@huges84 thanks for taking the time to challenge the post. You’ve raised some fair questions, and I’ll do my best to respond.
- Personal Experience ≠ Universal Truth
No, I’m not suggesting my experience applies to everyone. But there may be corollaries for others with similar training backgrounds. My observations are based on my own long-term experimentation and response to different training blocks.
- On Intensity and Sweet Spot Load
Sweet Spot work has been a uniquely effective stimulus for me, especially as a time-crunched athlete. It delivers the best bang-for-buck, and I’ve built up to higher volumes slowly and incrementally, with careful attention to rest and recovery.
If I have ~10 hours/week to train, I’ve consistently found that the old SSB plans offer the most reliable path to fitness gains.
Before Adaptive Training, TrainerRoad coaches advocated this exact approach. There’s a great back catalog of early podcasts that discuss it. If you haven’t seen them I’d recommend a watch\listen. eg. How Much Sweet Spot Training Should I Do Per Week? (Ask a Cycling Coach 284)
On a lighter note I can’t help but smile thinking back to the early days when Nate Pearson was training he and the coaches were all vocal proponents of the Sweet Spot Base plans.
One thing I didn’t mention in the original post is my training goal: general aerobic fitness, targeted toward Gran Fondos and long, hilly endurance events. For those kind of events SSB adaptation and bumps in FTP are really helpful.
- Are the FTP Differences Meaningful?
I agree, average FTP isn’t a great metric. I used it to provide a quick snapshot, but what matters more are the trends and context. I tried to highlight those in the table I shared.
It’s early days since returning to the older-style training, but the results already look promising. After ~4 years of following Adaptive Training + RLGL, I was either flat or slightly declining.
For context, I’ve always kept:
• TrainerRoad Aggressiveness: Max
• Masters athlete setting: Off
If I followed current TR guidance to the letter, my CTL would be half what it is now, and I expect I’d be regressing.
- Role of AI in This Post
Totally fair point AI can be terrible with numbers. I used ChatGPT to help structure the post and organize thoughts, but all the data analysis and interpretation is my own. AI helped make the post more readable, not draw conclusions.
- PD Curves and More Meaningful Metrics
Agreed, those would offer a much deeper look, especially for performance modelling. I’ll aim to compile my Power Duration curves across seasons soon.
A concern I have with the current model is that it seems heavily geared toward optimizing the bell curve of compliance across the entire athlete base. That’s great for new users or those struggling with consistency. It risks leaving well-adapted athletes behind.
For those of us who’ve already built up to a higher training load and have our nutrition and recovery dialed in, the system pulls us back toward the middle. It’s a one-size-fits-most model that may be suppressing gains in the upper end of the athlete spectrum, not out of bad design, but out of over-conservatism. It’s for this reason that personally I’ve retaken control of my training plan. I’m sure the AI models will eventually get to where they need to be. Perhaps TrainerRoad AI v2 will truly take over from the humans.