USR-2026-0017 · CMB

After look-elsewhere correction the joint large-angle boundary anomaly is 0.8–1.6σ. The 1.6σ is the most favourable end of that range, not a single answer: it assumes 5 effective independent looks (1.24σ at 10, 1.01σ at 15), and excluding the method-fragile alignment term leaves 0.75σ.

PROVISIONAL 0.8–1.6σ post-LEE — a range set by two stated choices, not a single number re-run & confirmed 2026-07-09 (R0 — founder CI; independence pending)

CORRECTED 2026-07-26: the registry published 1.6σ as the answer. Fisher-combining S_1/2 (p = 0.05) with the alignment (p = 0.03) gives 2.28σ pre-LEE, and the post-LEE value depends entirely on two choices — the trials count and whether the alignment is admitted at all (USR-2026-0014 flags it as method-fragile). The inputs are literature values, not this seam's own measurements (0.052 and 0.025).

Falsify-box — how to kill this claim

UNTESTED THE RANGE IS ALREADY ESTABLISHED, arithmetically, with no new data - I reproduced every number: p_s = 0.05 and p_a = 0.03 Fisher-combine to X = 13.00, p = 0.0113 (2.28 sigma), and the post-LEE value is 1.60 sigma at N_trials = 5, 1.24 at N = 10, 1.01 at the naive N = 15 that fig_joint itself names; dropping the alignment leaves 1.64 sigma pre-LEE and 0.75 sigma at N = 5. Both choices that pick 1.6 sigma out of that 0.75-1.6 sigma range are already known to be discretionary (N_trials = 5 hardcoded at run_boundary_deep.py line 243; the alignment flagged method-fragile by USR-2026-0014, whose figure gives SMICA 3.0 deg full-sky versus 23.8 deg masked, p 0.001 -> 0.085). A criterion of the form 'dead if N_eff >= 10 or if the alignment is inadmissible' has therefore ALREADY fired and cannot be registered as an open test - it is the reason for restatement. THE LIVE KILL TEST for the restated range claim: calibrate the trials factor by Monte Carlo instead of asserting it. For >= 10^4 LambdaCDM skies, compute every large-angle statistic this seam actually considered (D_2, S_1/2 full-sky, S_1/2 cut-sky, quad-oct alignment, parity ratio, and the l_c cutoff fit), Fisher-combine the same two the seam selected, and read the global p-value off the distribution of that combined statistic. The restated claim is DEAD if the MC-calibrated global significance falls OUTSIDE 0.75-1.6 sigma: above 1.6 sigma (1.6 was then not the favourable end and the published range is wrong low) or below 0.75 sigma (the floor is wrong and this seam's headline belongs under 1 sigma). Second, independent prong: the combined p-values must be this seam's own measurements (0.052 and 0.025, giving 2.33 sigma pre-LEE) rather than the hardcoded literature 0.05/0.03; DEAD if the seam's own inputs with the MC-calibrated trials factor land outside that same band.

Verification record — every quoted number, re-run 2026-07-09

quantitymeasuredverdict
pre-LEE combination 2.28σ (S_1/2 p = 0.05 with alignment p = 0.03) MATCH
post-LEE, N = 5 looks 1.60σ — the most favourable end MATCH
post-LEE, N = 10 / N = 15 1.24σ / 1.01σ MATCH
excluding the fragile alignment term 0.75σ MATCH

Provenance

script: boundary_seam/run_boundary_deep.py
script status: PRESENT
datasets: Planck SMICA · WMAP 9-yr ILC · claim-scripts (downloadable from /data — run it yourself)
re-run: 2026-07-09

Status, honestly

CORRECTED 2026-07-26 by the kill-test sweep; the audit flag it carried is resolved. Status here is computed from evidence — the author cannot set it, and neither can we. Independent reproduction would move it; nothing else will.

§ Figures 4 from boundary_seam/run_boundary_deep.py

What this claim's own script draws. These are the founder's committed outputs — the same plots the script regenerates on any machine that runs it.

deep_box_squeeze.png 109.1 KB download
deep_joint.png 222.4 KB download
deep_s_half.png 140.6 KB download
deep_skies.png 768.0 KB download
Share this claim X Bluesky LinkedIn Reddit HN Email