DETERMINISTIC BLACKJACK — S-STAGE PLATE-READY VERIFIED MANIFEST
CODE VERSION: V9.2.179-UAT
SESSION: 48 | 20261004_024655
STATUS: PRE-RENDER VERIFIED INPUT SET — PLATE REMAINS PENDING UNTIL POST-RENDER AUDIT
PURPOSE: S has already prepared the candidate publication values so plate creation is a controlled layout/audit step, not a second data-discovery exercise.
AUTHORITY: current-session evidence and this S-generated manifest are numeric authority. Previous plates are presentation precedent only. Sessionplate_structure.zip contains the canonical shuffle-robustness table template; that template is STRUCTURAL authority for the robustness table only and is never numeric authority.
S ANALYSIS SOURCE: blackjack_live_session_20261004_024655.txt | corrected analytical evidence supersedes raw capture only within its documented correction scope; raw capture remains preserved.
DEALER-NATURAL TIMING: player-actions-first | dealer hole card/natural revealed and resolved after permitted player actions | S17 | insurance not used.
ZERO IS DATA: never inherit a zero, N/A, score, status or number from a template.

================ OBSERVED / LIVE SESSION =================
SESSION PROVENANCE: REAL-MONEY ONLINE CASINO — OBSERVED | Paddy Power | Blackjack CashOut
SOURCE_CLASS: LIVE_CASINO
PLATFORM: PADDY_POWER
GAME: BLACKJACK_CASHOUT
EVIDENCE_PHASE: FORMAL_SESSION
CAPTURE_ORIGIN: CONTEMPORANEOUS_OBSERVATION_LOG
EXECUTION / RESEARCH SCALE: DECLARED BEFORE HAND 1 | experimental bankroll £10.00 | 1:10.00 | actual casino £1.50/£2.00/£3.00 = Frozen-equivalent £15/£20/£30 | scaled actual wagers are exact Frozen execution equivalents, NOT wager deviations
CAPTURE SCALE EVIDENCE: DECLARED EXECUTION SCALE | Paddy Power | experimental bankroll £10.00 | scale 1:10 | actual platform start £10.00 | protected floor £0.00 | normalized research start £100.00 | platform smallest wager £0.10 | scaling declared before Hand 1
DUAL MONEY FRAME: actual casino balance start £10.00 | protected floor £0.00 | experimental bankroll £10.00 -> £26.50 | actual balance final £26.50 || research-equivalent bankroll £100.00 -> £265.00
SHOE / DECK PROVENANCE: number of decks UNKNOWN | shoe size UNKNOWN | penetration UNKNOWN | cards remaining UNKNOWN | internal shuffle/replenishment mechanics UNKNOWN
SHUFFLE WORDING RULE: report only visibly OBSERVED shuffle events. Zero observed shuffle events must never be rewritten as proof that no internal shuffle/replenishment occurred.
PUBLICATION TOP-BLUE PROVENANCE BANNER: REAL-MONEY ONLINE CASINO • PADDY POWER — BLACKJACK CASHOUT | declared 1:10.00 execution scale • £10.00 experimental bankroll = £100 research frame
PUBLICATION MONEY RULE: established scorecard statistics use research-equivalent values for Sessions 1+ comparability; actual casino values are execution provenance and may be shown secondarily. Never infer non-Frozen play from scaled actual wager size.
MODE: HYBRID
Observed path: start £100.00 | final £265.00 | P/L +£165.00 | hands 10 | W/L/P 7/2/1 | exposure £270.00 | avg wager £27.00
Observed path extrema: peak £265.00 at hand 5 | trough £100.00 at hand 0 | maxDD £30.00 at hand 7
Observed synthetic-reference profile: N/A | requires a complete source-supported 30-hand observed path under the frozen scorecard equations

================ S EXHAUSTIVE PUBLICATION EVIDENCE CATALOGUE =================
CATALOGUE RULE: S attempts every established publication statistic defensible from retained current-session evidence. Supported values are emitted; unsupported values receive an explicit evidence-state reason and must not be inferred by the plate renderer.
Frozen vs Casual decision comparison | common source-supported hands 12 | Frozen decisions recorded 43 | Casual decisions recorded 33 | directly comparable decisions 11 | same 9 | different 2 | decision alignment 81.8% | base-wager comparable hands 13 | Casual-vs-Frozen wager match/above/below 10/0/3
PERCENTILE CONTEXT: NOT AVAILABLE FROM CURRENT ESTABLISHED SCORECARD MODEL | Risk/Journey values below are synthetic-reference profile outputs, not percentile ranks. No percentile may be invented until an explicit reference-corpus percentile mapping is implemented and validated.

================ S-STAGE PUBLICATION TEMPLATE / MODULE SELECTION =================
TOP-SECTION TEMPLATE CLASS: STANDARD ARCHETYPE B — HYBRID / EVIDENCE-STATE / INCOMPLETE JOURNEY
PRIMARY TOP-SECTION PRECEDENT: HYBRID / EVIDENCE-STATE STANDARD (canonical seed reference: Session 10)
APPLICABILITY SCORE: Complete-observed standard = 3 | Hybrid/evidence-state standard = 10
SELECTION REASON: Evidence favours Hybrid/evidence-state treatment. Complete-observed applicability=3, Hybrid/evidence applicability=10, complete observed=false, action differences=8, wager differences=0.
MANDATORY EVIDENCE MODULES: HYBRID / EVIDENCE-STATE STANDARD MODULES: evidence-state/N/A/last-source-supported treatment, decision provenance/divergence where present, five-controller running P/L bar and protected lower interpretation area. Carry complete-observed score modules only when current evidence actually supports them.
MANDATORY GLOBAL MODULE: full-width five-controller running P/L bar and protected lower interpretation area whenever those controllers/evidence states are applicable.
SELECTION RULE: S classifies the current evidence against two standard publication archetypes, independent of the current Session number. Canonical Session 7 and Session 10 plates are seed whole-plate visual references for those archetypes only; Session 9 is a HYBRID-DETAIL MODULE reference when current Hybrid-vs-Frozen decision comparison evidence exists. None are numeric authority. Sessionplate_structure.zip is always included in the ChatGPT publication handoff. Its CANONICAL_SHUFFLE_ROBUSTNESS_TABLE_TEMPLATE.png is structural authority for the shuffle-robustness table only; all table values must come from current S evidence.
PUBLICATION DENSITY CHECK: preserve the selected precedent's information density, canvas utilisation, typography scale, panel depth and graphical richness. Reject a sparse/generic reconstruction even when its numbers are correct. Do not optimise for small PNG file size; render at publication resolution.
COMPOSITION DIRECTIVE: PRIMARY BODY = HYBRID/EVIDENCE-STATE / SESSION-10 ARCHETYPE; ADD = COMPLETE-OBSERVED SCORE MODULES only where current evidence supports them; GLOBAL = FIVE-CONTROLLER RUNNING-P/L BAR; DO NOT CREATE A THIRD VISUAL ARCHETYPE.
PUBLICATION COMPOSITION MAP: COMPLETE OBSERVED + FIXED CASUAL = LEFT observed scorecard | MIDDLE Session Identification + Framework/Recorded Session + Frozen-vs-Casual Comparison | RIGHT Fixed Casual scorecard. These three components are protected. Stat-Watching, Ultimate AI and real-casino provenance are additive modules and must not displace them.
CANONICAL COMPONENT REFERENCES: benchmark_middle_session_comparison.png = middle-column structure authority; benchmark_fixed_casual_scorecard.png = Fixed Casual scorecard structure authority; benchmark_visual.png = final editorial/canvas finish authority. Current S evidence remains numeric authority.
SHUFFLE TABLE COLUMN LOCK: preserve Method / Shuffle Type, Description, Median Bankroll, Mean Bankroll, £200 Reach, Mean Exposure, Mean Max Drawdown, 30-Hand Completion, Highest Final Bankroll, Paired Final-Bankroll Difference (F-C), Mean Action Difference. Do not merge or omit columns.
CANONICAL GEOMETRY GATE: treat the primary archetype as the page skeleton, not loose inspiration. Preserve the accepted family's approximately 40% left / 20% centre / 40% right main-body hierarchy, icon-led metric cells, prominent bankroll arrow/summary, large Risk/Journey score boxes, dense full-width robustness table, and four equal-depth lower analysis panels. Hybrid additions must fit inside that skeleton or increase canvas height; they must not cause a new dashboard layout.
COMPONENT MAP GATE: before rendering, classify every visible module as SESSION-7 WHOLE-PLATE INHERITED, SESSION-10 WHOLE-PLATE INHERITED, SESSION-9 HYBRID-DETAIL MODULE INHERITED, or CURRENT-SESSION EVIDENCE SUBSTITUTION. Session 9 is module-only and must never become a third whole-plate archetype. If a module cannot be mapped to one of those categories, do not invent a new permanent panel unless current evidence makes it unavoidable.
VISUAL SIMILARITY GATE: before showing the image, compare it against both supplied canonical plates. It must be immediately recognisable as the same publication family by geometry, panel proportions, density, iconography, typography hierarchy, metric-cell structure, score prominence, table depth and lower-panel utilisation. Blue/white colour similarity alone is insufficient. A numerically correct but visibly redesigned or sparse plate is FAIL and must be rebuilt before user presentation.
ACCEPTED-PLATE CORRECTION LOCK: if a canonical current-session SessionN.png already exists and the task is an audit/data correction, that PNG is the IMMUTABLE VISUAL BASELINE. Preserve its canvas, geometry, panel positions, typography hierarchy, colour treatment, borders, icons and information hierarchy. Correct only manifest-failing text/value/state cells in place. DO NOT generate a replacement design, new dashboard, alternate theme or fresh plate composition unless the user explicitly requests redesign.
USER VISUAL ACCEPTANCE BOUNDARY: user visual acceptance is final human UAT after autonomous evidence AND canonical-visual gates pass; it is not the first design-review stage.
Decision provenance: total 52 | textbook/frozen-aligned 52 | research overrides 0 | primary H11v10 overrides 0
HYBRID vs FROZEN action comparison: Frozen decisions recorded/comparable 52 | same 44 | different 8 | decision alignment 84.6%
HYBRID vs FROZEN wager comparison: observed wagers 10 | match/above/below 10/0/0 | wager divergence 0/10 (0.0%)
HYBRID DETAIL NUMERIC LOCK — CURRENT SESSION ONLY: Frozen decisions recorded 52 | Hybrid/Frozen comparable decisions 52 | Same decision 44 | Different decision 8 | Decision alignment 84.6% | Wager match/above/below 10/0/0
HYBRID DETAIL ARITHMETIC LOCK: same + different = comparable -> 44 + 8 = 52 PASS | wager match + above + below = observed hands -> 10 + 0 + 0 = 10 PASS
HYBRID DETAIL REFERENCE SAFETY: Session9.png supplies visual/module structure ONLY. If a rendered Hybrid-detail value differs from this CURRENT-SESSION NUMERIC LOCK, NUMERIC AUDIT FAILS and the plate must be corrected before presentation.
SESSION-9 HYBRID DETAIL MODULE: REQUIRED visual module precedent for this Hybrid session. Use Session 9 only to enrich the left Hybrid-vs-Frozen analytical block; do not use it as a whole-plate template and never inherit Session 9 numbers.
Wager match/above/below 10/0/0
Action match/different 44/8
SOURCE: Paddy Power
GAME: Blackjack CashOut
SOURCE TYPE: Observed real-money online casino blackjack
SHOE / DECK CONFIGURATION: UNKNOWN — not observable from the captured casino interface
NUMBER OF DECKS: UNKNOWN
PENETRATION / CARDS REMAINING: UNKNOWN — do not infer from captured-card count
EXECUTION SCALE: DECLARED BEFORE HAND 1 | experimental bankroll £10.00 | 1:10 | actual £1.50 / £2.00 / £3.00 = Frozen-equivalent £15/£20/£30 | platform smallest wager £0.10
RESEARCH-EQUIVALENT MONEY: Start £100.00 | Final £265.00 | Exposure £270.00 | all Frozen comparison/statistical values remain on the established £100 frame
Shuffle events explicitly observed: 0
Total cards recorded: 105
Cards since most recent observed shuffle/start: 105
Validation warnings/overrides encountered: 6
Hands re-entered through amendment review: 6
Observed shuffle boundaries: none
SOURCE: Paddy Power | REAL-MONEY ONLINE CASINO | Blackjack CashOut | deck/shoe configuration UNKNOWN

================ SESSION SUMMARY — RUNNING P/L BAR CANDIDATES =================
Observed HYBRID: +£165.00
Frozen Reference: N/A | SOURCE_EXHAUSTED | NOT SOURCE-COMPLETE | last source-supported £150.00
Frozen Reference source boundary: hands 12 | W/L/P 5/6/1 | exposure £305.00 | cards consumed 105 | last source-supported bankroll £150.00 | termination SOURCE_EXHAUSTED
Frozen synthetic-reference profile | RISK 66.991 | JOURNEY 38.482 | Survival 50.906 | Path Exposure 40.962 | Recovery Activity 23.579 | inputs: maxDD £120.00 | giveback £100.00 | cumulative wager £305.00 | absolute movement £290.00 | recovery episodes 0 | successful 0 | recovery ratio 0.000000 | table_exit 0
FIXED-POLICY CASUAL — CURRENT CANONICAL PUBLICATION STATE: INCOMPLETE | NOT SOURCE-COMPLETE | hands 13 | final N/A | termination SOURCE_EXHAUSTED
FIXED-POLICY CASUAL STATE AUTHORITY: THIS LINE IS CURRENT MANIFEST AUTHORITY. Historical plates, benchmark images, prior handoffs and component references are presentation/provenance references only and MUST NOT override this completion/source state.
Fixed-Policy Casual: N/A | SOURCE_EXHAUSTED | NOT SOURCE-COMPLETE | last source-supported £85.00
Fixed-Policy Casual source boundary: hands 13 | W/L/P 4/8/1 | exposure £240.00 | cards consumed 105 | last source-supported bankroll £85.00 | termination SOURCE_EXHAUSTED
Casual synthetic-reference profile | RISK 68.019 | JOURNEY 37.364 | Survival 52.311 | Path Exposure 32.071 | Recovery Activity 27.711 | inputs: maxDD £120.00 | giveback £105.00 | cumulative wager £240.00 | absolute movement £225.00 | recovery episodes 1 | successful 0 | recovery ratio 0.000000 | table_exit 0
Stat-Watching Casual: Chosen behavioural path: hands 11 | bankroll £85.00 | W/L/P 3/7/1 | exposure £210.00 | termination BEHAVIOURAL_EXIT | chosen exit Hand 11 at £85.00
Ultimate AI: Ultimate AI: hands 12 | bankroll N/A | last source-supported £150.00 | P/L N/A | W/L/P 5/6/1 | exposure £305.00 | termination SOURCE_EXHAUSTED | NOT SOURCE-COMPLETE

================ FORMAL-START / CONTEXT EVIDENCE =================

================ HYBRID FORMAL-START COUNTERFACTUAL | SESSION 20261004_024655 ================
PURPOSE: replay the unchanged Frozen and fixed Casual policies from the same formal-session starting captured cardstream. The observed Hybrid journey remains separate and authoritative for what was actually played.
Captured formal chronology available: 105 blackjack-replayable cards (rank known; any U suit remains unresolved). Each policy consumes independently; no card rank or settlement is invented.
Frozen from formal start: outcome N/A | hands 12 | W/L/P 5/6/1 | exposure £305.00 | cards consumed 105 | termination SOURCE_EXHAUSTED | NOT SOURCE-COMPLETE | last source-supported bankroll £150.00
Frozen synthetic-reference profile | RISK 66.991 | JOURNEY 38.482 | Survival 50.906 | Path Exposure 40.962 | Recovery Activity 23.579 | inputs: maxDD £120.00 | giveback £100.00 | cumulative wager £305.00 | absolute movement £290.00 | recovery episodes 0 | successful 0 | recovery ratio 0.000000 | table_exit 0
Casual from formal start: outcome N/A | hands 13 | W/L/P 4/8/1 | exposure £240.00 | cards consumed 105 | termination SOURCE_EXHAUSTED | NOT SOURCE-COMPLETE | last source-supported bankroll £85.00
Casual synthetic-reference profile | RISK 68.019 | JOURNEY 37.364 | Survival 52.311 | Path Exposure 32.071 | Recovery Activity 27.711 | inputs: maxDD £120.00 | giveback £105.00 | cumulative wager £240.00 | absolute movement £225.00 | recovery episodes 1 | successful 0 | recovery ratio 0.000000 | table_exit 0
Publication rule: if a replay is NOT SOURCE-COMPLETE, its outcome is N/A; the displayed bankroll is only the last source-supported bankroll. Risk/Journey remain synthetic-reference profile outputs for the source-supported path, not completed-outcome claims.
================ END HYBRID FORMAL-START COUNTERFACTUAL | SESSION 20261004_024655 ================

================ STAT-WATCHING / ULTIMATE-AI EVIDENCE =================

================ STAT-WATCHING CASUAL | SESSION 20261004_024655 ================
EXPERIMENT STATUS: EXPLORATORY COUNTERFACTUAL / DETERMINISTIC / QUARANTINED FROM FROZEN-vs-CASUAL SCORECARD
REPLAY BASIS: FORMAL_START
VISIBLE-EVIDENCE RULE: wager and exit decisions use only completed prior-hand evidence visible to the player; no future cards, shoe composition, or unrevealed dealer hole card is consulted. A hole-card hunch may affect a current play decision only after the dealer up-card is visible and only from previously revealed hole-card history.
POLICY: ordinary wager £15; rare evidence-triggered 2x/3x/4x presses; rolling last-five-hand table impression; deterministic walk-away rule from Hand 8 onward. Existing Casual and Frozen policies are unchanged.
STAY POLICY: the stay-at-table counterfactual keeps the SAME Stat-Watching Casual personality, wager logic, hunch logic and play policy; only the behavioural walk-away trigger is disabled. It never switches to Frozen style.
Chosen behavioural path: hands 11 | bankroll £85.00 | W/L/P 3/7/1 | exposure £210.00 | termination BEHAVIOURAL_EXIT | chosen exit Hand 11 at £85.00
Stay-at-table counterfactual: hands 13 | bankroll £85.00 | W/L/P 4/8/1 | exposure £240.00 | termination SOURCE_EXHAUSTED | NOT SOURCE-COMPLETE
Exit value versus staying: N/A unless a behavioural exit occurs and the stay-at-table replay is source-complete.
PATH_CHOSEN: H0=100.00;H1=130.00;H2=175.00;H3=190.00;H4=175.00;H5=175.00;H6=160.00;H7=145.00;H8=130.00;H9=115.00;H10=100.00;H11=85.00
PATH_STAY: H0=100.00;H1=130.00;H2=175.00;H3=190.00;H4=175.00;H5=175.00;H6=160.00;H7=145.00;H8=130.00;H9=115.00;H10=100.00;H11=85.00;H12=70.00;H13=85.00
SW_TRACE | H1 | before 100.00 | after 130.00 | wager 15.00 | committed 30.00 | result W | dealerUp 5♠ | player 8♥,10♥ | dealer 5♠,10♣,7♥ | actions SPLIT,STAND,STAND | branches 8♥,10♥;8♣,10♣
SW_TRACE | H2 | before 130.00 | after 175.00 | wager 15.00 | committed 45.00 | result W | dealerUp 5♠ | player 8♥,3♥,10♦ | dealer 5♠,10♠,7♣ | actions SPLIT,DOUBLE,STAND | branches 8♥,3♥,10♦;8♣,4♠
SW_TRACE | H3 | before 175.00 | after 190.00 | wager 15.00 | committed 15.00 | result W | dealerUp 6♥ | player 4♣,5♦,10♠ | dealer 6♥,9♣,7♦ | actions HIT,STAND | branches 4♣,5♦,10♠
SW_TRACE | H4 | before 190.00 | after 175.00 | wager 15.00 | committed 15.00 | result L | dealerUp 6♦ | player 9♥,9♣ | dealer 6♦,5♣,2♠,8♥ | actions STAND | branches 9♥,9♣
SW_TRACE | H5 | before 175.00 | after 175.00 | wager 15.00 | committed 15.00 | result P | dealerUp 4♦ | player 10♠,10♥ | dealer 4♦,10♦,6♣ | actions STAND | branches 10♠,10♥
SW_TRACE | H6 | before 175.00 | after 160.00 | wager 15.00 | committed 15.00 | result L | dealerUp 5♦ | player 6♠,2♥,4♥ | dealer 5♦,3♣,2♣,9♠ | actions HIT,STAND | branches 6♠,2♥,4♥
SW_TRACE | H7 | before 160.00 | after 145.00 | wager 15.00 | committed 15.00 | result L | dealerUp 10♣ | player 2♥,2♦,5♥,4♣,2♠,3♦ | dealer 10♣,4♠,2♣,3♣ | actions HIT,HIT,HIT,HIT,STAND | branches 2♥,2♦,5♥,4♣,2♠,3♦
SW_TRACE | H8 | before 145.00 | after 130.00 | wager 15.00 | committed 15.00 | result L | dealerUp 4♦ | player 3♥,2♦,2♣,3♠,4♥ | dealer 4♦,2♥,5♠,2♠,3♦,4♣ | actions HIT,HIT,HIT,STAND | branches 3♥,2♦,2♣,3♠,4♥
SW_TRACE | H9 | before 130.00 | after 115.00 | wager 15.00 | committed 15.00 | result L | dealerUp 2♥ | player 5♥,4♦,4♣ | dealer 2♥,3♣,2♠,3♥,2♦,3♠,2♣ | actions HIT,STAND | branches 5♥,4♦,4♣
SW_TRACE | H10 | before 115.00 | after 100.00 | wager 15.00 | committed 15.00 | result L | dealerUp 4♠ | player 4♥,A♦,10♦ | dealer 4♠,2♥,2♦,4♦,A♣,3♣,A♦ | actions HIT,STAND | branches 4♥,A♦,10♦
SW_TRACE | H11 | before 100.00 | after 85.00 | wager 15.00 | committed 15.00 | result L | dealerUp A♠ | player A♥,2♣,3♥,3♠ | dealer A♠,3♦,2♠,4♣ | actions HIT,HIT,STAND | branches A♥,2♣,3♥,3♠
INFLUENCE | H11 | EXIT | walked away at £85.00 • last 5 hands: low cards 86%, 10-value cards 4%, W/L 0/5 • drawdown from peak £105.00
Publication boundary: exploratory side experiment only. Do not award established Frozen/Casual scorecard percentiles or alter the main comparison from this block without a later declared research phase.
================ END STAT-WATCHING CASUAL | SESSION 20261004_024655 ================

================ ULTIMATE AI | SESSION 20261004_024655 ================
AI POLICY: ULTIMATE-AI-V1-FROZEN-WEIGHTED-GATE | deterministic | Frozen baseline | repository through Session 47
REPLAY BASIS: FORMAL_START
TEMPORAL BOUNDARY: only repository evidence through Session 47 plus evidence visible before each current decision was available. Session 48 completed outcomes were not in the repository during this replay.
V1 GATE: Frozen wins weak evidence and ties. Action/wager overrides require convergent predeclared support; scorecard/percentile similarity has zero live-decision weight until a predecision-safe mapping exists.
Ultimate AI: hands 12 | bankroll N/A | last source-supported £150.00 | P/L N/A | W/L/P 5/6/1 | exposure £305.00 | termination SOURCE_EXHAUSTED | NOT SOURCE-COMPLETE
Frozen same-basis comparator: hands 12 | bankroll N/A | P/L N/A | exposure £305.00 | termination SOURCE_EXHAUSTED | NOT SOURCE-COMPLETE
OVERRIDES: action 0 | wager 0 | exit 0
PATH_AI: H0=100.00;H1=130.00;H2=175.00;H3=205.00;H4=250.00;H5=235.00;H6=235.00;H7=205.00;H8=190.00;H9=175.00;H10=160.00;H11=130.00;H12=150.00
AI_TRACE | H1 | before 100.00 | after 130.00 | wager 15.00 | committed 30.00 | result W | dealerUp 5♠ | player 8♥,10♥ | dealer 5♠,10♣,7♥ | actions SPLIT,STAND,STAND | branches 8♥,10♥;8♣,10♣
AI_TRACE | H2 | before 130.00 | after 175.00 | wager 15.00 | committed 45.00 | result W | dealerUp 5♠ | player 8♥,3♥,10♦ | dealer 5♠,10♠,7♣ | actions SPLIT,DOUBLE,STAND | branches 8♥,3♥,10♦;8♣,4♠
AI_TRACE | H3 | before 175.00 | after 205.00 | wager 15.00 | committed 30.00 | result W | dealerUp 6♥ | player 4♣,5♦,10♠ | dealer 6♥,9♣,7♦ | actions DOUBLE | branches 4♣,5♦,10♠
AI_TRACE | H4 | before 205.00 | after 250.00 | wager 15.00 | committed 45.00 | result W | dealerUp 6♦ | player 9♥,2♠,10♠ | dealer 6♦,5♣,4♦,10♥ | actions SPLIT,DOUBLE,STAND | branches 9♥,2♠,10♠;9♣,8♥
AI_TRACE | H5 | before 250.00 | after 235.00 | wager 15.00 | committed 15.00 | result L | dealerUp 6♣ | player 10♦,6♠ | dealer 6♣,5♦,2♥,3♣,4♥ | actions STAND | branches 10♦,6♠
AI_TRACE | H6 | before 235.00 | after 235.00 | wager 15.00 | committed 15.00 | result P | dealerUp 9♠ | player 2♣,2♥,2♦,4♠,5♥,4♣ | dealer 9♠,10♣ | actions HIT,HIT,HIT,HIT,STAND | branches 2♣,2♥,2♦,4♠,5♥,4♣
AI_TRACE | H7 | before 235.00 | after 205.00 | wager 15.00 | committed 30.00 | result L | dealerUp 3♦ | player 2♠,3♥,2♦,2♥,2♣,3♠ | dealer 3♦,3♣,2♠,3♦,4♣,5♥ | actions SPLIT,HIT,HIT,HIT,HIT,STAND,HIT,HIT,STAND | branches 2♠,3♥,2♦,2♥,2♣,3♠;2♣,4♦,4♥,5♠
AI_TRACE | H8 | before 205.00 | after 190.00 | wager 15.00 | committed 15.00 | result L | dealerUp 4♦ | player 2♥,3♣,2♠,3♥,2♦ | dealer 4♦,4♣,3♠,2♣,4♥ | actions HIT,HIT,HIT,STAND | branches 2♥,3♣,2♠,3♥,2♦
AI_TRACE | H9 | before 190.00 | after 175.00 | wager 15.00 | committed 15.00 | result L | dealerUp A♦ | player 4♠,2♥,2♦,4♦,A♣,3♣,A♦ | dealer A♦,10♦ | actions HIT,HIT,HIT,HIT,HIT,STAND | branches 4♠,2♥,2♦,4♦,A♣,3♣,A♦
AI_TRACE | H10 | before 175.00 | after 160.00 | wager 15.00 | committed 15.00 | result L | dealerUp A♠ | player A♥,2♣,3♥,3♠ | dealer A♠,3♦,2♠,4♣ | actions HIT,HIT,STAND | branches A♥,2♣,3♥,3♠
AI_TRACE | H11 | before 160.00 | after 130.00 | wager 30.00 | committed 30.00 | result L | dealerUp 2♥ | player 4♥,10♦ | dealer 2♥,2♦,4♥,A♠ | actions STAND | branches 4♥,10♦
AI_TRACE | H12 | before 130.00 | after 150.00 | wager 20.00 | committed 20.00 | result W | dealerUp A♥ | player 3♣,A♦,2♣,2♠,3♦ | dealer A♥,A♣,3♥,3♠ | actions HIT,HIT,HIT,STAND | branches 3♣,A♦,2♣,2♠,3♦
AI_OVERRIDE | NONE | Frozen baseline superseded all weaker/contradictory signals in this source-supported path.
Publication boundary: experimental synthesis controller. Descriptive deterministic comparison only; no claim of casino advantage, card prediction or guaranteed profitability.
================ END ULTIMATE AI | SESSION 20261004_024655 ================

================ SHUFFLE ROBUSTNESS A–H — 5,000 PAIRED REPLAYS EACH =================
A | Observed-card permutation: shuffle only the captured session-card reservoir; no cards added.
Median bankroll £ Frozen 40.00 | Casual 55.00 | Mean bankroll £ Frozen 53.93 | Casual 60.25 | £200 reach Frozen 3.62% | Casual 1.62% | Mean exposure £ Frozen 214.60 | Casual 203.24 | Mean maxDD £ Frozen 80.72 | Casual 73.56 | 30-hand completion Frozen 0 | Casual 0 | Highest final bankroll £ Frozen 302.50 | Casual 265.00 | Paired final-bankroll difference £ N/A | Mean action difference 0.9812
B | Full six-deck RNG: generate and uniformly shuffle a fresh 312-card six-deck shoe.
Median bankroll £ Frozen 77.50 | Casual 85.00 | Mean bankroll £ Frozen 96.78 | Casual 95.71 | £200 reach Frozen 28.42% | Casual 19.56% | Mean exposure £ Frozen 446.43 | Casual 412.90 | Mean maxDD £ Frozen 93.04 | Casual 84.23 | 30-hand completion Frozen 3252 | Casual 3484 | Highest final bankroll £ Frozen 542.50 | Casual 527.50 | Paired final-bankroll difference £ 7.23 | Mean action difference 1.1582
C1 | Penetration 50%: captured reservoir; reshuffle only remaining cards between hands at 50%.
Median bankroll £ Frozen 40.00 | Casual 55.00 | Mean bankroll £ Frozen 54.31 | Casual 60.29 | £200 reach Frozen 3.82% | Casual 1.44% | Mean exposure £ Frozen 215.09 | Casual 203.58 | Mean maxDD £ Frozen 80.71 | Casual 73.17 | 30-hand completion Frozen 0 | Casual 0 | Highest final bankroll £ Frozen 337.50 | Casual 272.50 | Paired final-bankroll difference £ N/A | Mean action difference 0.9866
C2 | Penetration 65%: captured reservoir; reshuffle only remaining cards between hands at 65%.
Median bankroll £ Frozen 40.00 | Casual 55.00 | Mean bankroll £ Frozen 54.03 | Casual 60.43 | £200 reach Frozen 3.78% | Casual 1.74% | Mean exposure £ Frozen 214.12 | Casual 203.29 | Mean maxDD £ Frozen 80.56 | Casual 73.23 | 30-hand completion Frozen 0 | Casual 0 | Highest final bankroll £ Frozen 347.50 | Casual 257.50 | Paired final-bankroll difference £ N/A | Mean action difference 0.9982
C3 | Penetration 75%: captured reservoir; reshuffle only remaining cards between hands at 75%.
Median bankroll £ Frozen 40.00 | Casual 55.00 | Mean bankroll £ Frozen 54.80 | Casual 60.35 | £200 reach Frozen 3.54% | Casual 1.56% | Mean exposure £ Frozen 214.98 | Casual 202.90 | Mean maxDD £ Frozen 80.60 | Casual 73.46 | 30-hand completion Frozen 0 | Casual 0 | Highest final bankroll £ Frozen 302.50 | Casual 295.00 | Paired final-bankroll difference £ N/A | Mean action difference 0.9892
C4 | Penetration 80%: captured reservoir; reshuffle only remaining cards between hands at 80%.
Median bankroll £ Frozen 40.00 | Casual 55.00 | Mean bankroll £ Frozen 54.03 | Casual 59.69 | £200 reach Frozen 3.74% | Casual 1.66% | Mean exposure £ Frozen 213.55 | Casual 201.73 | Mean maxDD £ Frozen 80.67 | Casual 73.34 | 30-hand completion Frozen 0 | Casual 0 | Highest final bankroll £ Frozen 277.50 | Casual 265.00 | Paired final-bankroll difference £ N/A | Mean action difference 0.9884
D | Physical-style proxy: captured reservoir with deterministic riffle/strip/cut-style mixing.
Median bankroll £ Frozen 40.00 | Casual 55.00 | Mean bankroll £ Frozen 53.73 | Casual 59.17 | £200 reach Frozen 3.62% | Casual 1.48% | Mean exposure £ Frozen 214.08 | Casual 202.31 | Mean maxDD £ Frozen 80.89 | Casual 73.91 | 30-hand completion Frozen 0 | Casual 0 | Highest final bankroll £ Frozen 280.00 | Casual 272.50 | Paired final-bankroll difference £ N/A | Mean action difference 0.9850
E | Automatic/continuous proxy: captured reservoir; after each complete hand, used cards return and the full reservoir is remixed.
Median bankroll £ Frozen 10.00 | Casual 10.00 | Mean bankroll £ Frozen 29.92 | Casual 36.04 | £200 reach Frozen 7.76% | Casual 4.56% | Mean exposure £ Frozen 332.42 | Casual 324.10 | Mean maxDD £ Frozen 106.89 | Casual 99.30 | 30-hand completion Frozen 1205 | Casual 1619 | Highest final bankroll £ Frozen 437.50 | Casual 332.50 | Paired final-bankroll difference £ -3.40 | Mean action difference 1.9240

================ SESSION 10+ STANDARD PLATE ARCHITECTURE =================
TOP BAR: mandatory full-width SESSION SUMMARY — RUNNING PROFIT / LOSS (ALL PLAYERS). Use only current-session verified P/L or evidence-state wording. Controller completion/exit status belongs in this bar; do not duplicate Ultimate-AI final/status in the main header.
FROZEN REFERENCE HEADLINE RULE: when a complete observed FROZEN session has exact observed wager and action alignment, the Frozen Reference box must show the SAME P/L AS OBSERVED SESSION and say SAME AS OBSERVED SESSION. A separately reconstructed formal-start Frozen replay that is SOURCE_EXHAUSTED is a secondary qualifier only and must not make that headline box N/A.
HEADER: retain shuffle/entry context at left, main title, and Journey ID at far right. Header is session-level identification, not a second controller-result area.
OBSERVED SCORECARD PRECEDENT: if the current Observed/Hybrid/Personal journey is a complete source-supported 30-hand path, show its OWN independently computed Risk and Journey in the main observed-player panel. Use the canonical complete-observed archetype treatment where applicable, while retaining the global five-controller running-P/L bar when supported.
FORMAL-START MIDDLE: when supported, show BOTH Risk and Journey beside BOTH Frozen and Casual rows. Large score row is exactly RISK SCORE | JOURNEY INTENSITY SCORE and must align cleanly; do not add a third duplicate Journey box.
COMPLETE-SCORED CONTINUITY: retain BOTH full-width continuity strips: SESSION SUMMARY — RUNNING PROFIT / LOSS (ALL PLAYERS) near the top AND SESSION N — OBSERVED FROZEN FORMAL SESSION SUMMARY with Final Bankroll | Hands Completed | Results | Exposure | Max Drawdown | £200 Reached? | Highest Bankroll. New modules do not replace either strip.
DECISION-POINT PRIORITY: keep the Frozen-vs-Casual comparable-decision / same-different / wager-alignment / action-alignment evidence prominent in the main body. Shuffle-return results are robustness evidence below, not the primary comparison story.
SCORE DISPLAY RULE: any valid current-session Risk/Journey score supplied above must be displayed. Do not turn a valid complete-observed score into N/A. Source exhaustion after a supported partial counterfactual does not erase its explicitly supplied synthetic-reference Risk/Journey values; label source state separately. Never invent absent percentiles/scores, and do not award Stat-Watching/Ultimate AI established Frozen/Casual percentiles unless explicitly authorised.
BOTTOM FOUR PANELS: KEY FINDINGS | MOST SIGNIFICANT OBSERVATIONS | ROBUSTNESS INTERPRETATION | OVERALL CONCLUSION. These panels primarily interpret the robustness table; routine audit housekeeping stays beside the evidence it qualifies.
PROTECTED ANALYSIS SPACE: do not obtain room for new modules by shrinking the bottom analytical area or reducing its established readable text. Increase canvas height if necessary.
WRITTEN CHATGPT SUMMARY: explicitly report running P/L/evidence status for every controller shown in the top bar.

POST-RENDER GATE: audit every number and substantive label against this manifest plus authoritative evidence. Plate remains PENDING until that audit passes.