A few people have posted about newcomer plans being too easy for too long. You might say that it converges too slowly to the correct PLs. The advice is to manually select harder alternates, demonstrating PLs near the user’s limit, thereby simplifying AI’s task of finding those limits. (Thanks to those who asked and answered).
The alternates appear ordered by PL. I’ve just started on the build phase of my plan. Should I simply choose the hardest alternates I think I can complete for a few sessions, and then see how it responds after? Or would that be overdoing it during the build phase?
In other words, should the build phase be done at PLs near my limit? Or should something be held in reserve until specialisation?
But not so very hard that you can’t complete the next scheduled workout properly. Even, some say, if that means leaving an interval in the bag (i.e. finish when you know you could only complete one more interval, rather than completing it).
Look at your forthcoming workouts as a series. If you are certain that the series overall is too light for you, then up the PL accordingly. But don’t let a relatively small increase on one workout in isolation prevent you from properly completing the whole build phase.
ideally, you are just trying to get the AI to get more accurate faster. Once it’s closer to the right place, then it should help keep you managed. However, that will require you to be consistent with the surveys after each workout. I would not necessarily jump to the hardest you can do because this “convergence” doesn’t necessarily need to happen all in one jump. If you try a harder workout and it’s still too easy, pick a harder one.
Did you do a complete, 12 week base prior to build? If so, then your PLs should be a fairly close representation of your abilities and there’s no need to make adjustments.
If the PLs are still not accurate, then you can use alternates to increase them but I wouldn’t go beyond a stretch workout in build. Build is hard and a breakthrough or not recommended workout might lead to failures in the coming weeks.
As others have suggested, it’s good to err on the side of caution. What is hard but achievable (just!) for a week could be completely overwhelming by the end of week 3. I made this mistake late last year, advancing too quickly, & by the second week every workout filled me with dread, I was bailing from some workouts, & even cut the block short. It was an effective way to derail the end of a block & leave me intimidated by the thought of anything more. Actually, cooling was a big factor too & I was trying to get accustomed to higher workloads so I could participate in the full Disaster shenanigans. Did it too aggressively, on too short notice, so I’ve learnt my lesson.
What’s the structure of your build? If it’s the trio of VO2/threshold/sweetspot you could manually accelerate one workout type & let AT guide the others. Ideally what you’re most familiar with. That way if you get it wrong it’s not so big a deal. When I first started with TR I’d done a good bit of sweetspot outside of TR but no threshold or VO2, so I knew where to bump sweetspot to. Sure threshold started slowly but it was delivering me appropriately hard workouts in about a month anyway. In that block I felt like the sweetspot work was the bread & butter, & threshold was just a bit of extra time in zone until it caught up.
After joining TR, I scheduled a 3 month plan, of which 1 month was base. I skipped the ramp test, because I already knew my FTP accurately. My PLs seemed much too low throughout and after that base month. The idea of waiting for 3 months (12 week build phase @mtbjones mentioned) for TR to adjust to my fitness level isn’t hugely appealing.
Yesterday, I replaced the recommended Build phase, Sweetspot workout with Venado -1 (stretch), and on completion, my Sweetspot PL rose from 3.9 to 6.2. My HR hit zone 5 near the end of the 5th interval (of 6), so it was a good workout, but I’m not feeling depleted.
My other PLs remain a lot lower, so I may need a few more alternate substitutions to make coarse adjustments to those before letting AI take over the fine tuning.
My takeaway: TR / AI isn’t great at coarse adjustment, but presumably is quite good at fine adjustment (which I guess is what it was designed for).
One thing to know is that PL aren’t relative across workout types. For example, Sweetspot seems more manageable up to higher PL’s than Threshold. At 6.2, you probably have room to grow before an FTP bump in sweet spot, but with Threshold, 6.0 may be an upper limit before an FTP bump. You will start to get a feel for PL’s in a certain range for each type after a while. VO2Max is one that can go pretty high as well, but Threshold just seems to be stingier.
