S11 WOODHENGE — CURRENT READING GUIDE, v1.50
Authors: Lucas Giovani Ribeiro and Marcos André Simonssini

Read Supplementary Information S11 and main sections 2.5, 2.10 and 3.8.
The comparison is retrospective: the known approximate date of 2600 BCE
informed selection of the target (28,72,20). The complete subsequent family
has 5,832 numerical targets, 33 optima in 2635–2575 cal BCE, and one also
below 0.2 degree RMS. The same target ranks sixth without the date condition.
These counts are descriptive family denominators, not p-values or an all-sky
field census. Material transcription and conditional decoding are separate
post hoc constructions. The geometric controls and their adverse outcomes
are retained. Durrington remains indeterminate at input eligibility.

All 35 files of the earlier exploratory record are preserved byte for byte,
including its Portuguese reading notes, protocol and verification outputs.
Their statements about not editing the main describe that earlier stage;
the current manuscript integrates the comparison in v1.50. The local
SHA256SUMS.txt verifies those 35 files. The parent archive manifest also
covers this guide, dependencies, plot script and figure added at integration.
The protocol is a private record, not a public preregistration.

REPRODUCTION (run in a disposable copy; scripts overwrite their own results)
  python -m pip install -r requirements.txt
  python denominator_audit/verify_results.py
  python denominator_audit/audit_words.py
  python material_bridge/check_material_bridge.py
  python lattice_decoding.py
  python geometry_check/check_BC_geometry.py
  python geometry_check/check_BC_controls.py
  python plot_temporal_comparison.py

verify_results.py checks the archived rows without repeating the full
optimization. audit_words.py repeats both grids, minima refinement and the
interval-constrained comparison. See the archived protocol and audit summary
for numerical conventions and the boundary-local-minimum discrepancy.
The physical model uses the supplied sky_model.py, chronology_engine.py and
cygnus_hipparcos.csv. No third-party image is required to run these scripts.
The source plan and report are cited rather than redistributed. The digitized
ring coordinates and numbered-ring inventory are included.

SOURCE
Marshall et al. (2024), Historic England Research Report 94/2024,
Fig. 2 (numbered rings), pp. 14–15 (unitary construction chronology).
https://historicengland.org.uk/research/results/reports/94-2024
The six numbered rings total 156; the report describes 168 postholes overall.
No unassigned positions were inserted into the input vector. The benchmark
is the 95% construction interval modelled from rings A–C, not a posterior
mode or a separately measured chronology for every ring.
