Do the “right” names fit their lives better than shuffled names?
RESULT · NOT SUPPORTED
We froze the existing A–Z designed concept dictionary, converted each name into its concept vocabulary, and compared exact shared concept words with the citizen's existing Purpose + Great Question + Capability labels. Then names were randomly reassigned 10,000 times.
Reality returned something stronger than “no effect”
The true names did not outperform shuffled names under this fixed designed-letter dictionary. Their mean lexical affinity (0.0736) was actually below the average shuffled assignment (0.0881).
This is evidence against treating the current A=Attention / N=Network / T=Test style dictionary as a predictive decoder of these 100 lives.
Why this matters
A flexible interpreter can select whichever concept from each letter feels appropriate after learning the biography.
If random names fit as well or better, the feeling of “uncanny fit” is not sufficient evidence.
The letter dictionary remains useful as a creativity/learning device, but Pilot 2 gives no basis for calling it a life code.
Boundary
Combined evidence so far
Pilot 1: objective Roman-letter structural similarity did not track Life-Space similarity (r ≈ −0.047; permutation p ≈ 0.211).
Pilot 2: fixed designed letter-concept readings did not make true names fit their Life labels better than shuffled names (p ≈ 0.905).
Therefore two different tests currently point in the same direction: the simple “name as compressed life code” formulations tested here are not supported.