# PRE-REGISTRATION — bouted release against continuous, on published sleep constants **Batter's box item: _"The duty cycle: release on a timer, total at the wall (period 1/r; bursty release is worse)"_. Status: on deck. Its note: *"Burst release beats continuous release at equal average (hazard −13.5% and −17.7%)."* The owed next step, flagged as the **"pursue first" item of 13 September**: *"on the published sleep constants, prove that bouted release [beats continuous]."*** Written 2026-09-29, **before any number exists**. --- ## The claim under test **At equal total release, releasing in one long bout beats spreading the same release thinly.** The original claim said the opposite — bursty release is worse — and was retired. This asks the published constants. ## The model, fixed before running The **two-process model** (Borbély, 1982), which the house already lands on in the Night room. Process S is homeostatic pressure: - **Waking:** dS/dt = (1 − S)/τ_r, rising toward 1, with **τ_r = 18.2 h**. - **Releasing:** dS/dt = −S/τ_d, decaying toward 0, with **τ_d = 4.2 h**. Both are the standard published values. Nothing is fitted. ## The two schedules, at equal total release Over a 24-hour cycle, **8 hours of release in both arms**: - **BOUTED:** 16 h accumulating, then 8 h releasing, unbroken. - **DISTRIBUTED:** the same 8 h split into **n equal naps** spread evenly, n ∈ {2, 4, 8, 16, 24, 48}. Run to steady state (30 cycles), then measure the last cycle. ## What is measured 1. **Peak S** — the worst pressure reached. 2. **Mean S** — the time-integrated load. 3. **End-of-cycle S** — whether the schedule is sustainable or drifts up. ## The pass criterion, fixed now - **BOUTED WINS** if its **mean S** is lower than every distributed arm by at least **0.01**. - **DISTRIBUTED WINS** if any distributed arm beats bouted on mean S by at least 0.01. - Within ±0.01 across the board is **A WASH**, reported as such. - **Peak S is reported either way**, because a schedule can win on mean and lose on peak, and that trade is the interesting part rather than a nuisance. ## What is predicted, and why Decay is **proportional to S**: dS/dt = −S/τ_d. So release is most efficient when pressure is high. Bouting lets S climb before releasing, so each hour of release does more work. **I predict bouted wins on mean S and loses on peak S** — it is the schedule that tolerates a worse maximum to get a better average. ## Deviations Any change to the above is written here, with its reason, **before** the run it affects.