ASPDEN: stage-by-stage digits of agreement with today's value (provenance.py version 2)

W1 1/α (measured 137.035999177)
   struct 108π(8/N)^(1/6); N = 1843 is selected by his stated rule (odd, at or below the zero-potential value     137.035914847  D =  6.2  added 6.2
   total D = 6.2; target 0.0, input 0.0, struct 6.2, choice 0.0
   digits beyond chance (per formula): 2.9 to 3.9   [working paper 11.2: 65-650 structural variants, 4 rounding rules]
   note: pure structure, but the mechanism neglects effects of 10-30% (working paper 11.4), and its assumptions were assembled in 1972, when α was known to about 1 ppm

W4 proton/electron mass ratio (measured 1836.15267343)
   struct P = Q/(√(3/2) − 1), Q = 944784π²/1843^(4/3); no new adjustable number (1975)                              1836.152316  D =  6.7  added 6.7
   total D = 6.7; target 0.0, input 0.0, struct 6.7, choice 0.0
   digits beyond chance (per formula): 2.3 to 2.6   [working paper 11.2: 464,100 lattice formulas]
   note: closer to today's value than the 1975 measurement was

W5 G (measured 6.6743)
   struct his G from the lattice and the 1.875 cluster (struct, history unverified)                                     6.67216  D =  3.5  added 3.5
   total D = 3.5; target 0.0, input 0.0, struct 3.5, choice 0.0
   digits beyond chance (per formula): 0.1 to 1.2   [working paper 11.2 (116 cluster variants), scaled from a 34 ppm miss to his 321 ppm]
   note: his published G moved from 6.6676 (1966) to 6.6705 (1986), 6.672 (1997), 6.686 (1998) and 6.67435 (2000, 'guided by the empirical data'); whether the changes followed the measurements is not shown (working paper 7.2)

W6 deuteron magnetic moment (measured 0.8574382335)
   choice the fraction 6/7 = (2 + 4)/7, set by the state weights 2:1:4                                           0.857142857143  D =  3.5  added 3.5
   struct plus the mass-deficit term of his chain-state model (1986)                                               0.8574389111  D =  6.1  added 2.6
   total D = 6.1; target 0.0, input 0.0, struct 2.6, choice 3.5
   digits beyond chance (per formula): 0.16 to 2.1   [working paper 11.2 (7,680 combinations of states and weights 1-4), scaled from 0.33 ppm to his 0.79 ppm]
   note: working paper 8: most of the moment comes from 6/7, fixed by weights the chance test treats as free

P01 muon/electron mass ratio (measured 206.7682827)
   target the integer 207, the muon mass ratio rounded (1983 gives no derivation of it)                                   207.0  D =  3.0  added 3.0
   struct three-charge formula 209 − (9/4)·207/(207 + δ), δ = √3 from his stated geometry (1983, P02)             206.768670416  D =  5.7  added 2.8
   choice δ = 1 + 2π/9, a 'wave resonance' (1983, P01)                                                            206.768307765  D =  6.9  added 1.2
   total D = 6.9; target 3.0, input 0.0, struct 2.8, choice 1.2
   digits beyond chance (per formula): not estimated
   note: he tabulated δ = 1.688-1.744 against the measured range before picking the resonance value. Another reading of 207: his lattice 'virtual muon', 206.3329, has 207 as its nearest odd integer; on that reading the target share falls

P04 muon anomaly (measured 0.001165920715)
   input  the measured electron anomaly, which his formulas take as input                                           0.001159652  D =  2.3  added 2.3
   struct 1982 derivation: resonance distance 2(R + r)                                                             0.0011658682  D =  4.3  added 2.1
   choice 1983 revision (P05)                                                                                        0.00116589  D =  4.6  added 0.23
   choice 1984 revision: average over two states (P04)                                                              0.001165918  D =  5.6  added 1.1
   total D = 5.6; target 0.0, input 2.3, struct 2.1, choice 1.3
   digits beyond chance (per formula): not estimated
   note: the 1983 and 1984 revisions each moved toward the measurement of their day

P06 electron anomaly (measured 0.00115965218059)
   struct cavity formula 2/(1 − δ), δ = 1/(2π/α + 4/√3), α from N = 1843 (1981)                                  0.001159646913  D =  5.3  added 5.3
   choice plus a gravitational term of 5.281 × 10⁻¹⁰                                                               0.0011596522  D =  7.8  added 2.4
   total D = 7.8; target 0.0, input 0.0, struct 5.3, choice 2.4
   digits beyond chance (per formula): 1.5   [working paper 9.3: his cavity formula among 21 natural variants; the gravitational term adds about 0]
   note: the result equals the 1981 measured value, 0.001159652200(40), in every printed digit

P24 neutron magnetic moment (size) (measured 1.91304276)
   struct his model's bare moment, 2 μN (the measured antiproton moment is 2.793 μN)                                        2.0  D =  1.3  added 1.3
   choice times (1 − 1/23): 44/23, from his four-state model (46 transitions; 23 is the largest prime factor o    1.91304347826  D =  6.4  added 5.1
   total D = 6.4; target 0.0, input 0.0, struct 1.3, choice 5.1
   digits beyond chance (per formula): 3.0   [Part 2 block E: 20 fractions with denominators up to 46 lie in 1.90-1.93 (as one formula; about 1 in 30, i.e. 1.5 digits, allowing for the 36 particle numbers tested)]
   note: the 46 comes from a four-state model with several free choices, which he called tortuous; the same populations give a neutron-proton mass difference 63σ off. Still within 1.6σ today

P25 proton magnetic moment (measured 2.79284734463)
   struct 1980: Physics Unified eq. (199), with his own α⁻¹ = 137.0359 and mμ = 206.7689                               2.792846  D =  6.3  added 6.3
   choice 1986: cavity resonance N/n = 9.373948                                                                       2.7928467  D =  6.6  added 0.32
   choice 1988: 2:1 mixture of the 205 and 207 resonances                                                           2.792847068  D =  7.0  added 0.37
   choice 1988: plus 1.07 × 10⁻⁷, his proton–electron charge difference, fitted to the proton/electron mass re      2.792847367  D =  8.1  added 1.1
   total D = 8.1; target 0.0, input 0.0, struct 6.3, choice 1.8
   digits beyond chance (per formula): not estimated
   note: each revision moved toward the measurement of its day; the charge difference in the last step is excluded experimentally (working paper 6.2)

P10 charged pion mass (in mₑ) (measured 273.1324399)
   struct (π:P:π:π) system, π/P = 0.14874412, non-resonant (he notes it is 6σ off)                                      273.117  D =  4.2  added 4.2
   choice resonant pion, bare half the time (1989)                                                                   273.127432  D =  4.7  added 0.49
   total D = 4.7; target 0.0, input 0.0, struct 4.2, choice 0.49
   digits beyond chance (per formula): not estimated
   note: his 1980 value, 273.1262, used no integer and reaches D = 4.6

P13 neutral pion mass (in mₑ) (measured 264.143008513)
   target integer search: B/C = 2641/10 ('best C = 10'), the π⁰ mass to 4 digits                                          264.1  D =  3.8  added 3.8
   choice plus the surplus at B/8A, the lower end of his range, which he preferred                                       264.11  D =  3.9  added 0.11
   total D = 3.9; target 3.8, input 0.0, struct 0.0, choice 0.11
   digits beyond chance (per formula): not estimated
   note: the upper end of his range (P13b, 5.1 digits) was picked in this study with today's value in view, so it is not counted

P20 Λ mass (measured 1115.683)
   choice 2^(1/4) × the proton mass; the factor is empirical, matched to the known Λ (1986)                       1115.79984456  D =  4.0  added 4.0
   total D = 4.0; target 0.0, input 0.0, struct 0.0, choice 4.0
   digits beyond chance (per formula): 1.9 to 2.0 (local and global toolkit density)
   note: one of 13 published matches; on 10 states he did not address the toolkit does no better than chance

P07 tau mass (measured 1776.93)
   choice 3^(7/12) × the proton mass; empirical factor, matched to the known tau (printed 1781 MeV, 1986)                1781.0  D =  2.6  added 2.6
   choice revised 1996 after the 1992 measurements (P08)                                                                1776.51  D =  3.6  added 0.99
   choice revised again 1998 (P09)                                                                                      1777.02  D =  4.3  added 0.67
   total D = 4.3; target 0.0, input 0.0, struct 0.0, choice 4.3
   digits beyond chance (per formula): 0.49 to 0.65 (local and global toolkit density)
   note: beyond chance is for the first (1986) value

P21 Ω⁻ mass (measured 1672.45)
   choice (τ + 3Q)/3^(1/3); empirical, matched to the known Ω⁻ (1986)                                                    1674.0  D =  3.0  added 3.0
   total D = 3.0; target 0.0, input 0.0, struct 0.0, choice 3.0
   digits beyond chance (per formula): 0.88 to 1.0 (local and global toolkit density)

P18 J/ψ mass (measured 3096.9)
   choice (√6 − 3/2)·2^(1/3)·g, with g = 2.587 GeV fitted in 1966 to the Σ, K and π; published 1978-79, after            3095.0  D =  3.2  added 3.2
   total D = 3.2; target 0.0, input 0.0, struct 0.0, choice 3.2
   digits beyond chance (per formula): 0.84 to 1.2 (local and global toolkit density)

P27 W mass (measured 80.3625)
   choice 2(2N + 1)^(4/3) × proton mass with the integer N = 8, published after the W was found (1983)                    82.03  D =  1.7  added 1.7
   total D = 1.7; target 0.0, input 0.0, struct 0.0, choice 1.7
   digits beyond chance (per formula): 0.59
   note: scored against his threshold series (spacing about 16%), not the toolkit, which stops at 12 GeV

P31 Z mass (measured 91.1879)
   choice 1984: same formula on the graviton, N = 4, after the Z was found                                                96.86  D =  1.2  added 1.2
   choice 1986: f(2) − f(1)                                                                                             93.7935  D =  1.5  added 0.34
   choice 1989: (g/ω)³ g                                                                                                  92.59  D =  1.8  added 0.27
   choice 1997: minus 2 × 0.785 GeV                                                                                       91.02  D =  2.7  added 0.92
   total D = 2.7; target 0.0, input 0.0, struct 0.0, choice 2.7
   digits beyond chance (per formula): 0.11
   note: beyond chance is for the first (1984) value; about 0 after four attempts

P14 neutral-pion lifetime (measured 8.43)
   struct (2/1843) × his electron 'lifetime' 7.55 × 10⁻¹⁴ s (1987: 8.19 × 10⁻¹⁷ s; the 1989 paper gives 8.2 by              8.2  D =  1.6  added 1.6
   total D = 1.6; target 0.0, input 0.0, struct 1.6, choice 0.0
   digits beyond chance (per formula): not estimated

P22 neutron mean life (measured 878.3)
   choice 1981: simultaneous muon hits, condition (3), the one that matched 918 ± 14 s                                    921.8  D =  1.3  added 1.3
   choice 1986: dual-encounter formula, 'exactly the median of the recommended value'                                     898.0  D =  1.6  added 0.34
   total D = 1.6; target 0.0, input 0.0, struct 0.0, choice 1.6
   digits beyond chance (per formula): not estimated


QED AND THE STANDARD MODEL: stage-by-stage digits of agreement with today's value

electron anomaly (measured 0.00115965218059)
   struct Schwinger term α/2π, with α measured by rubidium atom recoil                                           0.0011614097318519  D =  2.8  added 2.8
   struct plus electron loops to fifth order (computed coefficients)                                             0.0011596521757262  D =  8.4  added 5.6
   struct plus muon and tau loops (lepton masses are Standard Model parameters)                                  0.0011596521784737  D =  8.7  added 0.36
   input  plus hadronic part (collider data) and electroweak part                                                  0.00115965218023  D =  9.5  added 0.77
   choice α from rubidium rather than cesium: D 9.5 against 9.1 with the alternative, so 0.44 digits move from 'struct' to 'choice'
   total D = 9.5; target 0.0, input 0.77, struct 8.3, choice 0.44
   note: the choice of rubidium is not the most favourable: the weighted mean of the two α values would give D = 9.8; the 2024 revision of the five-loop coefficient moved the prediction away from the measurement (D 9.6 before, 9.5 after)

muon anomaly (measured 0.001165920715)
   struct QED (α and lepton masses)                                                                                  0.001165847188  D =  4.2  added 4.2
   struct plus electroweak part                                                                                      0.001165848732  D =  4.2  added 0.0092
   input  plus hadronic part from collider data (2020)                                                                 0.0011659181  D =  5.6  added 1.4
   choice leading hadronic part from lattice QCD instead (2025)                                                       0.00116592033  D =  6.5  added 0.83
   total D = 6.5; target 0.0, input 1.4, struct 4.2, choice 0.83
   note: the switch removed a discrepancy of 4.2σ (2021), 5.2σ (2023) and 5.8σ against the 2025 final measurement; the lattice result (2021) preceded Fermilab's final result but followed BNL's (2006); about 23 of the 223 × 10⁻¹¹ change came from the light-by-light term

hydrogen Lamb shift (measured 1057.8298)
   struct QED with a point proton                                                                                   1057.6937967019  D =  3.9  added 3.9
   input  plus the proton-size term, radius 0.84075 fm (muonic hydrogen dominates it; the muonic-only radius g            1057.8322  D =  5.6  added 1.8
   choice the smaller (muonic) proton radius rather than the 2014 value: D 5.6 against 4.9 with the alternative, so 0.76 digits move from 'input' to 'choice'
   total D = 5.6; target 0.0, input 0.99, struct 3.9, choice 0.76

W mass (measured 80.3625)
   struct lowest order from α, the muon lifetime (G_F) and the Z mass                                               80.939228555988  D =  2.1  added 2.1
   input  with α run to the Z mass (uses collider data)                                                             79.966616526599  D =  2.3  added 0.16
   struct plus loop corrections with the top and Higgs masses (Standard Model parameters; Gfitter 2026)                     80.3558  D =  4.1  added 1.8
   total D = 4.1; target 0.0, input 0.16, struct 3.9, choice 0.0

neutral-pion lifetime (measured 8.43)
   input  the anomaly formula, with the measured pion decay constant                                                           8.38  D =  2.2  added 2.2
   struct plus two-loop corrections with estimated constants                                                                   8.04  D =  1.3  added -0.89
   total D = 1.3; target 0.0, input 2.2, struct -0.89, choice 0.0
   note: the higher-order result moved away from the measurement

neutron mean life (measured 878.3)
   struct Standard Model relation, with V_ud (a Standard Model parameter) and the lattice g_A                                894.17  D =  1.7  added 1.7
   total D = 1.7; target 0.0, input 0.0, struct 1.7, choice 0.0

lattice QCD hadron values (4 constants fixed once, to the π±, K and Ξ masses, and the J/ψ for charm):
   proton/electron mass ratio         1832.0  D = 2.6  struct 2.6
   (mₙ − mₚ)/mₑ                         2.95  D = 0.78  struct 0.78
   neutral pion mass               264.25886  D = 3.4  input 1.5, struct 1.9
   Λ mass                             1114.0  D = 2.8  struct 2.8
   Ω⁻ mass                            1676.0  D = 2.7  struct 2.7
   neutron moment (size)               1.893  D = 2.0  struct 2.0
   proton moment                       2.739  D = 1.7  struct 1.7


SUMMARY (Aspden)
  W1    1/α                                D  6.2  target     —  input     —  struct   6.2  choice     —   beyond chance: 2.9 to 3.9
  W4    proton/electron mass ratio         D  6.7  target     —  input     —  struct   6.7  choice     —   beyond chance: 2.3 to 2.6
  W5    G                                  D  3.5  target     —  input     —  struct   3.5  choice     —   beyond chance: 0.1 to 1.2
  W6    deuteron magnetic moment           D  6.1  target     —  input     —  struct   2.6  choice   3.5   beyond chance: 0.16 to 2.1
  P01   muon/electron mass ratio           D  6.9  target   3.0  input     —  struct   2.8  choice   1.2   beyond chance: not estimated
  P04   muon anomaly                       D  5.6  target     —  input   2.3  struct   2.1  choice   1.3   beyond chance: not estimated
  P06   electron anomaly                   D  7.8  target     —  input     —  struct   5.3  choice   2.4   beyond chance: 1.5
  P24   neutron magnetic moment (size)     D  6.4  target     —  input     —  struct   1.3  choice   5.1   beyond chance: 3.0
  P25   proton magnetic moment             D  8.1  target     —  input     —  struct   6.3  choice   1.8   beyond chance: not estimated
  P10   charged pion mass (in mₑ)          D  4.7  target     —  input     —  struct   4.2  choice  0.49   beyond chance: not estimated
  P13   neutral pion mass (in mₑ)          D  3.9  target   3.8  input     —  struct     —  choice  0.11   beyond chance: not estimated
  P20   Λ mass                             D  4.0  target     —  input     —  struct     —  choice   4.0   beyond chance: 1.9 to 2.0 (local and global toolkit density)
  P07   tau mass                           D  4.3  target     —  input     —  struct     —  choice   4.3   beyond chance: 0.49 to 0.65 (local and global toolkit density)
  P21   Ω⁻ mass                            D  3.0  target     —  input     —  struct     —  choice   3.0   beyond chance: 0.88 to 1.0 (local and global toolkit density)
  P18   J/ψ mass                           D  3.2  target     —  input     —  struct     —  choice   3.2   beyond chance: 0.84 to 1.2 (local and global toolkit density)
  P27   W mass                             D  1.7  target     —  input     —  struct     —  choice   1.7   beyond chance: 0.59
  P31   Z mass                             D  2.7  target     —  input     —  struct     —  choice   2.7   beyond chance: 0.11
  P14   neutral-pion lifetime              D  1.6  target     —  input     —  struct   1.6  choice     —   beyond chance: not estimated
  P22   neutron mean life                  D  1.6  target     —  input     —  struct     —  choice   1.6   beyond chance: not estimated

SUMMARY (QED and Standard Model)
  electron anomaly                   D  9.5  struct   8.3  input  0.77  choice  0.44
  muon anomaly                       D  6.5  struct   4.2  input   1.4  choice  0.83
  hydrogen Lamb shift                D  5.6  struct   3.9  input  0.99  choice  0.76
  W mass                             D  4.1  struct   3.9  input  0.16  choice     —
  neutral-pion lifetime              D  1.3  struct -0.89  input   2.2  choice     —
  neutron mean life                  D  1.7  struct   1.7  input     —  choice     —

Aspden, all 19 formulas together: target 6.7 digits (7.6%), input 2.3 digits (2.6%), struct 43.0 digits (48.0%), choice 36.0 digits (41.0%)
QED / Standard Model, the 6 staged predictions: target 0.0 digits (0.0%), input 5.6 digits (19.0%), struct 21.0 digits (74.0%), choice 2.0 digits (7.1%)
QED / Standard Model, with the 7 lattice values added: target 0.0 digits (0.0%), input 7.1 digits (16.0%), struct 36.0 digits (80.0%), choice 2.0 digits (4.5%)

Do revisions go both ways? Aspden: every revision in the ledger moved toward the measurement of its day.
  Standard Model: the 2024 five-loop revision moved away (D 9.6 -> 9.5 ); the two-loop pion lifetime moved away (D 2.2 -> 1.3); the 2025 muon hadronic switch moved toward (D 5.6 -> 6.5)

wrote out/provenance_*.md and out/provenance.csv
