🗓️ 27062026 2200
How top athletes are actually coached — the training science the training_log and the
skating-coachskill reason from. The plan tells you what to do; these principles tell you when to push, hold, or back off.
autoregulation
A written plan is a starting hypothesis, not a contract. Elite coaching adjusts the day's load to how the athlete actually shows up — readiness varies with sleep, fatigue, stress.
Gauge each working effort with two cheap signals:
- RPE (rate of perceived exertion, 1–10): how hard the last set felt.
- RIR (reps / seconds in reserve): how much you had left before form would break.
Drive load from these, not from the number on the plan. Same plan, lighter day → do less. Same plan, strong day → the plan's progression is available, take it.
the quality gate
A rep only counts if form holds. Once toe height drops or posture breaks, you're past the effective zone — extra reps don't build the skill, they groove the wrong pattern and steal energy from the technical work that's the actual point of the session.
Quality is the cap, not the clock. End the set when form goes, not when the timer says so.
This is why chasing duration alone (e.g. 4 × 2:30 with the last set sagging) stalls: the junk reps at the end teach your body to wheel with a bent knee and weight on the ankle — the exact faults in wheeling.
progress / hold / back off
| Signal | Read | Action |
|---|---|---|
| Target work hit at RPE ≤ 7, form holds to the last rep, ~2 sessions running | Adapted | Progress one variable |
| Form holds but RPE 8–9, or mixed results | Not yet | Hold — repeat the load |
| Form breaks before target, RPE 9–10, or "too drained to continue after" | Overreached | Back off — more support / fewer sets / shorter |
Progress one variable at a time. Changing several at once hides which one drove the result (and is how quality collapses).
- Capacity order: remove support → add duration/reps → raise intensity (toe height / cones / speed)
- Skill order: raise success rate at current difficulty → add difficulty → perform under fatigue / inside a routine
discipline lenses
Not everything progresses like a hold. Classify what you trained, then apply the matching lens:
| Lens | Examples | Progress when… | Driven by |
|---|---|---|---|
| Capacity | toe wheeling holds, pt, fitness/strengthening | RPE drops at constant load | periodization + autoregulation |
| Skill acquisition | technical tricks, basics, spins | success rate rises, then holds under fatigue | ksj/practise_slowly, technical_tricks_matrix (basics set the ceiling) |
| Performance | choreo, improvisation | execution stays clean while expression opens up | huangke/improvisation_flow_state |
Capacity work is measurable and periodized. Skill work is about clean repetition and success rate — slow it down until it's right, then speed up. Performance is about doing the known cleanly under pressure.
monitoring
Watch trends across sessions, not single days:
- RPE at constant load drifting down → getting stronger, time to progress.
- RPE at constant load drifting up → accumulating fatigue, deload soon.
- The same form fault 3+ sessions (e.g. knee bending → fitness/dorsiflexion) → stop adding load, prescribe a targeted fix.
- Chronically "drained after" → overreaching; the session is too big to leave room for the skill work that matters.
deload
Every 3–4 weeks, or when fatigue markers stack up, cut volume ~40% while keeping intensity. Recovery is when adaptation locks in — see periodization.
process over outcome
bob_bowman_golden_rules: judge a session by whether you executed the details right, not by whether you hit a number. One focus per session ("What's Important Now"). The coach's job each day is to name that one thing.