PHASE 3.13 — LAYMAN RESULTS EXPLANATION
HARD 12–13 AGAINST STRONG DEALERS

WHAT QUESTION DID 3.13 ASK?
Phase 3.12 left two different sources of damage:
1. residual low totals that still DOUBLE; and
2. HARD 12–13 against strong dealer cards.

We deliberately compared those groups before changing anything.

WHAT THE 3.12 DATA SAID
HARD 12–13 vs dealer 7–Ace under automatic DOUBLE:
- Primary: 18,075 hands; mean net -0.7296 units/hand; bust rate 33.49%.
- Holdout: 18,245 hands; mean net -0.7604; bust rate 34.39%.

Residual HARD 5–8 vs dealer 2–6 under automatic DOUBLE:
- Primary: 4,887 hands; mean net -0.2202.
- Holdout: 4,737 hands; mean net -0.2681.
- Immediate bust rate: 0%.

This comparison made the next choice unusually clear. HARD 12–13 vs 7–Ace had roughly four times the exposure and substantially greater average damage.

WHAT CHANGED
Phase 3.13 changes ONE family only:
HARD 12 and HARD 13 versus dealer 7–Ace now HIT instead of Maximum Aggression automatically DOUBLING.

All other 3.12 policy layers remain frozen.

WHY HIT?
For these visible states the frozen Textbook control prescribes HIT. We are therefore not inventing a new action or using knowledge of future cards. We are selectively substituting the corresponding Textbook decision only where the finite-shoe Primary/Holdout evidence justified intervention.

RESULT
Primary average ending bankroll rises from 95.203 in 3.12 to 96.617 in 3.13: +1.414 units.
Holdout rises from 95.065 to 96.547: +1.482 units.

Positive sessions rise from 28.26% to 32.98% Primary and from 27.86% to 32.08% Holdout.

This is a substantial and closely replicated improvement.

WHAT HAPPENED TO THE GAP TO TEXTBOOK?
Textbook remains approximately 100 units average ending bankroll:
- Primary Textbook: 99.956.
- Holdout Textbook: 99.972.

Selective Aggression has now recovered to:
- Primary: 96.617.
- Holdout: 96.547.

So the hybrid policy is now much closer to Textbook than the original Maximum Aggression baseline, while still retaining Maximum Aggression behaviour wherever the evidence has not justified an override.

PEAK BANKROLL AND MULTIPLIERS
Selective Aggression maximum observed bankroll:
- Primary: 127.0 units at Hand 25.
- Holdout: 127.5 units at Hand 30.

No session finished at 2x, 3x, 5x or 10x.

The programme is still improving average outcomes and positive-session frequency rather than producing extreme 30-hand multipliers.

WHAT THE REMAINING EVIDENCE NOW LOOKS LIKE
After removing the HARD 12–13 vs 7–Ace family, HARD 8 against strong dealers rises prominently in the remaining ranking:
- HARD 8 vs 10: 1,004 Primary / 1,054 Holdout, mean net -0.813 / -0.708.
- HARD 8 vs Ace: 189 / 209, mean net -0.656 / -0.842.
- HARD 8 vs 9: 255 / 281, mean net -0.596 / -0.747.
These have 0% immediate bust rate and Textbook HIT.

We also see a different unresolved group: HARD 13 vs dealer 2, where Textbook says STAND and the aggressive DOUBLE has substantial bust exposure.

WHY THIS MATTERS
The remaining damage landscape is no longer one simple descending sequence. It contains:
- low-total forced-one-card aggression such as HARD 8 vs strong dealers;
- mid-total states such as HARD 13 vs 2 where the correct Textbook action is STAND;
- soft-hand states;
- and some split-child states.

The next experiment therefore needs another evidence comparison rather than mechanically changing the next total.

NEXT QUESTION — PHASE 3.14
Compare the aggregate leverage of the now-prominent HARD 8 strong-dealer HIT family with the remaining dealer-weak mid-total states such as HARD 13 vs 2. Change only one coherent family after that comparison.

PUBLICATION NOTE
All explanatory interpretation for Phase 3.13 is intentionally contained in this Layman Results report. The remaining TXT/CSV files are the technical evidence and audit trail rather than separate explanatory reports.
