Why elites skip the recovery week — and the 100→180 km/wk safety net

1 — Eight reasons elites don't need a weekly cutback

The reasons divide into "strong evidence," "moderate evidence," and "speculative." Almost all of them probably contribute — there's no single answer.

2 — The real framing: elites have recovery, just shaped differently

"Elites don't need cutbacks" is the wrong framing. The more accurate version is: total recovery dose is probably comparable; it's just distributed across a different time scale. Elites recover daily; non-elites recover weekly. Both work; both are calibrated to the granularity of monitoring the athlete has access to.

3 — Safety net for the 100 → 180 km/wk band

A calibrated upper-bound function for safe weekly progression in the Norwegian-double-threshold target range. Default ceiling: 180 km/wk (the Ingebrigtsen/Bakken anchor; user-overridable). The safety net is the envelope around the published plans we've benchmarked — if a user's proposed weekly step exceeds this curve, we warn.

The three components

// Calibrated to Pfitz 18/55, 18/70, 18/85 max-build-week steps
function maxBuildStepKm(v: number): number {
  if (v < 60)  return 6;     // Higdon novice band
  if (v < 100) return 6;     // Pfitz 18/55 range
  if (v < 140) return 5;     // Pfitz 18/70, 18/85 build
  if (v < 170) return 4;     // Sub-elite tier
  return 3;                  // 170-180 elite tail
}

// Inverted from current Tren code; matches Higdon family inversion (r=0.72)
function cutbackPct(v: number): number {
  if (v < 60)  return 0.10;  // Higdon Novice 1: ~10%
  if (v < 100) return 0.15;  // Higdon Inter, Hansons
  if (v < 140) return 0.15;  // Pfitz 18/85: 12-15%
  if (v < 170) return 0.18;  // Sub-elite
  return 0.15;                  // 170+: shallower (elites trend toward continuous)
}

// Above 150 km/wk, the discrete cycle dissolves into continuous EWMA control.
function useDiscreteCycle(v: number): boolean {
  return v < 150;
}

// Tren's default ceiling — anchored to Norwegian double-threshold target band
const TREN_DEFAULT_CEILING_KM = 180;

// User can override but loses our calibrated safety net (warning shown)
function overCeilingWarning(v: number): boolean {
  return v > TREN_DEFAULT_CEILING_KM;
}

4 — Algorithm envelope vs published plan ramp rates

The safety net should be the upper bound of safe progression. If our curve sits above the published plans' actual max-build-week steps, our model permits everything the plans permit (without prescribing them). If it sits below, we're more conservative than published practice.

5 — Multi-year ramp simulator using the safety net

Apply the safety-net rules week-by-week. Build 3 weeks, cut 1 week, repeat. Above 150 km/wk drop the cycle and grow continuously at the safety-net cap.

6 — Honest framing

• The safety net is an upper bound, not a prescription. It says "more than this in a build week is risky." It doesn't say "you must do this much."
• Calibrated to published plans only — not to RCTs. No prospective study has powered a contrast between, say, 80 vs 130 km/wk in trained adults. The curve sits on coach-authored plans (Pfitz, Hansons, Higdon, Lydiard) and elite case series (Tjelta 2011, Stellingwerff 2012, Kipchoge / Ingebrigtsen / Bakken).
• Above 150 km/wk the cycle dissolves on purpose. Kipchoge "only slight variation"; Norwegian model uses macrocycle deload (winter → in-season) instead of microcycle. Our continuous EWMA model is the right shape at this tail.
• 180 km/wk ceiling is a soft default. Users targeting the elite tail can override; we just lose calibration confidence above the published plan range. The product reason for the ceiling is "Norwegian double-threshold target" — Tren's stated coaching anchor.
• Frandsen L30 single-session cap matters more at the high tail, not less. A 40 km long run at 180 km/wk chronic could pass our chronic-based checks while exceeding the spike threshold. Ship Layer 2 regardless of any safety-net decision.