When two different HFGCS EAM messages are slid against each other character-by-character to find the alignment offset that produces the most matching characters, the best-fit offsets cluster at multiples of 22 and 44 far beyond what random chance predicts.
This holds across messages from different prefix groups, different dates (spanning 2022–2026), and different message lengths. The messages being compared are independent intercepts — different content, different days, often different originators — yet they structurally align at 44-character intervals.
The technique is a standard sliding window comparison. For two messages A and B:
Msg A: ...K Z O C C C C X 5 L G V O T O 5 Z Q V H H 4 C K H L L L L G 3 F A E | . . | | | | | | | | | . . . . . | | | . . . . . | | | | | | | | | Msg B: ...K 2 K C C C C X 5 L G V W N 5 T J Q V H 4 H P 4 I L L L L G 3 F A E
This is performed within each prefix group (messages sharing the same 2-character prefix). 243 unique messages from 21 prefix groups yield 1,512 pairwise comparisons.
If message structure were unrelated to a 44-character period, best-fit offsets would scatter roughly uniformly across all possible values. Instead, multiples of 22 dominate:
The three multiples of 22 (excluding zero) — |22|, |44|, |66| — account for 10.5% of all pairs despite representing only 3 out of ~200+ possible offset values. Under uniform distribution, they would represent roughly 1.5%.
Below are actual message pairs whose best sliding alignment falls at multiples of 22. These are different messages, intercepted on different dates. Green characters are positions that match at the given offset.
Beyond the 22/44 peaks, there is a broader structural signal: 67.6% of all best-fit offsets are even, versus 32.4% odd. Under random expectation, even and odd would each be ~50%.
| Category | Count | Percentage | Expected (uniform) |
|---|---|---|---|
| Even offsets | 1,022 | 67.6% | ~50% |
| Odd offsets | 490 | 32.4% | ~50% |
| Divisible by 22 | 360 | 23.8% | ~2–3% |
| Divisible by 44 | 242 | 16.0% | ~1–2% |
This even-offset dominance suggests a 2-character subunit (character pair) as a building block, with the 22-character and 44-character periods built from groups of 11 and 22 such pairs.
The 22/44-character periodicity appears across all 21 Group 4 prefix groups analyzed. These prefixes are believed to represent different message originators, routing paths, or message types. Despite different content and temporal separation, the structural alignment period is the same.
| Prefix | Messages | Pairs at ±44 | Example |
|---|---|---|---|
| PD | 6 | 3 | 31 matches, Nov 2022 |
| 6K | 9 | 4 | 39 matches, Jun–Jul 2023 |
| ET | 9 | 2 | 13 matches, Jan–Feb 2023 |
| BH | 20 | 5 | Visible in slider heatmap |
| BV | 15 | 4 | Visible in slider heatmap |
| JC | 20 | 3 | 12–13 matches, Jul–Aug 2023 |
| JO | 18 | 4 | Visible in slider heatmap |
| + 14 others |
The cross-message alignment at 44-character intervals means that structurally similar positions recur every 44 characters not just within a single message, but across different messages from different dates. This constrains the possible explanations:
The 22-character half-period (offset ±22 is also elevated, as is ±66 = 3×22) suggests a possible 22-character base unit, with 44 being a double period. The even-offset asymmetry further suggests a 2-character pairing as the smallest structural element.
A follow-up analysis asks: is the 44-character periodicity a repeating cipher key (where position i and position i+44 are always governed by the same rule), or a discrete block structure (where messages are composed of independent 44-character blocks)?
The discriminating test: when two messages align at offset ±44, a periodic cipher predicts all overlapping block pairs should match at similar rates. Instead, the match rate is concentrated in one or two specific block pairs, with adjacent blocks falling to near-random levels. This pattern is consistent across all prefix groups examined.