# PRE-REGISTRATION — The Climb Out, attempt 2: is the published number in simulation steps? Written 29 September 2026, after attempt 1's hypothesis was refuted, before this run. ## Attempt 1 — FAILS, and my hypothesis was wrong `PREREG_climbout_prefactor_2026-09-29.md` predicted the **exact** double integral would give the published 2,294, and that entry 29's 18.6 came from evaluating Kramers' asymptotic outside its domain (ΔU/D = 1.60). **Refuted.** The exact integral gives **23.19** at σ = 0.3 and **180.49** at σ = 0.2. The asymptotic gives 18.65 and 138.27. **The two forms agree within 25%** — so the approximation was never the problem, and my diagnosis was wrong. Both disagree with the published pair by 99%. **The verdict FAILS stands and the hypothesis is recorded as refuted.** ## ⛔ And it exposes an error in entry 29 that must be corrected Entry 29 says the gap is "123.3× at σ = 0.3 and 130.2× at σ = 0.2 — **not constant, so not units**". That was measured against the **asymptotic** form. Against the **exact** integral the gap is **98.9× and 99.4×** — constant to **0.5%**. **My conclusion that it is not a unit conversion was reached from the wrong baseline, and entry 29 needs amending on the page.** ## The new hypothesis A constant factor is a unit conversion. The formula returns **natural time**; the room reports **cycles**. If the room's numbers came from a simulation stepping at dt, then cycles = time / dt, so dt = exact integral / published cycles giving 23.19 / 2294 = 0.010109 and 180.49 / 17938 = 0.010062 — **both ≈ 0.01**, a wholly ordinary integration step. ## Criterion, fixed before the run 1. **One step size explains both.** The implied dt at σ = 0.3 and at σ = 0.2 agree with each other to within **2%**. Two independent σ landing on one dt is the test; a single ratio would explain nothing. 2. **dt = 0.01 reproduces both published numbers.** `exact integral / 0.01` matches 2,294 and 17,938 to within **3%** at both σ. Three percent because the room's own simulation sits 4.5% from its formula value, so demanding tighter would claim more than the published pair carries. 3. **Reported, not scored.** The same check against the room's *simulated* numbers (2,397 and 18,496), and the implied dt from those. **WHAT KILLS IT:** the two implied dt values disagreeing by more than 2%. Then it is not a single time step, the factor of 99 is unexplained, and the open item stays open and is named as a real discrepancy in the room's published numbers rather than a unit convention. **Prediction:** holds, with dt = 0.01. **Recorded plainly: I have now been wrong twice about this item — first that it was not a unit conversion, then that it was the asymptotic's fault. A third wrong reading is possible and this criterion is set so it would show.** ## Discipline No prose conclusion in the script. Attempt 1's FAILS and its refuted hypothesis are published beside this.