If you do read this, don’t expect a concrete conclusion, I’m not running a controlled experiment. This is a reflection on letting TrainerRoad do more, and spending less time monitoring and evaluating my training data.
This year I’ve leaned into trusting TrainerRoad to set the routine, get on with riding and using features like RLGL, adaptations, and TrainNow to guide.
My main constraint: two hours commuting by bike, twice a week. I need tools that work around that and help me avoid overdoing it. The flexibility TR has built in over the past couple of years has been getting progressively better for working with my lifestyle.
Motivation & Goals
There’s a perception on here that TR is for racers — I’m not currently racing. My “A” event is 1–3 weeks in the French Alps every August.
As an aside: my partner also uses TR, training on relatively low hours for big bike tours. She has also been making great improvements using the new features.
Shift in approach
Before 2025: spent quite a bit of time looking at the PMC fitness chart on intervals.icu, posted occasionally in the PMC score thread, and prescribed most of my own sessions mixed with TR — lots of z2 early in the year, then sweet spot closer to my event. Time training and TSS guided.
In 2025: Around February, I stopped looking at intervals.icu entirely and let TR prescribe my work. My goal shifted to hitting at least two hard sessions per week. If RLGL downgraded a hard session, I usually accepted it (unless I was about to have several days off, a holiday, etc.).
TrainNow: I started the year with no Progress Levels and used the “smarter” TrainNow for a few weeks to “test” it and set those levels before adding in a plan. Most suggested workouts were very hard but given it was (presumably) working from outside, unstructured data, it was not far off. Based on that, I’ve made a habit to pick alternates to take the edge off, especially when revisiting an energy system I haven’t touched in a while (I always need to work into Vo2).
I still self coach a little, but find it easy to adapt day by day without getting sucked into too much analysis.
Caveats
Worrying about PMC scores early in 2025 might have influenced my training outcomes.
In 2024 I started from a lower FTP than normal, after my longest cycling break in maybe 10 years (a bit over two weeks off).
Total hours is somewhat useful but a couple of slow and easy bike tours can really ramp the hours up.
Year stats heading into the Alps trip:
| Year | Distance | Hours | FTP Change | w/kg | +Watts | 6wk TSS Avg | TSS Peak |
|---|---|---|---|---|---|---|---|
| 2024 | 2,900 m | 244 h | 261 → 284 | ~4.3 | +23 W | 59 | 72 |
| 2025 | 2,400 m | 185 h | 275 → 294 | ~4.5 | +19 W | 56 | 59 |
Year on Year TSS - chasing hours/TSS in 2024 after a longer than normal Xmas holiday.
2025, trusting TR from February:
Reflections
I’ve enjoyed the new and continually improving features. I do still self coach a little - listen to my body, etc. and the tools make it easy feed this in.
Feel I’ve achieved a lot this year on fewer hours by keeping intensity consistent but limited, without feeling the need to keep adding z2 volume.
RLGL and TrainNow have made it easier to balance training with life while avoiding overreach. More time has been spent on the PlayStation and less on looking at charts.
Happy to be sitting at 4.5 w/kg on ~6-10 hours per week (where ~4 are commuting). The real proof is in the outcome and how good I’ll feel in the high mountains next week 
This is likely my peak for many years and it could well be a ceiling based on current training.
