# Names table test (#312), 10 Oct 2026
Names from ppdata tables (r6nat 2,501, r6weth 19,009 named). Script it24/t9b.py (t9 with the table names).
Same sample (A1 n=1268 base 1.09, H3000u n=243 base 3.31). No theme or repeat-name bucket differs >=1.0x at n>=30 with both halves agreeing:
- A1 animal n49 1.55 (halves 2.02/1.10), repeat-name True n478 1.26 vs 0.99 (diff 0.27), capitalized n1123 vs 145: noise-level.
- H3000u: only ai n14, repeat-name n9 (0.16): too small.
Verdict: KILLED. No arm to testmgr.