Another way to do it that I’ve tried to use instead of looking at workouts and past rides (although they’re still useful) is to look at the whole context of the workout suggestion, not just XSS targets to help dial it in via the Focus field.
Basically I take in the blue, orange and red targets that can’t be tracked in realtime (yet) I also make sure to look at what it’s called (e.g polar breakaway specialist) and the description (e.g. Focus on attacking with strong power efforts with light pedaling rest in between)
Then on the free ride I aim to accumulate the target XSS. While having my focus field match for both parts (colour for pure, mixed etc; and time for rider type). So for polar breakaway that would be blue, which is polarised (I think) and around the 5:00 mark for breakaway specialist.
Using the description and XSS breakdown I guide my free riding to hit those targets in the ‘right’ way. I.e. am I smashing a hard sprint that puts my power number to red, then recovering easy so focus rises back up to 5:00 (to accumulate red XSS), or am I surging up to threshold riding with my power number yellow-ish to hold 2-3 mins at my 5 min (breakaway power) and then riding tempo for example which I believe would build more orange strain. I could also be doing both, like surging into orange and finishing with a sprint or sprinting and then holding (like a race).
It’s not a perfect hit on the numbers, but so far it’s gotten me close enough. If you’re keen you can also use the planner to sense whether you want to ride to make sure you hit high intensity targets at risk of going over them, or ride conservatively to not exceed them at the risk of going under. This is based on when your next hard session is.
These things are of course all linked somewhat too where if it’s the sprint version it might be called polar because lots of easy low XSS is needed to get that focus, whereas the threshold version might be mixed because you’re spending longer in yellow and less in blue as less recovery than smashing it.
Autogen question - apologies if it has been mentioned elsewhere.
I have a high intensity workout in the planner which gives me target intervals at vo2 max power (330w) etc. but the autogen feature plans a workout with almost 100w more power at a shored duration around 1min. I wonder which is better? I understand that higher power includes both red and yellow number targets, but would time spent at vo2 not be more beneficial here?
I would expect a further increae in training load, threshold, HIE etc. I have checked and the same happens with other event types. Perhaps this is a bug?
However, this behaviour is expected - notice how the ‘Event Readiness’ drops from 1 (Excellent) to 2 (Good) as you increased the duration.
What the system does as the TL’s become larger (and less-attainable from where you are vs. the target date) is to increase the # of days needed to recover after the event. The idea is the less prepared you are for an event, the longer it will take you to recover from it. Or put another way, athletes can accomplish a huge event with a less-than-ideal training load… it just comes at the cost of being red status for longer after the event Hope that makes sense!
I get the system but maybe it should not jump like that but do it gradually?
If 1.0 readiness is not possible let the user be 1.1 or 1.2 ready for a slightly longer event with the same values as he would have had to do to get 1.0 ready for the shorter event?
This jump to very different values after a slight change does seem strange.
Not really an option (at this point anyways) - adding a full extra day of red status recovery drops the necessary training load for an event quite a bit. Just going from ~2 days of post-event redness to 3 is already a 50% increase in post-event recovery time, which allows TL to be significantly lower.
You can ‘game’ the readiness by setting the longest duration event possible before the readiness drops. That should allow you to reach the highest TL possible at that readiness value.
In my case, the race is actually going to be about 6.5hrs long at a decent intensity. So I’m some way off with 3hrs
The plan I have currently was generated before the updated UI here, so not sure what ‘readiness’ I will have. Based on training & racing history, I’m sure I’ll make it through the race. But I honestly have no idea how I will do relative to others. I’m still intrigued by what the assumptions are “under the hood” for different races!
I’d also noticed this so good to get the explanation on why it behaves like that @ManofSteele. My issue with it but have been assuming ‘race’ mode will solve it, is that it leaves a lot of room for error or misinterpretation.
For example:
I am training for a 142 km 1200 m elevation gain event with a Cat 2 climb, 2x Cat 5s and a few other notable rises.
The waves you can sign up for are sub ~4 hours (35 kph+), 4.5 hrs (32 kph), 5 hrs (28 kph), 5.5 hrs (25 kph) and then 6+ hrs (~22 kph to avoid cutoffs). My goal is to be as fit and prepared as possible.
Using the Xert tools alone it would reinforce over interpreting my ability because I could say well I’ll sign up for the 4.5 hour option because Xert says I’ll be excellently prepared with high TP, HIE and PP achievements compared to if I was more realistic and said actually I can probably avg 26 kph at best so ~5.5 hr. Over that duration I might see I’ll only be ‘fair’ prepared and have relatively lower TP, HIE and PP outcomes.
Like I said race mode I’m hoping will solve because it should be able to tell me what is the best possible time I can do that 142 km in based on where I am now right? or at least that 4.5 hrs is not acheiveable?
I think both of our examples highlight a similar problem; what exactly are the physiological demands of a certain race result?
For something like a road race, where tactics etc are a huge part of things, there is no clear answer. I would guess similarly for CX or MTB racing, where skill is a big coponent. Of course, the winners in every event will still have excellent physiology.
However, for the gravel rides/sportives/gran fondos/time trials etc, things are slightly more controlled, and the race demands can be approximated relatively well. What we don’t know (but Xert maybe does) is how strong the opposition are. If Xert can leverage their data to base required pysiology on percentiles (for example), then instead of choosing “2 hours” or “3 hours”, you could instead choose “top 10% finish position”. To me, that would be cool!
For both of you, I suppose the ‘Race’ option will be what many users are looking for - combines the necessary training load and signature needed to compete at a higher level. For the ‘Event’ option (at the moment), it’s more designed from the perspective of what does it take to be adequately prepared - especially for users who may be looking to complete an ultra-endurance event.
Would you get better results with the AI forecast beta for achieving a Goal? Or still use the XATA with continious improvement? Are the benefits by using the beta (still beta so maybe some flaws?)
Continuous Improvement will still help you improve general fitness (centred around your selected athlete type), but I’d say that the ForecastAI will be much more effective in helping you reach a targeted goal (e.g. 5 min power) etc.
Good looking training program put together by XFAI:
I focus more on endurance efforts, so I set a 20 min power goal (320 W) by May 25 w/ a max weekly hrs of 28 hrs/week (the max value), though this program appears to max out only around 16.2 hrs. The only other things I’ve adjusted was to always take Sundays off, as well as slightly moving the ‘Recovery Demands’ Slider to the left (Less Recovery).
I don’t have any times that I’ll be away planned (YET), but I’m sure there will be times when I’m unable to train. I’ll see how the system adapts the program as weekend(s) become unavailable. This also doesn’t account for any outdoor (group) rides once the 2+ ft of snow melts & the weather warms a bit
Corresponds with the forecast circle colours in the fitness planner - Green meets the XSS recommendation, yellow indicates at least one of the XSS recommendations is close, but not fully met, & red indicates that the recommended training is quite a bit off of the recommendation:
Wow, that is a very nice forecast ;-). I also focus on endurance and have a 20min power goal (298W). But already in March…Maybe I have to get my goal date little further away in the past like your date.
I joined the beta and a couple of things stood out straight away:
The current week is pretty much empty.
Almost every Monday for the last 2 years I’ve done a 60ish xss zone 2 ride - xfai has taken the average ride time (including the hour sat chatting) and filled it up to 120 xss. Ideally I’d like to hint or restrict that day by xss. I could mark it as unavailable but I’d like to define xss given I know it’s a certain thing.
Does this look correct, 5 High Intensity days in a row? Or, 8 in 2 weeks.
I have set my goal date to early June. This seems to be a lot of High Intensity early in the stages of the period?
Not sure if this is easier for younger fitter riders, but as a 50 year old with less than stellar ability this looks like a high amount of training strain.
What are your complete Program and FAI Settings?
What is Availability set to? (flex or fixed, etc.)
What is your starting stars count and ending stars count?
Which weeks are shown? Example, weeks 7 and 8 out of 19.
Two changes coming in the next update should help.
You’ll be able to run a forecast and instruct it not to minimize the training needed to reach the goal. This should reduce rest days especially for longer training plans. (Btw, you can nudge XFAI to maintain a more consistent training plan for extended periods by lower the periodization value).
You can currently set your availability to 1 hour on Mondays which will restrict the XSS but you can’t currently force it to be low intensity. In the coming update, you’ll be able to choose a workout/activity on the Monday for the forecast and pin it as your training. You would then just pick that activity for each Monday and pin it, then run the Adapt process to plan everything around it. Pinning is a new feature that will work like planning does only it would apply to forecasted training.
Hope that makes sense!
