I have a question regarding building my plan. ATM I’m training with chatgpt (a room created by someone else to act as a coach, based on scientific approaches) as a coach.
I’ll start racing around half March with a long season till september, october..
My coach is now giving me alot of endurance based rides, with some sweetspot and some torque efforts (because I asked). Once every two weeks there is a mild VO2 effort.
However the intensity is always really low. I train about 10-15 hours a week. (Depending on indoor (atm) or outdoor).
The bot really focusses on keeping HR drift low during the week, which leads to easy workouts.
Is this the right way to approach base season? Or should I train more diverse? I dont want to be flat, but I don’t want to peak to soon either… Which happened during previous seasons when I trained with apps like Trainerroad or JOIN…
I think the right way to train would be - don’t trust ChatGPT for this, it’s known to Hallucinate (Recent Studies show as high as 40% Hallucination Rate). You’re better off and will see better progress with a platform like TR, especially with the new (pending) TrainerRoad AI release.
If you think you peaked too soon, I think it comes down to adjusting number of intensity workouts per week, total volume, and the specific plan/event you have in the calendar and then respecting your rest/recovery times.
(Regarding ChatGPT or any other similar platform - they can be great tools. But, they frequently get stuff wrong, so you have to be critical and use it raise questions and approaches that you then are able to challenge and evaluate yourself. It’s not to be followed blindly, whereas platforms like TR are specifically designed for the purpose and trained on data relevant to the purpose)
I’m not using ChatGPT in its original form. When I ask how it functions it shows me it uses a pyramidal form now and switches over to a more polarised model closer to the season.
I feel that the knowledge and vision is pretty nice. But I dont know for myself how loaded my workouts now need to be.
I also started using Chatgpt because I found that other trainingplatforms always work with single day events. So they peak. Or they offer programs in blocks of 3 months.
If you’re referring to TR, I don’t think your characterization of “other trainingplatforms always work with single day events. So they peak. Or they offer programs in blocks of 3 months” is correct. I think it would handle this better than your ChatGPT setup.
Regarding your workouts: The recommendations it’s giving you could be decent, but also maybe not. Not enough information to really respond.
I too do a lot of Sweet Spot and Z2 this time of year, but intensity builds as I get into a winter build phase. February, March, and April for me will be focused around VO2 and Threshold work.
Torque work for me is very sporadic, more a specialty phase focus as I get closer to my target event.
I have never once trained based on Heart Rate Drift other than trying to keep Heart Rate below a certain level for some Z2 workouts, so no, I’m not sure that is the right focus.
Also - is Kermesse like a Crit? Short, fast, punchy? No idea what that means other than a quick google search.
If so - right now I’d be doing Base / Build, but as I got closer I’d be doing something like a Criterium Plan that had a lot more VO2, Anaerobic, and Sprint?
First of all, than you very much for replies and time.
Kermesse are longer than crits. Usually 75-120 km on higher intensity. But on a local circuit of around 5km or more.
Chatgpt is giving me one SS a week now 3 x 18min etc. The rest is base and one VO2 light. I then have to input the duration, avg watts,NP, avg HR, HR dirft and RPE. Or when blocks the same in the blocks. It then analyses and gives me a workout based on how well I did the workout and how it was.
Well, if they’re 75-120 KM, they’re not higher intensity as compared to crits, but I think I get what you’re saying. Probably a 2-4 hour effort?
My only changes to what I posted above would be the specialty phase, I’d still do Base and Build focused on trying to raise FTP. For specialty Not as much Sprint and Anaerobic Focus, and more extending Time in Zone for Sweet Spot, Threshold throwing in some Over/Unders and VO2 work to prepare for surging during the race. I think the equivalent TR Event Characterization for their plans would be “Rolling Road Race”?
Neither of those are really relevant - time is the first variable that matters (1.5-3 hours based on average speeds and Distances?) Then your NP needs to be in context of IF or percentage of your FTP.
But, for those time ranges, recommendation stands. Rolling Road Race, extended time at Sweet Spot and threshold, with over/unders, and VO2 Bursts throughout the event to account for in Specialty. Unless you’re racing it as a TT - in which case it’s basically Sweet Spot the whole way.
Belgian Kermesses are literally one of the hardest formats of racing there is. I have fond (terrible ) memories of doing some in the 90s.
My view is that your winter training will be driven by when you want to be in the best shape, eg if you want to be flying in spring and or want a couple of peaks you’d need more high quality in winter. Then rebuild etc for second peak. If you just want to be in decent shape more or less all season (not easy) then I’d be trying to do as much volume as possible in winter to build biggest base possible (obviously with some intensity as races approach) - I’m guessing that is what your program is trying to do..
TR can get you into great shape too but it has a lot of intensity. But based on the length of your season you’ll get plenty of intensity in races o If I had to choose between TR and your program I’m choosing yours
Maybe on low volume it has to but if you have the time for the volume then you wont get too many intense sessions and if you plan the events in and put them in your calander in advance even better.
I think your chatGPT plan sounds sensible it’s what anyone would suggest if you asked them to look on the internet and report back.
) memories of doing some in the 90s.
it’s what anyone would suggest if you asked them to look on the internet and report back.