Have you tried refreshing the page after setting the slider? It sometimes seems to l9ad outdated info from a cache and refreshing the page can help in those cases.
HI, yes, after changing the slider, I wait until all numbers are re-calculated, then I refresh the page, since that is the only way I found to see if the stars change color from the change of FF.
Some days I feel fresher, some days less fresh. Would be nice to be able to adjust FF to refelct how I feel this particular day and then just have the right type of workout served by the AI
This section (picture) change depending on the colors of the stars, when changing the FF slider, so that part works as I think it should work
Aha, I notice it has changed. Now the recommended workouts change after I change color of the stars between blue ↔ yellow when using the FF slider.
Move slider, wait for numbers to recalc, refresh page, new workouts proposed ![]()
Now it works as “assumed” Thanks!
What is the shortest intervals XERT uses to calculate TP in seconds?
I get the feeling that it is a very short window. That would explain why XERT overestimates my TP by approx 25w. For a light guy like me, that’s about 0.41 wkg
I had the same issue with intervals.icu with their default value they have. Since I am more anaerobic than aerobic phenotype it also gave me unrealistically high FTP in the beginning, only because my short burst power numbers are relatively high, vs my FTP/TP. Now I set that timer in intervals to 410 seconds and it is much more aligned with what I can actually do over longer durations. Is it possible somehow to tell XERT to use a larger min-window to calculate my signature profile?
If you do some very short all out efforts to breakthrough, it should increase your PP and HIE and reduce TP automatically… is your PP correct? Interesting to know your overall signature, and you can check that vs some rough norms
Are there errors in my Fitness Signature? – Xert (may have been posted before but still could be useful
How short are “very short” efforts?
Almost every sessions I do with MB, at some point Garmin-MB tells me I make a new 30 sek BT but it is never reflected in the training log for that session on Xert training Activities, but I think I read something about that already so that is not an issue.
I think above TP, PP, 3min Power are too high, approx 20-30 watts.
HIE is maybe too low but I am not sure about that.
Both Xert HIE and Intervals W’ fluctuate wildly based on the inensity of the activities recently have done last few weeks. In Intervals it swings changes between 30KJ and 16KJ. I would assume that HIE/W’ would be more stable value in it’s nature.
.
Occationally Xert thinks my TP is 286w while Zwift and Intervals place my CP at 263. This leads to Workouts having too high z2-z3 so I quickly die and cannot perform most of the high power intervals in the pursuiter workouts.
Going through my activities and check if I have any oddities in there or have faulty outliers, I am not skilled enough to identify those things. For now I turn down the power a bit between intervals so I can manage to complete the XSS for High+Peak and pay less attention to Low XSS. Maybe that totally destroys the purpose of the workout?
Are you sure it’s saying you have a new BT and not saying something like “BT if you hold this power for another X seconds”? That was confusing me for a little bit until I squinted enough to read it properly mid effort
Garmin says “new 30 sec BT detected” once in a while when I do my PP intervals pushes
For PP I’d probably sprint all out for 10 seconds…
Not sure for HIE, maybe out to 2 min…?
They do need to be all-out and when fresh, not just what you happen to hit in a normal training ride (as you mention getting a BT on most MB rides). Also, MB rides typically draw down MPA over quite a long time over multiple intervals, so while you are hitting all ‘systems’ including HIE / high strain, you will also be hitting TP / low strain in any BT you get in those. A single short effort to BT may yield a different result.
If that doesn’t clear it up you can write an email to support for more personalized advice, including helping hunt down data errors
Sounds like you are actually seeing the magic buckets message that you will hit a breakthrough IF you hold your current power for another 30 seconds.
OK: I will pay more attention to that message, you are probably right ![]()
As others have mentioned the XMB message refers to a potential BT. ![]()
If you continue to ride for 30 more seconds at current strain a BT is likely.
Go harder and you can shorten that estimate but you’ll be cranking up the pain.
If a BT is detected a BT badge will appear below HR.
Another way to look at it is signature validation. Just getting close confirm your signature is dialed in and that portion of the workout was very hard.
Xert doesn’t rely on MMP sampling to determine your fitness signature and power curve.
Your signature is derived by analyzing sec-by-sec performance under fatigue.
I don’t think you should try to match terms and numbers between platforms. The methodologies and math differs. They are not meant to match although they may be close for some.
Have you tried switching to Slope mode when the going gets too tough? It impossible to “fail” a workout that way especially if you monitor XMB on a Garmin while you execute the workout on EBC. You don’t have to match the intervals exactly to fill the buckets for the day. Same as it would work riding XMB outdoors.
I just subscribed today and wondering if I need to do anything for Xert to sync data more than 3 months old?
We recommend syncing historical activities from Strava! ![]()
Does Forecast AI not have anything that asks for subjective feedback about how I feel about my freshness or how a workout went and adapt accordingly? The platforms I’ve used before this would ask me to assign an RPE to my workouts after completing them and would also ask me how I was feeling before the workout. If I had to adjust the resistance of a workout or couldn’t complete the prescribed intensities, they would ask why and adjust my future workouts accordingly.
Does Forecast AI do anything similar or do I just need to know myself better and say “I don’t feel strong enough to do the prescribed workout today and I’m going to do something different.”
@ManofSteele Am I correct that FAI abandons any form of RPE feedback and I just have to trust it will adapt based on how I complete a workout or something else? I’m trying to wrap my head around that because every coach run and adaptive training plan platform I’ve used stresses how important it is to provide that type of feedback so the plan knows how well it is adapting to your abilities/fatigue.
Adaptations are based on whether you did more or less XSS than prescribed, and also more granularly (if that’s a word) in that it also looks at the mix of work (and form) across low, high, and peak loads. It doesn’t analyze workout compliance, RPE or HR or anything for this.
To adjust for how you feel you’re recovering you can adjust the recovery demands setting (permanently adjusts how aggressive / conservative the plan is), or freshness feedback (for temporary adjustments).
There are pros and cons of all the flexibility - definitely more self-coaching to optimize though
I don’t think that applies to Forecast AI plans, according to what I’ve read.
The freshness feedback slider isn’t available with forecast Ai plans for some reason.
Great questions @Mtnphotog .
Subjectivity in training is a how coaches and athletes would handle poor data analysis. Zones, TSS and using %FTP in general are pretty blunt instruments and two workouts with the same TSS or time in zones are rarely the same workout. Add in how trained athletes handle workouts differently than untrained athletes and you end up needing subjective data to fill in the gaps left by the data. A good coach will ask for subjective info and use it to adjust. Most often, they’ll adjust the exact same workout based on this feedback they get from you. Using feedback from one workout in fine-tuning another very different workout is a skill only advanced coaches have.
Xert most does away with the older, more blunt tools in assessing workouts / prescribing training and instead introduces new concepts like XSS in 3 dimensions, Training Loads in 3 dimensions. Recovery Load in 3 dimensions, workout difficulty and MPA. With this richer palette of information, we can make more data-driven assessments rather than subjective, guesswork on how athletes need to be trained. Coaches have a better idea if two workouts will be similar in difficulty or similar in training benefit. Subjective assessment is dramatically reduced.
Note that with Xert, you’re in control of much of what it’s doing too. This is where you apply your subjective (or personal data-driven method). For example, if you’re feeling fresh before workouts and the system often indicates a Tired or Very Tired status, you can reduce the setting for Recovery Demands . This instructs the Forecast AI to reduce the recovery time for the Recovery Loads in between workouts. If you’re finding you need to do more difficult workouts to complete the required training within the time you have to train, you can move the XSSR Preference (Workout Difficulty) slider to the left to instruct Forecast AI to give you less XSS to do in your available time and thus make workouts easier.
So keep these in mind. What’s most important is laying down a workable training plan with good doable workout targets with enough day-to-day whilst meeting your availability and personal riding preferences. Get these all set correctly and you’re good. Need to adjust? Adjust. Adapt Forecast. Everything adapts to your changes. The subjective feedback loop with uncertainty as to what will change and how it will affect you reaching the outcome you want is all but eliminated.
That’s great info! Thanks!
I would suggest taking a good chunk of that and putting it in the support article about Freshness Feedback to expand on the little note it says about how you don’t use it with FAI and you could also add more of that to the FAI support article.
![]()
