Bijlage B
The Numbers: the Vault, the Method, and the Silence
The forensic core that God's Algorithm (Appendix E) entrusts to this book: the complete letter-audit, the gematria analysis, the structural architecture, and the statistical weighing of Daniel 9:24-27. In this book the numbers are a secondary matter — they confirm an exegesis that already stands on words, grammar, and canon. But a secondary matter is no license: this appendix fixes the rules, shows the count, and reports just as loudly what the text refuses to count. Whoever wishes to verify will find here the complete toolkit.
1. The method, fixed in advance
The church has been branded by number-mysticism and has a right to suspicion. That is why three rules apply here, fixed in advance and applied identically — exactly the discipline that God's Algorithm (Appendix C) and the text-critical literature demand.
Exact, or nothing. No rounding, no ±1, no letter dropped or added to make a sum come out. A number that only works out "approximately" does not count.
Content-bound. The number must land on the content of its own location and must square with the theology of that clause. A number that bears on nothing bears on nothing.
The test of the silence. The strongest evidence is not what the text counts, but what it refuses to count (§6). Whoever reports only the hits and conceals the misses proves nothing.
Overt versus secret gematria
The scholarly literature distinguishes two kinds (E. Goodfriend, Jewish Bible Quarterly 2024; critically M. Bar-Ilan, Association for Jewish Studies Review):
Overt gematria — both the number and the word are present in the text and are congruent in content. The classic example is Genesis 14:14, where Abraham's 318 servants coincide with the name Eliezer (אליעזר = 318). Defensible, because both the number and the word stand in the text.
Secret gematria — a number that appears only after the analyst chooses counting units, draws boundaries, or adjusts spelling forms. Largely arbitrary, because the degrees of freedom do the work.
The look-elsewhere trap
This is the decisive statistical correction, and most number-books ignore it. With enough degrees of freedom — which counting method (mispar hechrachi, katan, or one of a dozen others), which counting unit (words, letters, cola), which starting and ending point, raw or skeleton, qere or ketiv — almost every total lies next to a meaningful word. "Nearness" therefore proves nothing; it is inevitable. (Illustration: the body-total of 9:24-27 lies right next to both 358 = Mashiach and 359 = Satan; see §7. And the ELS test in §6 makes this hard-measurable.) The only defense is: minimize the number of degrees of freedom, and report the silence completely.
2. The verified count
Recalculated afresh from the Hebrew text itself (standard mispar hechrachi), per verse, for both the raw form (BHS/B19a Ketiv, all letters) and the SSER skeleton (the consonantal backbone after reduction of the matres lectionis). Independently verified against a variant ledger of the Leningrad Codex (B19a) and against a second, separate count: all 34 cola check out exactly, raw letters and gematria.
| Verse | Words | Raw letters | Raw gematria | Skeleton letters | Skeleton gematria |
|---|---|---|---|---|---|
| 9:24 | 23 | 96 (ketiv) / 95 (qere) | 6572 | 88 | 6500 |
| 9:25 | 22 | 97 | 8348 | 87 | 8268 |
| 9:26 | 21 | 86 | 6938 | 80 | 6882 |
| 9:27 | 20 | 80 | 5872 | 71 | 5794 |
| Total | 86 | 359 / 358 | 27,730 | 326 | 27,444 |
3. The pillars (Tier 1 — hard and content-bound)
(a) 86 words = Elohim (אלהים = 86). The word-count descends in a flawless staircase 23 · 22 · 21 · 20 — no step missing — and sums to 86, the numerical value of the divine name Elohim (א1 + ל30 + ה5 + י10 + ם40). The container pronounced over "your people and your holy city" is built to the measure of the Name. Decisive: this signal has zero degrees of freedom — no matres to strip, no boundary to shift (the pericope is the pericope), no qere/ketiv choice (words, not letters). This is the "one whole number, one target" type that evades the look-elsewhere trap. The strongest find of all.
(b) v. 26 raw letters = 86 = Elohim (a second landing). The wound-verse, where the Anointed One is cut off, counts 86 letters — a second, independent landing of the Name, now over the place where redemption is juridically accomplished.
(c) 34 → 40: ga'al → Goel. The cutting-off clause 26a-c — "and after the sixty-two sevens an Anointed One shall be cut off, and He has nothing" — counts 34 letters, exactly the value of גאל ga'al, "to redeem" (ג3 + א1 + ל30 = 34). The sentence in which the Anointed One loses everything is precisely as long as the word for the deed of the redeemer. And by one added letter — the waw (hook, connection, becoming-flesh) — ga'al (34) becomes גואל Goel (40), the Redeemer Himself. The deed and the Person, separated by the letter of the incarnation. (And the 40 = Goel lands also on the desolation-block 26e-h: the Goel is at the same time the gō'ēl haddām, the avenger of blood who lays the desolator waste — redemption and blood-vengeance in one span.) (A wink, INTERPRETED: the Anointed One is cut off at the age of 34 — born in the autumn of 5 BC, crucified on 14 Nisan AD 31 — and He began at 31, the value of אל El (Three Days and Three Nights, chapter 10). That is one coincidence, not two: with 3½ years of ministry the 34 follows from the 31. An age is a calendar ordinal, like AD 34 and 2034 below.)
(d) The 34-motif, woven threefold. The 34 does not stand alone: the oracle falls apart into 34 cola (9+10+8+7), and three coherent units each count exactly 34 letters (8 words each: 25a-d, 25e-g, 26a-c). One number — that of redemption — is at once the count of the building blocks, three times a panel-length, and the gematria of ga'al on the redemption-clause. That is convergence, not coincidence; it "breathes" the ga'al principle (see §10).
4. The year-numbers (lowest tier — honestly tiered)
The most vulnerable kind: numbers that look like a year-number. They remain in because they land each time on precisely the right clause, but they bear the weakest marks — at most weighed, usually speculative — with the weaknesses named aloud.
536 (24d, lekhalle hapesha, "to close the transgression"): hard as a sum (raw = skeleton, no reduction needed) and a defensible content-fit — 536 BC is the year of the return of the remnant, in which the exile-for-sin came to an end. The strongest of the year-numbers, but the coupling "completing the transgression ↔ the return" asks one interpretive step → weighed.
457 (25c + the directional lamed of 25d): a signal, provided one takes the directional unit. The gematria of the "from [...] to" construction מן מצא דבר (427) + the lamed of 25d (30) = 457 — exactly the starting year of the count (457 BC). Content (the going-forth command = the starting point) ↔ number (the starting year). Not the gematria of a whole colon, but of the grammatical directional unit; the only "cost" is the inclusion of one grammatically necessary letter. Weighed, tending toward well documented. (Not to be confused with 457 as a chronological anchor, which is well documented in any case: 457 + 483 = AD 27.)
2520 (26f-h, "until the end [...] desolations are decreed"): raw gematria = exactly 2520 = 7 × 360 = the "seven times" (Leviticus 26) and 2 × 1260. The end-time number on the end-time clause. A strong content-fit, but 2520 is round and much used (high look-elsewhere), and it leans on the 360-day measure (see Appendix D). Disputed — offered in its strongest form, the most content-rich of the gematria year-numbers, but resting on a contested measure.
The end, twofold. The closing sentence 27f-g ("until the consummation is poured out upon the desolator") has raw = skeleton = 1794 (granite, no matres-reduction). Within it stands 794 = the Jewish year 2034 in perat katan (the count without the thousands); the leading 1000 in 1794 is unexplained. The skeleton-sum of the whole of verse 27 is 5794 (raw = 5872) = the full Jewish Anno Mundi year for 2033/34 — here the "5000" is thus explained, but the sum is matres-dependent (lowest category). Speculative.
206 = dabar. The distance between the starting dates of the Jewish count (year 1 = autumn 3761 BC) and the genealogically corrected count of God's Algorithm (year 0 = autumn 3967 BC) is exactly 206 years. The year-numbers themselves differ by 205: the Jewish year 5794 (2033/34) is 5999 on the clock, and AM 6000 begins in the autumn of 2034 (Jewish 5795). And 206 is the numerical value of דבר dabar, "word" — precisely the starting word of the clock in 9:25. Striking and content-tight, but dependent on the two AM-anchors: speculative of the highest order, not a load-bearing pillar.
Ga'al (34) on the Gregorian calendar. The Redeemer-number lands also on AD 34 (the end of the 490) and 2034 — on a calendar whose zero-point itself stands askew (Jesus born in the autumn of 5 BC). The weakest mark, speculative: a coincidence with a human calendar ordinal, presented as a providential wink, never as proof.
5. The noted resonances
358 = Mashiach = Nachash. The body of the prophecy counts 358 letters (qere) — and both משיח Mashiach (Anointed One) and נחש Nachash (serpent) carry that value: the two combatants of Genesis 3:15, woven into the same number. Deeply apt in content. But it holds only on the qere (95 letters → 358); the written ketiv gives 96 → 359 (= ha-Satan). Because this book reads the written consonantal body, 358 is marked as speculative in the deeper register, not load-bearing, until the qere/ketiv choice is decided on independent text-critical grounds (see Appendix A).
Numerical echoes: וְאֵין (we'ein, "and nothing", 26c) = 67 = נָגִיד (nagid, "Prince") — the Prince who has nothing, in one number; and עֹלָמִים (olamim, "everlasting", 24g) = 190 = קֵץ (qets, "end", 26g). Fine antithetical echoes; weighed.
The Janus density. Through 9:24-27 runs a striking series of words that look both ways (see Appendix E). That this is a deliberate feature of a "sealed" oracle is a strong thesis — but it requires a baseline measurement. Until that control-measurement exists: speculative, not proof.
6. The silence — the backbone
The real test of whether these numbers are designed and not projected is what the text does not do. That is why, straight through the count, it is reported just as hard what is not there.
330 (l'khapper, "to atone") — a beautiful number, which would square with the content — has been deleted. The book initially claimed the skeleton-total = 330; the B19a ledger gives 326. Better to lose a beautiful number than to make a claim that a reader can undo with a pencil.
1260, 1290, 1335, 490, 483 — the numbers expositors would most love to find — are, as a letter-sum in 9:24-27, absent. (One exception, honestly reported: 2520 = 2 × 1260 is present on 26f-h; see §4.)
ELS / "Bible code" — tested and negative. A complete scan for equidistant letter sequences (skips 2-50, both directions) yields "hits" such as ישו, יהוה, דוד — but the chance-baseline unmasks them: the real text (~997 hits of ≥3 letters) performs within the range of 200 random reshufflings of the same letters (mean 631, maximum 1147). ELS here is therefore indistinguishable from chance. Reported just as hard as the pillars.
No formal mirror symmetry. The weight-curve (colon gematria) has been tested for palindromic symmetry: r ≈ −0.14 (within the margin of chance). The text is not a numerical chiasm; the order lies in the spine, the inclusios, and the strophes (§10), not in a mirror.
The grand totals (27,730 raw / 27,444 skeleton) carry no name and no clean factor: tested, and empty.
That is not embarrassment; that is the backbone. A pattern-hunter text lights up everywhere; this one does not. The finds that remain (86, 34/40, and the architecture of §10) are scarce, exact, and content-bound — and it is precisely the silence around them that makes them signal rather than chance. The signature is real because the paper around it is blank.
7. An honest little calculation (the statistical weighing)
Why does 86 count as a pillar and 358 not? The difference is the number of knobs you were allowed to turn.
Take the weak kind first. How many meaningful Hebrew words lie between, say, 350 and 366 — a finger's width on either side of a body-total? Dozens, among them Mashiach (358) and Satan (359). With so many targets in a narrow band, and with the choice between raw (359) and qere-skeleton (358), a large total can almost not fall beside something meaningful. That both 358 and 359 land on a name is no wonder but the look-elsewhere trap itself: nearness is inevitable and therefore proves nothing.
Set 86 against it. It is a word-count: no matres to strip, no boundary to shift, no qere/ketiv choice — zero degrees of freedom. One whole number, no wide band full of targets, and it does not fall just beside something but on Elohim, the Name that ought to stand above "your people and your holy city," via a flawless staircase with no missing step. That is a different kind of event: a bullseye without your being able to adjust the weapon. The same holds for 34/40 on the cutting-off clause. The distinction between pillar and resonance is therefore not "prettier" or "uglier" — it is statistical: how much freedom of choice did the analyst need to make the number appear?
8. Summary tier-table
| Finding | Status | Mark |
|---|---|---|
| 86 words = Elohim; staircase 23·22·21·20 | ✅ closes, zero degrees of freedom | Well documented (pillar) |
| 34 → 40 = ga'al → Goel (on 26a-c) | ✅ closes, on the cutting-off | Well documented (pillar) |
| 34-motif: 34 cola + 3× 34-letter panel | ✅ convergent, raw | Well documented (structural) |
| the architecture (spine, inclusios, strophes; §10) | ✅ multiple convergence | Well documented (structural) |
| v. 26 raw letters = 86 = Elohim | ✅ closes | Well documented |
| 457 (directional unit 25c+lamed) | ⚠️ requires one grammatical letter | weighed |
| 2520 (26f-h, end-time clause) | ⚠️ round number, 360-day measure | disputed |
| 206 = dabar (distance between the starting dates of the counts) | ⚠️ dependent on the AM anchors | speculative |
| 358 = Mashiach = Nachash; we'ein=nagid=67 | ⚠️ qere / small numbers | speculative |
| Janus density | ⚠️ baseline measurement needed | speculative |
| 536 / 794 / 5794 (year-numbers) | ⚠️ content-fit variable / skeleton | weighed / speculative |
| ga'al 34 on AD 34 / 2034 | ⚠️ calendar ordinal | speculative (a wink, not load-bearing) |
| 330 = l'khapper | ❌ deleted (ledger = 326) | lapsed |
| 1260 / 1290 / 1335 / 490 / 483 as letter-sum | ❌ absent | the silence |
| ELS hits (ישו / יהוה / דוד [...]) | ❌ within chance (chance-baseline) | the silence |
| formal mirror symmetry | ❌ r ≈ −0.14 | absent |
| Grand totals (27,730 / 27,444) | "empty" — no name/factor | negative, correct |
9. Methodological conclusion
The strongest signals (86 = Elohim; 34/40 = ga'al/Goel) are macro, exact, content-bound, and virtually free of degrees of freedom — exactly the type the strict rules permit. Alongside them stands a structural architecture (§10) that leans not on gematria but on counting, division, and repetition, and that converges under several independent lenses. Everything below is honestly downgraded: weighed to speculative resonances that note but do not bear, year-numbers with the weakest marks, and a fully reported silence — including the demonstrably negative ELS and symmetry tests. This does not undermine the thesis — the numbers are a secondary matter in any case — but it is the integrity that the book claims. Better an honestly downgraded number than a Tier 1 claim that goes under with a pencil. The exegesis hangs on the words; the numbers and the structure merely seal it.
10. The architecture — structure alongside gematria
Alongside the individual numbers the pericope carries a multiply ordered construction, independently verified from the raw text. The evidential force lies in convergence — word-count, letter-count, colon division, key-word repetition, inclusio, and content all point again and again to the same seams:
The macro-skeleton. The four verses count 23-22-21-20 words (a flawless descending staircase, sum = 86 = Elohim) across 9-10-8-7 = 34 cola (= ga'al). Name and redemption, in the count of the building blocks themselves.
The decree-spine and the center. The root of decreeing (חרץ/נחת) stands at word 3, 43, 65, 83 — just after the opening, the exact center, and just before the close. Word 43 is וחרוץ (charuts, "decreed"); the central letter (#180 of 359) is the ח of that word. Word-count and letter-count bisect on the same word, and that center is also the thematic turning point: from restoration to cutting off.
Two inclusios. Kalah ("to complete") frames the text (word 9, "to finish the transgression" ↔ word 82, "the consummation"); and the decree-root frames it likewise (3 ↔ 83). At the end the text folds itself doubly shut.
Parallel panels (equal word- and letter-counts): 3× 8w/34L (25a-d, 25e-g, 26a-c, the salvation timeline), 2× 10w/44L (26a-d ∥ 27a-d), 2× 5w/22L (27a-b ∥ 27c-d), and verse 27 = 10 + 10 words. These panels exist only in the raw text and vanish under skeleton-reduction — the structure lives in the written consonantal text.
The arc of tension. The weight (gematria) per colon peaks on the run-up before the cross (26a = 1708, the heaviest) and falls to the absolute minimum on the cutting-off (26c, "and He has nothing", = 103, the lowest colon of all 34). The form enacts the content: the emptiness-clause is literally the emptiest line.
The six goals split 2-2-2 / 3-3-3 (taking sin away ↔ bringing righteousness in), with "to seal" (חתם) at the mirror positions 2 and 5, and the first triad = 26 letters = YHWH.
The edges. The very first letter (שׁ) and the very last (ם) together form שם ("Name") — the envelope of a Name-oracle; and the first colon (1322) weighs almost the same as the last (1300). Light (two letters), but noted.
Not loose lucky hits, but a coherent construction in which count, division, and content cover one another — landing each time on the theological core: God (86) has decreed (the spine) the redemption (34), issuing in the judgment upon the desolator (the close). This is, alongside the two pillars, the strongest answer to "signal or eisegesis": an architecture is harder to explain by chance than a single gematria.
11. The ELS test and the alphabet-completeness — honestly weighed
Two much-heard claims, tested to the bottom.
ELS (equidistant letter sequences). A complete scan (skips 2-50, both directions) against a word-list fixed in advance, set against a chance-baseline of 500 random reshufflings of the same letters. No "meaningful" hit survives:
David (skip 7/40) and YHWH (skip 26): exactly at chance level (real ≈ shuffle). Yeshua (−31) and Or (34): even below chance level. In placement, moreover, David and Yeshua land on the desolator-zone (the coming prince / ishchit) — the opposite of their meaning. No matches; deleted.
Yeshu (without the ayin) has been removed as a candidate: three hyper-frequent letters (י/ש/ו), pure frequency noise.
Two light resonances remain worth noting, not load-bearing: Or (skip 34 = ga'al) and Torah (skip 36 = ¼ × 144, the number of the maskilim; also 6² and the "36 righteous ones") both run through the same revelation-cluster — vision · prophet · word (24h, 25c). YHWH (skip 26) spans a fine arc through the judgment-block (26a→27a), but out of purely grammatical particles. All three: speculative, on-theme, no proof.
The alphabet-completeness. Earlier research pointed out that the pericope uses all but one of the 22 basic letters (only samekh is missing) and all five final forms. Both are factually correct — but as a coupon-collector problem they are the norm, not the miracle:
On average some ~241 letters are needed before one may expect all 22 basic letters; the pericope has 359. Full coverage is therefore expected — 21/22 lies even slightly below the average (P(all 22) ≈ 0.86). In the 27-grapheme model (including the final letters) the expected coverage at 359 letters is ~26.4 out of 27; the text has 26 → precisely at the average. "Nearly all letters/final letters present" is therefore statistically unremarkable.
The only thing that is mildly striking: that it is precisely samekh that is the absent one, while the rarer final letters ף and ץ are indeed present (twice each). The chance that precisely samekh is missing and everything else present ≈ 1 in 33. Modest — no proof, but a weighed hook.
That hook is above all thematic: samekh is the pictogram of support/shield; the absence reads as the laid-down protection — the kenosis (Philippians 2; "and He has nothing", 26c) and, in the end-time, the removed Restrainer (2 Thessalonians 2). To be used thus: homiletically/typologically, honestly labelled as interpretation, not as statistics.
Both show the same line as §6: what does not survive the fine-tooth comb is reported just as hard as what remains standing. The ELS is noise; the alphabet-completeness is expected; and only the targeted absence of the samekh bears — and even then chiefly as an image, not as proof.
Text criticism qere/ketiv: Appendix A; the Janus words: Appendix E. Compare God's Algorithm, Appendix C (Lexicon of Numbers) and Appendix E. All counts independently verified from B19a (ketiv); ELS and symmetry tests with chance-baselines via random permutations.