Until recently, I had not done an all-out short power ride. I managed 10x1’ into 10x30" with equal rest at an average power of 447w for the 1’ and 528w for the 30". The resulting breakthrough brought my 5 minute power from 400w to 440w and my threshold from 297w to 319w.
It seems that similar past rides have really affected my power curve, to the point where I don’t know whether I can trust it. How can I make sense of all of this?!
This isn’t uncommon. The algorithm is the same for everyone on every activity so if there is a jump of 24W to your TP, you’ll need to assess why. Does your training volume explain it? Is it a value you have had before? Did you change power meters / need to calibrate? Were there power data errors? Was your previous signature under valued due to the system’s signature decay method?
You should keep in mind that you produced the power in the activity and can produce it again presumably so your signature should be something you can achieve again. If so, then these are correct numbers and should form a basis for assessing your fitness. If not, then there is something wrong with the data in your ride.
I’ve included a screenshot of the workout. Perhaps what explains this is that I am very good at anaerobic efforts, but not sustained power. So, while the signature changes for shorter durations
(30s-8:00) seem reasonable given this performance, I am suspicious that Xert is giving me unrealistic expectations for durations over 20 minutes.
I can achieve these results again, and my power meter was calibrated before the effort. They are Favero Assioma duos and quite reliable.
Looks like a case of not enough data after the BT event.
There are similar posts from others whose BTE occurs at a point at which they stopped pedaling afterwards which is where the algorithm relies on more data to establish your performance while fatigued.
If you don’t feel confident about the changes you can flag the activity to remove it from signature calculations.
For more accurate BT calculations the activity should include a range of power curve data as shown in BT workouts (search library for “breakthrough”). Or outdoor rides that include a similar range of efforts and fatigue before and after the BTE.
For BTE examples (including some problem ones) see this thread as well as the blog article linked below for best ways to achieve BT’s that dial in your signature.
Let’s See Some Recent BT’s - General - Xert Community Forum (xertonline.com)
Reference –
Breaking Through the Xert Way! – Xert (baronbiosys.com)
How would/could the program gauge fatigue after the breakthrough?
By the continued interaction with MPA. Not an abrupt halt to the activity.
What does that mean, specifically, “continued interaction with MPA”. The rider could have done light LTP intervals at the end. What information could the program gain from that? I can understand that the program can gain insight from successful maximal efforts, but just random riding at the end? Can you provide a specific example of what you mean, and how the program gains information from the example?
Check out the links I provided and notice that the best BTE’s occur with a sustained drawdown followed by a 5-7 sec minimum sprint before collapse and you continue to ride afterwards. That is where the algorithm works best.
Or study the BT workouts in the library which are specifically designed to produce BT’s.
A BTE followed by a halt in further data doesn’t necessarily include the “right stuff” to accurately update your signature. If you doubt the results you can always flag the activity which excludes it from signature calculation while the strain score still counts.
I don’t know how the algorithms work. I’m just another Xert user. But this issue has been mentioned before and is documented.
I don’t think the algorithm gets lots of valuable information when you do some easy stuff after the BT. I think what’s important is to not simply stop pedaling right when you feel like you can’t hold the power any more but to actually ‘fail’ as in drop the power naturally. And maybe try again to push a bit more and fail again.
Like @wescaine described here:
I’m going to try that next time, when I’m fresh and motivated enough, the last one doesn’t happen too often for me 
It seems that you confirm my experience with workouts with short intense intervals. I may improve my power on short durations, but the estimation of my TP by Xert is always way too much higher than what I can really push.
You need to distinguish between can or willing to push. Sometimes you have the power but simply can’t muster the motivation to ride at that for extended periods. If your MPA comes down above TP and recovers below TP, that’s the marker to use for training and not whether you think you can maintain it for an hour.
I already mentioned it on another topic, but when, after a recent BT test, Xert told me to push 312W for a Seiler 4x8 min workout, and I could only push 272W… 40W of difference: I can have all the motivation in the world, I would never ever scrap a quarter of those 40W. I reached 272W by pushing myself to the limit; luckily with that workout, you have 2 minutes of recovery at… 0W ! otherwise, I’d be dead after maybe… 12 minutes !? If I had to go for only 5 additional watts during the intervals, I would probably abort after the first interval, maybe even before the end of that first interval.
As I told on that previous topic, after one month of FLP, on my last attempt at the Seiler 4x8, I could only reach 291-292W… for the first 2 intervals, and then had to abort the workout. Still far away from the 312W suggested by Xert.
You should probably review your breakthroughs then to see if they are inline with general expections. See our FAQ: https://baronbiosys.com/faq-items/is-my-fitness-signature-right/.
Also look to perform a better breakthrough where you reach both high and low points of failure rather than using a sharp bursty single effort which can introduce greater error. If you see small circles floating higher than all the others on your XPMC, this is often a sign that the breakthroughs might have errors in them.
I started another FLP, I still have the last two weeks to finish it. Then I will try to compare my power on some local segments, with the power that I was able to hold the previous years.
I don’t know how to estimate my HIE, except from the graph from your link: as I haven’t done any PP training for months, could it be a good idea to try some sprints, to adjust the signature ? From your graph, and my experience, it seems that my HIE could be too high, because that would put my PP somewhere between 1000W and 1300W, and I think I barely reached the 900W a couple of years ago…
There may likely be problems with your historical or recent data if all 3 fitness signature values are that much over-estimated . It would mean MPA could never be reached at all yet your data indicates you did reach it during recent breakthroughs. Check your power meter sources and try and maintain the same source. As mentioned, if you have “floating small circles” on your XPMC, some data might not be correct.
Well, I checked my power-meters two days ago: I do have an Elite Direto XR home-trainer, and I switched the crank based PM on my bike, between the “Stages G2” from 2016 and the XCadey-L (one-year-old).
Results (on intervals between 140 and 350W: I’m on my second week of FLP, I should not corrupt the recovery days !):
But:
Finally, during the Winter I did a lot of workouts with short intense intervals (I hate doing SST sessions, above all when indoors !): about 3 HIIT days per week, then 2-3 days of easy recovery, and 1-2 days off. The only explanation that comes to my mind, is that Xert’s algorithms are not well suited for those workouts, on the long term. I will try to confirm that presentiment in the next weeks, when I will have the opportunity to test my power of some short intervals outdoors: I hope that Xert is at least right on these short efforts.
Ronnestads and Billats are standard breakthrough workouts.
I do Ronnestad workouts regularly, I sometimes get a BT from that, even if I fail it because I went to hard on the first block ! 
