an estimate but doesn’t quite capture it. As we’ve mentioned before, sprinting between stops at 600W should not average out to 600W because you had to stop at stoplights. (There are people that sprint when they approach a red light so that they have less moving time and higher speed averages. What’s up with that?)
So we have some better ideas but the changes go pretty deep into all our software so not something we can easily add but it’s in the plans.
Another zombie resurrection by me. Since this thread I’ve been more disciplined with breaking up rides into 2 files if I’m stopping for lunch or some other extended pause. This has gotten my average weekly ride time (and suggested for future weeks) down to an appropriate level. However, today I had a massive puncture that, after about 50 minutes of attempts to fix, resulted in the end of my ride. I did not stop the ride recording initially because most punctures I can sort out in <10min. But now I want to chop off that last hour of no activity because it will throw off my recommendations for next week. The only tool I’m aware of out there that will do this is at fitfiletools. So, did the trimming quick and easy uploaded to XERT, but! Now when I open the ride all the stats show the ride lasting 1:59 (including non-moving time at stoplights) However, from the activities dashboard it still clearly shows a duration of 2:51. Why would this be?
I’ve found the fitfilestools a bit strange when I upload the files into Xert. I recently did a 600km audax, so had to record it as one activity to get my ride validated by the Audax organiser, Xert however doesn’t like rides over 24hours so I had to split the ride. When I uploaded the two files the timing was way out.
Have you tried using the “Split Ride” facility in Strada itself? I’ve never used it myself but noticed the function a while ago.
No, not noticed that. I find the route planner in Strava quite basic, but I’ll have a look at it.
What was your ride recorded with? I am seeing similar behavior from a ride today.
Trimmed out the bogus section but Duration under Activity tab (or table view) still shows the longer time while activity details is correct time for the recorded section.
Unfortunately it didn’t work,once I’ve split the route I can only export either a GPX file, which Xert won’t accept, or export the original file which is too long due to the 24hour cut off. On a positive note, my Strava now shows the ride split properly without the sleep section.
@ridgerider2, ride was recorded with Pioneer CA-600 head unit: a dedicated bicycle gps computer. I dont know a ton about the actual formatting of data in a .fit file, but i do know that before and after the actual GPS waypoint recordings (mine is once per second of elapsed time) there is some meta data about the ride. I suspect that fitfiletools is not modifying that meta data which probably still states the unmodified duration of the ride and is what is displaying on the main activities dashboard. Probably a quick coding fix…if xert is even using that particular number in the training advisor algorith?