Hi, currently on Base-Build-Peak, but only because I cannot get Forecast AI to generate a plan. I’ve uploaded a FIT File for a race 16 May 26 (that I completed 17 May 25) but even at 16 hours per week (currently doing about 10 as it’s winter) is not enough it seems. Same thing happened last year when in Feb 25 I was at similar Fitness to now but it wouldn’t run a plan based on what I did in May 24 (same race). Yet come the actual race in May 25 these numbers (in the FIT File) were achieved! I’ve reset back to Base-Build-Peak for now. But why does it think I’ll not reach the numbers? For reference I’ve attached 3 files/charts: Progression Chart for 1 year, Target Information/Estimated Training Needs and Forecast AI resulting plan. The sharp increase in Threshold Power in August/Sept is when I did a few TTs and far exceeded what Xert determined was my Threshold Power. Same will likely happen this year I suspect.
Are you able to make sense of this and see why an AI plan cannot be generated for an event I do each year. Would really appreciate your help as this is an important training period.
I am interested in this also. I am riding a 200km gravel (Dirty Reiver) at the end of April. I tried running forecast AI but also got an event readiness of 5 and not able to generate plan. I have on multiple occasions ridden our local 100 mile road event at the same time of year so I am pretty confident that I am capable of doing the event (albeit I am planning to put in some committed training between now and then)
Hi @rhorn I’m also doing Dirty Reiver which I do every year. Was 2nd fastest in age cat 2 years ago and I’m usually top 5……but Xert reckons I can’t even do the event with a score of 5! The event above is the Gralloch UCI Race a few weeks later. I’m updating what I’ve sent above as I’ve also tried to create a Forecast AI by all the other means and no matter what I when it’s run it won’t generate a plan. Even when it looks ok before running the forecast AI. See screeb below…this looks ok but it failed to generate a plan.
Update: seems I had a quite a lot of Activities with Locked Signatures. Not sure why as I hadn’t locked them. So’ I’ve unlocked them and it’s increased by Threshold Power by 11w to move closer to what I think it really is. Looks about right now for this time of year. However still cannot generate a Forecast AI plan even when it says it looks ok. See below screenshot… this one failed to generate a plan
Hi johnnybike MAX WEEKLY HOURS set at 14.5 hours which is higher than my typical week I’ve set READINESS at 3 so a very comfortable level. Does WEEKLY AVAILABILTY affect it so must that be at least 14.5 hours?
I was wondering whether time was the limiting factor (for the model). If you up it to say 20 hours availability does that produce a workable plan?
What makes me think along those lines is that you have a TL of ~120 and it thinks you need 160. That sort of level is not sustainable to most people. Even a TL of 120 is quite high (to my mind at least)
@johnnybike hi yes I completely agree. I’ve input enormous hours and yes it produces a plan (but with a warning). It should be able to produce a plan even based on 12 hours. That Training Load is enormous for something I could go and ride today but with a slightly slower time. That’s what makes me think there’s a problem with the AI v my data.
I had set a max of 14hrs per week in Forecast AI. I had sketched out a rough plan in TP, but wanted to explore using Forecast AI combined with the XMB app to maximise the benefit from my commutes which account for a lot of my training mile but the route I take to work is not great for putting in structured intervals over the whole route (the first third and last third of my route are ok, but the middle is very stop/start through a town). In TP I had planned to max out on training volume mid March, partly as I am skiing for a week end of March (~35hrs training :o), and the max week was ~14hrs including a long gravel ride:
Yes @rhorn you’d think that volume would enough (I don’t know what fitness level you’re at) to generate a plan for Dirty Reiver. As part of your Dirty Reiver entry you’ve got a code to get 2 months of Wahoo SYSTM which includes a Dirty Reiver Training Plan. It’s not as science based as Xert (It’s a one plan fits all +/- wattage) but it requires nowhere near as much volume as XERT looks for the same event. I’m a 4 and bit star athlete on XERT…I hope the experts at XERT can come back to me soon. Meanwhile I have the Wahoo SYSTM training plan on standby as it’s served me well in the past.
Hi @IanJackson , thanks for the info. Do you feel that your current signature represents your current fitness? If you haven’t really pushed yourself in a while and your signature is underestimated, the system might think you need to do more training to achieve the goal signature for this event.
In the final image of the forecast chart, it looks like there are a couple days near your target event that have a magenta outline - I suspect those days are calling for more training time than you have available on that day of the week, which is throwing the training plan can’t be forecasted message.
You could still accept the program and prepare for a little extra training on those day(s). If you get a bit ahead of your training between now and then, it might not be an issue later! I’ve temporarily reduced my available training time for the winter months (no desire for 5hr trainer rides), but I will update with a new forecast once outdoor season returns. Perhaps a similar approach would work for you as well.
You can also use the drop-down arrows on the right-hand side of the Race screen to see what the best readiness value is for Low, High, Peak system before your target event. With an event of ~400 Low XSS, it will be very hard to achieve a low Readiness Score - looks to be a long & hard race!
Hi @ManofSteele. That was also my thinking after I’d posted the message. So I targetted a Breakthrough today which was Gold without too much trouble. In fact smashed PP by over 100 watts. Also just watched your Breakthrough feature on You Tube and have commented (you tube name is @highpike). Going to re run the AI plan and also going to amend how my signature decays which I learned about in your feature. I think I need to do that until I’m fully back up to usual volume outdoors. Thanks for the response and support
Hi @ManofSteele . See below which I’m sending just to confirm that I have successfully run a FORECAST AI plan that suits my needs but also to perhaps to help you and the team by way of of what I think are odd results when trying to generate an achievable plan. All have been run with the GENERAL SETTINGS at default. Any comments or observations would be welcome. Happy reading.
So I’ve run Forecast AI again for each of EVENT, GOAL, and RACE and have managed to create an achievable plan using EVENT.
EVENT
EVENT READINESS is manually set as 3, so middle ground. The number of HIGH INTENSITY days is 2 so that fits with my typical week, or how I like to train, although I did have to increase MAX WEEKLY HOURS to 15 hours (+25 to 50% over current volume). I may fall behind until the better weather comes. So, this is what I’m going to run with instead of BASE, BUILD, PEAK. The result is below.
For GOAL I set a 1 watt (yes 1w ) increase against each of the different types of FOCUS to try to create a short 6 week plan. I set the MAX WEEKLY HOURS to way above what I’m doing now 15 Hours (so plus 25-50%) and yet I get a warning message on the generated plan. I also note that 3 of my 6 training days are HIGH INTENSITY, I usually try to limit that to 2 days..at 62 years old I need some recovery . Most noticable of all is that when FORECAST AI is run, my numbers decrease from 325w to 321w. The result is below. Note the decrease in TRAINING LOAD despite the increase in volume and more HIGH INTENSITY workouts than currently.
With regards to RACE (using my 2025 FIT File for this race) then the hours required are enormous increases over what I’m doing now just to get an achievable plan. Have spent an hour or so just moving sliders (what ifs) without success, other than if I just about double my MAX WEEKLY HOURS. Alas, I gave up on the RACE option
Rest day tomorrow and then into the new plan on Tuesday
If not a reasonable uptick in one table or the other (or both), try pushing into yellow text warning or even red text and try running the forecast again. The schedule may end up tight but doable.
You can verify what’s projected to change – before saving – by viewing the forecast chart and isolate each key.
Example 60-day Goal –
Something needs to trend upward during the term.
If not, you are essentially creating a lateral progression.
The main problems you will run into is when Availability constraints conflict with where you want to go. Or the term is too short. For example, your TED (target event date) is only 42 days out. What is your current status stars count?
Xert’s model does indeed put an emphasis on increasing TL and hours during the term.
If you can’t do that the alternative is to pack more high/peak strain into the forecast such as raising XSSR.
Or more HIT per week which is common with XFAI plans since many HIT days will be "mild’ HIT (fresh in 24 hours) unless you purposely raise the bar on those days.
Is two HIT per week following a traditional polarized approach what you prefer to do?
The final point I’ll make is an XFAI projection is conservative. I have always reached my target watts or exceeded the goal during the times I’ve used XFAI. When that happens, you can keep going as planned or update the forecast with new settings.
My preference, however, is XATA Continuous or BBP, but I don’t have a rigid schedule to adhere to. YMMV
Hi @ridgerider2 and thanks for the comments. The date for which I’ve saved the EVENT Forecast AI is 16 May 26 so plenty of time. I think you’ve looked at the GOAL Forecast AI which indeed I ran for only 6 weeks as I set that at a modest 1w increase, just playing around with ‘what-ifs’.
With regards to the 2 days of HIT, yes that’s just my preference. I’m currently on just over 4 stars and really don’t need much improvement for the 16 May event, but if I have more potential then why not go for it
I ran the ADAPT FORECAST today as it had a red dot against it, possibly because it had added a very easy 27XSS workout for today on my REST day but workout was not done? I’ve just seen it, it wasn’t there yesterday
The plan now contains 4 days of HIT this week and a 1.8 READINESS. Looking forward in the planner it varies each week to between 2 and 4 HIT days.Interesting that you say the forecast is conservative which is good to know.
I must admit I also like XATA and keeping that needle pointing in the right direction . So, I’ve taken a look at Continuous (seeing as you mentioned it). I like the look of a 6 week plan which ends 7 weeks out from my first proper event this season. My events (apart from TTs which I treat as training) are all between 4 and 9 hours, so have set it as below. Not too fussed about PP, want to maintain HIE but with some Threshold improvement. This looks ok to me. Thanks for mentioning it
) increase against each of the different types of FOCUS to try to create a short 6 week plan. I set the MAX WEEKLY HOURS to way above what I’m doing now 15 Hours (so plus 25-50%) and yet I get a warning message on the generated plan. I also note that 3 of my 6 training days are HIGH INTENSITY, I usually try to limit that to 2 days..at 62 years old I need some recovery 
. Most noticable of all is that when FORECAST AI is run, my numbers decrease from 325w to 321w. The result is below. Note the decrease in TRAINING LOAD despite the increase in volume and more HIGH INTENSITY workouts than currently.
