What do the coloured circles on my XPMC mean? – Xert (baronbiosys.com)
Hardness Test workouts are always going to be tough to complete even when especially fresh and motivated.
When I know I’ll be short of the High strain goal (and I’m not cooked), I’ll add some more red intervals during cooldown at whatever watts and duration I can muster. That may require skipping back to start of cooldown to extend the workout as necessary.
I havne’t had a breakthrough since Nov and workouts weren’t feeling quite hard enough so last Saturday I loaded up a long climb on the smart trainer and started my Garmin with the MPA IQ data field.
My plan was hold 10-20 watts above threshold until MPA was depleted, hang on, and then give it full beans when I couldn’t hang on anymore.
Previous
Current
Thought I had a breakthrough. Nice evening ride with impromptu sprint with buddy. For some reason peak power not revised??? Not even noted as a near breakthrough. Will have to do the sprint again and see what happens.
What was duration of time spent above MPA? Did it give ‘fakethrough’ instead of breakthrough?
Although I’d have thought PP would be a pretty instantaneous one.
In short, too short. 
The Xert server treats any short spikes as potential power spikes that should be excluded from signature calcs.
If you want to prove your PP value is higher than it is include some short sprints with at least 5-7 secs of max effort. Up a short hill is the easiest way to do that.
@ridgerider2 is spot on here. The system can ignore very short, high-power efforts which are sometimes just erroneous/wacky data from a power meter. If possible, hold on a bit longer next time
If you’re sure the data is valid, you can manually override the PP value which you circled in your image and click the ‘Save’ button and Xert will update accordingly!
Congrats 
I wish I had 1300 W max power! haha!
LOL, who has time for 5-7 seconds of max effort??? 
I’ll have to move things around in my calendar.
We’ll have to see what can be done, need to convince my buddy that we should do it again; although I’m fairly certain we regretted our decision for the rest of the ride.
Yesterday I had a very good ride with a silver breakthrough. I did 3 longer intervals just under 10 min and one shorter run in the end.
I got a breakthrough in the first 3 runs (MPA was down around 7 min in the first 2 intervals and after 8 mins in the third interval). I was nearly at the limit but could have done a small sprint in the end.
Anyway, in the last interval I did a shorter run (4 min) with more power, but I couldn’t nearly achieve a breakthrough.
Do you think my Fitness signature is off or do I have just a low vLamax? I did mostly base work below threshold in the winter (apart from many illnesses). It seems that I can not achieve a breaktrough if the interval is shorter than 8 min.
I had my last breakthrough just 1 week ago as I did 4x 6 min intervals on the same climb.
Let my signature decay for a few weeks (went skiing, had some time off and ill). I usually breakthrough with A Pain (3 sets of 13x30/15’s).
Thought I’d try the 33/30 approach. I don’t think I’m quite as strong here; the recovery is slightly too long for me and so each interval needs to be a lot more anaerobic rather than the vo2max vibe of 30/15s. Averaged 515w across the 30 x 33s intervals, but honestly it wasn’t really my CV system that was overtaxed (except the final interval where I went for it); more just muscle fatigue in the legs.
This was a bronze, and brought my TP back up to 2w below where it was from the last breakthrough. I hope this means that the model of my fitness is accurate.
According to strava I set new PB 30 min power of 350w, but I’m not sure 30/30’s at 500w is the ideal pacing for a steady effort 
.
Ok, now something dodgy is going on. Second hard workout this week. Breakthrough earlier this week (last post) working hard put me at 365w TP from 33/30 drawdown. This feels about right (if I imagine an epic alpine climb of an hour; that might be my pacing aim).
Today:
(1650 / 27 / 375)
But I think - I don’t think I hit 1600W in those 10s sprints (most were 1000-1200w) - and when I look at the raw file I find a dodgy spike just AFTER one of the sprints. Ok, so → fitfiletools → remove spikes over 1300w.
Now the model thinks adjusts to 1340W PP / 28.5kj HI / 395TP.
30w on my FTP? 395W? No way. Even if you’re generous, I did 420w avg for 10 minutes in the first set of 30/15s, which would be a 380w ftp (if taking 90% of two efforts).
It also thinks my top end Z2 is 319w. Thanks, but I’m not Tadej fucking pogacar! There is definitely something out.
The only thing I can think of is that actually my HiE is significantly underestimated - i.e. I have really good repeatability above threshold but not actually that high a threshold.
But to get that to fit I have to go to a HIE of 38 and a threshold of 380w.
Any ideas?
There could be some other dodginess in the data, especially even minor ones that occur during the on/off periods (sticky power is common although that’s not necessarily what happened here).
You can always flag the breakthrough and/or manually adjust to something that is more within your expectations and see if it is reachievable using a different protocol or even a different power meter.
Power data isn’t always free of errors.
I went through it with a fine-tooth comb - there was only one bit of power data. All of the other intervals look consistent to what I was pushing live, and end at the right time.
I’m interested in how the model takes ~415 for 10mins into a 390+ TP. Unless I pull HiE really far out. I can bring the TP down to something approaching reasonable (375) but my HiE needs to go to 42kJ?
Because you had 20s rest between each 30s hard effort for 10 minutes. Your average power isn’t meaningful in this scenario.
Small errors accumulate. Efforts above TP are not 100% offset by efforts below TP (i.e. you can’t just take an average) so errors will bias a TP higher. This happens to some (not all) users. The sensitivity to accumulated errors is high with the analysis, even with the smoothing algorithm we employ. Adding more smoothing is an option but then other breakthroughs will get missed.
