PHASE 3.395 NEXT DESIGN — REPLICATION OF THE CLEAN RULE COMPARISON

PURPOSE

Phase 3.394 found the same direction in PRIMARY and HOLDOUT: Clean Peek produced a higher average final bankroll than the declared Clean No-Peek model.

Phase 3.395 should remain simple.

Repeat the clean full-session comparison on a fresh deterministic shoe-seed range that was not used for the Phase 3.394 PRIMARY/HOLDOUT test.

Keep all rules, GBP wager execution, £100 starting bankroll, 30-hand limit and comparison measures unchanged.

QUESTION

Does the Phase 3.394 Peek-versus-No-Peek difference repeat on fresh card sequences?

CONNECTION TO AMIT'S BACKBONE

No historical result is rewritten.

This is a robustness check on the clean dealer-Blackjack rule comparison only. If the same direction repeats, the programme can move toward synthesis rather than adding new complexity.

DECISION GATE

If the direction and practical scale remain broadly consistent, treat the rule-timing result as replicated and move toward final interpretation/closure.

If it reverses or changes materially, investigate the discrepancy before closure.

The extension remains due for closure no later than Phase 3.400.
