How to Test or Challenge the Theory

Step 1 — Select a qualifying book and first sentence

• Use the first complete sentence of a real book published before January 1, 2019.
• Identify the title, author, publication date, publisher, edition, original language, and source of the text.
• Use the sentence from that identified edition in its original published language.
• State how its sentence boundary was determined, especially if editions or punctuation differ.
• The candidate may not reproduce Genesis 1:1 character for character or simply reproduce all of its character-by-character numerical values.
• For formal eligibility, disclose how the book and candidate were selected and confirm compliance with the computer-use rule.

Step 2 — Fix the text, encoding, and challenge version

• Record the exact first sentence, source edition, transcription, encoding, and language rules before examining its formal results.
• Do not add, remove, reorder, or selectively shorten words or letters.
• Do not switch editions, translations, spellings, punctuation, transliterations, or numerical systems after seeing results.
• Disclose treatment of spaces, accents, capitalization, final letter forms, punctuation, and nonalphabetic symbols.
• Preserve copies of rules.htm, allrequirements.htm, all linked definitions, and the verification software used.

Step 3 — Independently document and freeze the attempt

Do not email or register an attempt before testing. Before examining the formal results, create and retain a dated record that identifies:

• The candidate book, edition, and bibliographic information
• The exact original first sentence and proposed sentence boundary
• The prepared transcription and numerical encoding
• The treatment of punctuation, spaces, accents, and special letter forms
• The challenge and software versions being used
• How the candidate was selected and whether computers or artificial intelligence were involved

Preserve dated copies of the applicable Rules page, all 304 requirement definitions, linked supporting materials, and verification software. Those records identify the version tested; no acknowledgment from the website owner or judges is required or provided. Later additions or substantive revisions cannot be imposed retroactively on a properly documented attempt.

Step 4 — Prepare the program input

Place the prepared candidate in the supplied file:

If Hebrew text is used, apply the published procedure consistently:

Retain the exact input file and a readable copy of every transformation used to create it.

Step 5 — Download and preserve the verification materials

Use the current all.zip, including the replacement Pattern 91 test. Record download dates and versions, keep the files unchanged, and preserve complete outputs. Identify the source and version of every mathematical-constant or genome dataset used.

Step 6 — Run TBP-001 through TBP-100

From the extracted program directory, run:

Preserve the complete unedited output: every pass, failure, error message, program version, input, and date. A failure means the candidate has not completed the first-100 stage. Passing all 100 is a major intermediate result, but 204 documented requirements remain.

Step 7 — Optionally reproduce the random-sentence control

To request one trillion generated sentences, run:

A shorter declared number may be used for a verification run. Report the requested and completed trials, generator version and settings, seed when applicable, runtime, and complete count reaching each pattern.

Reported earlier 3/7/37/73 one-trillion-sentence result

Final 2/5/25/52 generated-sentence result

Detailed survivor counts were retained for 10,000,000,000,000 trials. These results describe the supplied generator and fixed test sequence. They are descriptive experimental results and should not be interpreted as a universal probability for all sentences or all published-book openings.

The observed frequency of passing the first 20 tests was 3 in one trillion, or about 1 in 333 billion, under that generator and fixed test sequence. This is a descriptive experimental frequency, not a universal probability for all sentences.

Applying that rate to an estimated 150 million published-book openings would give an expected value of about 0.00045, but that is an extrapolation, not a direct test. Real opening sentences may differ materially from the program's generated distribution.

Step 8 — Evaluate TBP-101 through TBP-304

Use the frozen definitions and supporting materials linked from allrequirements.htm:

• Follow each linked page's exact input, calculation, rounding, indexing, comparison, and pass criterion.
• Apply every defined 25/52 counterpart without switching between controls according to the result.
• Treat every designated composite as one complete package; its component equations do not become separate requirements.
• Obtain a written clarification before testing whenever a frozen definition is materially ambiguous.

Step 9 — Document the book-level and personal counterparts

• For each of the 20 comfort subjects, cite an identifiable passage in the candidate book by edition and page or stable location.
• For each of the 24 extra subjects, cite the candidate book's relevant naturalistic explanation and address the objective standard published on that page.
• A keyword, passing mention, or unrelated passage is not sufficient.
• Submit the evidence supporting the equivalent personal experience required by angel.htm.
• Separate verifiable documents and testimony from interpretations, inferences, and material that would be available only if later requested.

Step 10 — Prepare the complete report

The submission must include:

• Complete bibliographic identification of the candidate book and edition
• The exact original first sentence and prepared input
• The encoding, transcription, and sentence-boundary rules
• The dated records establishing the independently frozen challenge version
• How the candidate was selected and all computer or AI use
• Every candidate, edition, variant, encoding, and parameter examined
• Program and dataset versions, commands, and complete unedited outputs
• A TBP-001-through-TBP-304 checklist documenting a claimed pass for every requirement, with no failures, unresolved ambiguities, or unsupported claims of inapplicability
• All book citations and personal-experience evidence
• Enough material for independent reproduction and adjudication

Step 11 — Prize submissions are not open

No monetary award is currently offered, and no prize entries are being accepted by email or otherwise. Do not submit a claim for prize eligibility. If a funded award is formally launched, its official rules will specify the entry channel, deadlines, and supporting documentation. Work performed before launch does not reserve eligibility or priority. General research questions and independent criticism may still be sent through the site’s ordinary contact channels; they are not prize submissions.

Step 12 — Independent review and unanimous adjudication

• The panel consists of exactly three impartial, appropriately qualified judges.
• Judges must disclose relevant personal, professional, and financial conflicts.
• Objective calculations are independently reproduced. Subjective equivalence is decided only where a frozen requirement expressly delegates that decision to the judges.
• Every final adjudication must be unanimous. A two-to-one vote cannot complete the challenge or approve an associated reward.
• If unanimity is not reached, the disputed matters remain unresolved and the challenge is not declared completed.
• The panel is not required to review entries in submission order, diagnose an unsuccessful entry, identify where it failed, or issue a requirement-by-requirement rejection.
• Written notice is sent to the challenger only when the panel unanimously determines that every one of the 304 requirements has been satisfied.