DETERMINISTIC BLACKJACK DECISION ENGINE RESEARCH
PHASE 1.19 — RECENT-CARD MEMORY AND HUMAN DECISION PSYCHOLOGY

LAYMAN PURPOSE
--------------
This phase tests a familiar feeling: after many hands without a 10-value card, a player may believe a 10 is now due. The player may fear taking another card, or expect a dealer showing 2–6 to receive the overdue 10 and bust. After many recent tens, the player may believe another 10 is less likely.

The experiment does not predict hidden cards and does not provide live gambling assistance. It compares fixed psychology profiles with a textbook control inside a deterministic simulator.

CONTROLLED CONDITIONS
---------------------
10,000 sessions; 100 hands per session; £100 nominal starting bankroll; £1 normal wager; six decks; dealer stands on soft 17; visible completed cards only. A drought means 12 consecutive visible cards without a 10-value card. Abundance means at least 6 tens among the latest 12 visible cards.

PRIMARY RESULTS
---------------
TEXTBOOK                average ending £  98.49 | total net £ -15139.00 | staked £1097067.00 | W/L/P 431099/483922/84979 | activations 0 | better/worse/same 0/0/0
PLAYER_BUST_FEAR        average ending £  98.46 | total net £ -15446.50 | staked £1097135.00 | W/L/P 431088/484165/84747 | activations 4034 | better/worse/same 565/940/2529
DEALER_BUST_BELIEF      average ending £  98.47 | total net £ -15331.50 | staked £1097069.00 | W/L/P 431017/484007/84976 | activations 226 | better/worse/same 54/48/124
TEN_ABUNDANCE           average ending £  98.29 | total net £ -17136.50 | staked £1097066.00 | W/L/P 429510/484322/86168 | activations 22316 | better/worse/same 4703/6052/11561
MIXED                   average ending £  98.21 | total net £ -17931.50 | staked £1097062.00 | W/L/P 429281/484785/85934 | activations 26625 | better/worse/same 5346/7020/14259

BELIEF-SPECIFIC COUNTS
----------------------
TEXTBOOK | player-fear 0 | dealer-bust 0 | abundance 0
PLAYER_BUST_FEAR | player-fear 4034 | dealer-bust 0 | abundance 0
DEALER_BUST_BELIEF | player-fear 0 | dealer-bust 226 | abundance 0
TEN_ABUNDANCE | player-fear 0 | dealer-bust 0 | abundance 22316
MIXED | player-fear 4056 | dealer-bust 231 | abundance 22338

LAYMAN INTERPRETATION
--------------------
An individual altered decision can win and feel like proof that the recent pattern mattered. The better/worse/same totals show the complete population, including the less memorable failures. The holdout run uses different fixed seeds to test whether any apparent pattern survives new simulated sessions.

HOLDOUT RESULTS
---------------
TEXTBOOK                average ending £  98.47 | total net £ -15273.00 | staked £1096443.00 | W/L/P 430775/483695/85530 | activations 0 | better/worse/same 0/0/0
PLAYER_BUST_FEAR        average ending £  98.44 | total net £ -15594.00 | staked £1096484.00 | W/L/P 430713/483954/85333 | activations 3912 | better/worse/same 559/878/2475
DEALER_BUST_BELIEF      average ending £  98.47 | total net £ -15288.50 | staked £1096430.00 | W/L/P 430768/483710/85522 | activations 210 | better/worse/same 58/47/105
TEN_ABUNDANCE           average ending £  98.23 | total net £ -17700.00 | staked £1096571.00 | W/L/P 429141/484396/86463 | activations 22103 | better/worse/same 4569/6004/11530
MIXED                   average ending £  98.18 | total net £ -18158.50 | staked £1096575.00 | W/L/P 429037/484777/86186 | activations 26202 | better/worse/same 5169/6934/14099

REPLAY STATUS: PASS

LIMITATION
----------
Because actions can consume different numbers of cards, profiles begin from identically generated session seeds but may diverge later within a shoe. The isolated better/worse comparison uses a deterministic counterfactual stream and is supporting evidence, not a claim about a live casino shoe.
