PHASE 3.04 — SELECTIVE-AGGRESSION INSTRUMENTATION SUMMARY

PRIMARY resolved selective-aggression hands: 171721; mean net/hand -0.3993; mean committed stake 1.7962; bust rate 26.65%; third-card bust rate when a third card was taken 33.47%.
Textbook would have chosen on those same visible first-decision states: HIT 70274, STAND 68077, DOUBLE 15807.

HOLDOUT resolved selective-aggression hands: 171786; mean net/hand -0.4068; mean committed stake 1.7946; bust rate 26.75%; third-card bust rate when a third card was taken 33.67%.
Textbook would have chosen on those same visible first-decision states: HIT 70460, STAND 67880, DOUBLE 15749.

INTERPRETATION
Phase 3.04 carries forward the Phase 3.03 split-child HARD 20 STAND rule and adds one new family only: HARD 19 now STANDS instead of automatically DOUBLING, whether ordinary or split-child. All other Maximum Aggression behaviour remains frozen. The purpose is to measure whether removing this single replicated damage mechanism improves the 30-hand bankroll distribution.
