{ "schema_version": "1.3", "rebuild": "v2", "provenance_class": "dealt_hand", "table_size": 5, "id": "trips-block-bet-small-1088", "format": "multi_option", "free": true, "updated_at": "2026-08-08T00:00:00Z", "source": { "session_id": "NutsandBetsizing", "verdict_audio": "quiz/audio/trips-block-bet-small-1088.m4a", "verdict_beats": "RE-CUT 2026-08-08 from the CANONICAL session (the old clip came off the twin 2025-11-24 s78-83, 51.0s). New beats, all NutsandBetsizing word ranges, registered in tools/cut_quiz_audio.py: w2374-2391 (s254, t=935.76-939.78, head 'So so I think I think' trimmed) + w2414-2426 (s257, t=947.94-951.36) + w2427-2430 (s258, t=953.22-954.60, a separate beat so the 1.9s hole at 951.4-953.2 never airs) + w2442-2481 (s260-261, t=963.58-972.40). Result 19.89s. At the first aired word the screen is a Snipping Tool screenshot of THIS hand frozen at THIS turn node (own frame t=936: board 8-T-8-2, 'Pot: 6.76 BB', 'Hero - 95% / 96.37 BB', 'P2-3055527b6... - 5% / 92.49 BB', red hand-annotations); across beats 2-4 a HoldEq range-entry dialog floats over that same screenshot (t=950, t=966) and is closed again by t=971. Frames READ-ONLY, never aired.", "lesson_slug": null, "lesson_t": null, "master_slug": "NutsandBetsizing_master", "master_t": 723.1, "src_t": 935.8, "lesson_link_note": "registry/quiz_lessons.json carries src_t 240.0 / master_t 86.4 for this quiz — BOTH STALE AND WRONG. tools/map_quiz_lessons.py resolves src_t by reading QUIZ_AUDIO's beats against ser/.json, and the old registry entry indexed sentence 78 of the TWIN (2025-11-24) against the CANONICAL SER, landing on NutsandBetsizing s78 (t=240.0) — eleven minutes before this hand exists. Values above are re-derived here from the re-cut beats: first aired word t=935.76; NutsandBetsizing_master.cutlist.json segment {src_t0 932.5, src_t1 1001.68, out_t0 719.88} -> master_t 723.1. No NutsandBetsizing lesson cutlist covers t=935.76 (the one lesson cut from this session spans 2604-3842), so lesson_slug stays null and map = MASTER. Re-run tools/map_quiz_lessons.py --write to refresh the registry.", "ser_sentences": [254, 257, 258, 260, 261], "hand_record": { "lane": "REPLAYER", "site": "Ignition hand, reviewed in the PokerTracker 4 hand replayer (frames READ-ONLY, never aired)", "hand_id": "#4652230422", "dealt_at": "2024/06/17 2:39 pm (log line 1, own frame t=676)", "stake": "$5 NL (6 max)", "stake_source": "replayer title bar verbatim 'Playback for Ignition Hand #4652230422 (2) - $5 NL (6 max) - Holdem' — own frame t=684, title-bar zoom title684.png (3x)", "units_note": "The replayer is running in BB display mode: every amount it prints is already a big-blind figure ('posts the small blind 0.50 BB', 'Hero raises to 2 BB', 'Hero bets 1.13 BB', 'Pot: 6.76 BB', seat stacks '96.37 BB'). This fixture copies those BB figures verbatim — NO $-to-bb conversion is performed or claimed anywhere. The window's other dollar readouts are left uninterpreted: the hand-selector box '$39.22' and the left-margin HUD 'Avg Stack: $830' have no on-screen mapping to this hand's chips, and the title bar's '$5 NL' is quoted as the table's name, not used as a scale.", "video_t": [676, 1010], "frames": [ "t=676 (scratchpad dh3/trips-block-bet-small-1088/frames/f_676.png + log zoom f_676_log.png) — top of the replayer log, verbatim: 'Ignition Hand #4652230422 begins at 2024/06/17 2:39 pm / Seat 5 (Hero) is the dealer / P6-9d69ed76f4a6e8ceefc7b6e418483d1a posts the small blind 0.50 BB / P2-3055527b64b4c28be7f316bf243f6f1a posts the big blind 1 BB / P3-f45672775726ad15d4cd691f24391353 folds / P4-c1ef03b732dbe2cf8360fb99d543f9af folds / Hero raises to 2 BB / P6-... folds'", "t=684 (f_684.png + log684_a.png + hero684.png 3x) — FLOP node: title bar; five occupied seats at a six-seat ring with the top-right seat drawn empty (emptyseat702.png); hero's cards K on a WHITE card with a black SPADE pip and 8 on a BLUE card with a DIAMOND pip = K♠8♦; board 8 (GREEN club) T (GREEN club) 8 (WHITE spade); 'Pot: 5.63 BB' with 4.50 BB of chips in the middle and 1.13 BB in front of hero; dealer disc drawn at hero's seat (btn702.png); seat stacks P6 395.83, P2 93.62, hero 96.37, P4 100, P3 71.93; log continues 'P2-... calls 1 BB / Flop: [8c Tc 8s] / P2-... checks / Hero bets 1.13 BB'", "t=702 (f_702.png + log702.png) — THE DECISION NODE: board 8♣ T♣ 8♠ + 2 (RED heart), 'Pot: 6.76 BB' with 6.76 BB in the middle, villain tagged CHECK, hero 96.37 BB behind, 'Hero - 95%'; log 'P2-... calls 1.13 BB / Turn: [2h] / P2-... checks'", "t=921, t=934, t=936, t=971 (f_921.png, f_934.png, f_936.png, f_971.png) — the verdict window: a Snipping Tool screenshot of THIS SAME node (board 8-T-8-2, 'Pot: 6.76 BB', 'Hero - 95% / 96.37 BB', 'P2-3055527b6... - 5% / 92.49 BB'), red hand-annotations over it: 'JT ==> 9bb' at the right (ann_right.png, 6x), an illegible ratio scribble top-centre (ann_top.png), and a two-glyph note near hero's seat that reads as either '6%' or '60' (ann_left.png, 8x — genuinely ambiguous, not used for anything)", "t=916 (f_916.png + holdeq_board.png 8x) — the ~60s interlude BEFORE the verdict (SER s229-250): HoldEq v1.1.8, Range 1 'K8o' 42.69% vs Range 2 (a typed list '22,A6s,K8s,Q6s,J8s,T8s,98s,87s,68s,A8o,...') 57.30%, Board field typed by hand 'JsTc8c8?2s'. Earlier in the SAME interlude (own frames t=899, t=902) that dialog carried a DIFFERENT typed board, 'Tc8c8d4h2s' — this hand's flop+turn with re-typed suits plus a blank 4h river — where 'K8o' showed 51.027%; that is the board behind Tyler's spoken '51 percent' (s243), see unverifiable_claims. That board is INVENTED — it carries a J and a spade deuce this hand never dealt (s232 'I don't know what we want the river to be', s240 'I need to get the suits right'). A hypothetical, not the source of the felt; excluded from the fixture and from the aired clip.", "t=1006 (f_1006.png + log1006.png 4x) — the real continuation, log verbatim: 'Turn: [2h] / P2-... checks / Hero bets 4.82 BB / P2-... calls 4.82 BB / River: [Kh]'; table 'Pot: 16.39 BB', hero 91.56 BB, P2 87.67 BB", "t=1160 (f_1160.png) — AFTER the graded moment: the screenshot underneath has been replaced by a RIVER screenshot (board 8-T-8-2-K, 'Pot: 16.39 BB') annotated '8x ==> 2x pot / Tx ==> 3/4 pot / Kx ==> 1.25x pot', with the Windows calculator showing 0.02 x 38 = 0.76. Everything Tyler says from s277 on (the 2x-pot / three-quarters / 1.25x-pot region work) is RIVER sizing on that screenshot — a different node, excluded from this fixture." ], "log_verbatim": [ "Ignition Hand #4652230422 begins at 2024/06/17 2:39 pm", "Seat 5 (Hero) is the dealer", "P6-9d69ed76f4a6e8ceefc7b6e418483d1a posts the small blind 0.50 BB", "P2-3055527b64b4c28be7f316bf243f6f1a posts the big blind 1 BB", "P3-f45672775726ad15d4cd691f24391353 folds", "P4-c1ef03b732dbe2cf8360fb99d543f9af folds", "Hero raises to 2 BB", "P6-9d69ed76f4a6e8ceefc7b6e418483d1a folds", "P2-3055527b64b4c28be7f316bf243f6f1a calls 1 BB", "Flop: [8c Tc 8s]", "P2-3055527b64b4c28be7f316bf243f6f1a checks", "Hero bets 1.13 BB", "P2-3055527b64b4c28be7f316bf243f6f1a calls 1.13 BB", "Turn: [2h]", "P2-3055527b64b4c28be7f316bf243f6f1a checks", "Hero bets 4.82 BB", "P2-3055527b64b4c28be7f316bf243f6f1a calls 4.82 BB", "River: [Kh]" ], "table_structure": "A six-seat ring running FIVE-HANDED — the top-right seat is drawn empty on every frame of the hand (t=684, t=702, zoom emptyseat702.png). No straddle, no ante, no live blind posts: the only forced money is 0.50 BB and 1 BB. 'Seat 5 (Hero) is the dealer' is log-verbatim, so hero is the BUTTON and acts last on every postflop street; P6 is the SB, P2 the BB, and the two folders P3 and P4 are the UTG and CO seats of a five-handed ring (P3 acts first preflop, P4 second, hero third). Clockwise from hero the ring is Hero(BTN) -> P6(SB) -> P2(BB) -> P3(UTG) -> P4(CO), which is the order the seats are drawn in.", "stacks_bb": { "hero_at_decision": 96.37, "villain_at_decision": 92.49, "sb_at_decision": 395.83, "co_at_decision": 100, "utg_at_decision": 71.93, "hero_start_derived": "99.50 — 96.37 behind at the turn node plus 1.13 (flop bet) plus 2 (preflop raise); HUD line prints 'Hero / 99 BB'", "villain_start_derived": "95.62 — 92.49 behind at the turn node plus 1.13 plus 2; HUD line prints 'P2-30 / 96 BB'", "effective_note": "Effective depth at the decision is the villain's 92.49 BB behind (~95.6 BB starting). Hero and villain are NOT the same depth, which is why the context line quotes no single 'N BB DEEP' figure — the felt shows each seat its own true stack instead." }, "result": "Hero bet 4.82 BB into the 6.76 BB pot (71.3%), villain called, river K♥, pot 16.39 BB (log + frame t=1006). Recorded for provenance; NOT displayed and NOT graded — this quiz grades Tyler's review verdict at the turn node, exactly as kt-top-two-bet-the-scare-turn does. See unverifiable_claims for the tension between that 71.3% and his spoken 'So I went 60'." } }, "context": "TYLER'S OWN HAND · $5 NL 6-MAX, FIVE-HANDED · BUTTON vs AN UNKNOWN IN THE BIG BLIND", "hero": { "pos": "BTN", "cards": [ "Ks", "8d" ], "stack_bb": 96.37, "avatar": "lono_ghoul", "button": true }, "seats": [ { "pos": "SB", "avatar": "ghoul_2942", "stack_bb": 395.83, "folded": true }, { "pos": "BB", "avatar": "ghoul_1776", "name": "LIKELY REC", "stack_bb": 92.49, "villain": true }, { "pos": "UTG", "avatar": "ghoul_1257", "stack_bb": 71.93, "folded": true }, { "pos": "CO", "avatar": "ghoul_4505", "stack_bb": 100, "folded": true } ], "streets": [ { "name": "preflop", "actions": [ { "pos": "UTG", "act": "fold" }, { "pos": "CO", "act": "fold" }, { "pos": "hero", "act": "open", "to_bb": 2 }, { "pos": "SB", "act": "fold" }, { "pos": "BB", "act": "call" } ], "pot_bb": 4.5 }, { "name": "flop", "board": [ "8c", "Tc", "8s" ], "actions": [ { "pos": "BB", "act": "check" }, { "pos": "hero", "act": "bet", "to_bb": 1.13 }, { "pos": "BB", "act": "call" } ], "pot_bb": 6.76 }, { "name": "turn", "board": [ "8c", "Tc", "8s", "2h" ], "actions": [ { "pos": "BB", "act": "check" } ] } ], "prompt": "K♠8♦ · YOU FLOPPED TRIPS AND HE CALLED · THE 2♥ TURNS AND HE CHECKS AGAIN — HOW BIG?", "pot_bb": 6.76, "options": [ { "id": "half", "label": "BET ABOUT HALF POT" }, { "id": "pot", "label": "BET POT" }, { "id": "overbet", "label": "OVERBET" } ], "correct": "half", "option_feedback": { "half": "Correct. Your customer on this card is a ten, and he holds far more tens than eights — Tyler puts the ratio at three to one at least, and the eight in your own hand is exactly why: it is one of the eights he cannot have. So price the bet for the ten, not for the eight you are hoping he has. Half the pot keeps every ten and every club draw paying you, and it is the one size he might attack with a raise — getting bluff-raised is Tyler's other way to win here. He talks the sizes through out loud and lands right there: \"you could consider half.\"", "pot": "Too big. Pot-sized runs off the club draws that would have paid you twice more, and the tens — the hands you are actually farming — stop paying at the end. Tyler's aim on this card is to be called again and again by a ten; a bet this size buys one call and then leaves you with nothing good to do on the last card.", "overbet": "This is Eric's read — the board is connected and wet, so charge the draws. Tyler's answer is that an overbet charges them right out of the hand: \"you probably get ten-x to call, but you probably lose all his draws.\" And once the pot is that size he is not calling a second big bet with a ten, so the three streets of value that make trips worth flopping are gone." }, "verdict_text": "The only reasonable way to play this hand is to pick a sizing that they might bluff-raise. Our holy grail here is getting called three times by a ten — or getting bluff-raised. On this trips board, what I really want to do is make sure I'm betting a sizing that keeps all of his tens in for as many streets as I can, and hopefully gets raised.", "aired_transcript": "I think the only reasonable way to play this hand is to pick a sizing that they might bluff raise because our holy grail here is actually called three times from a 10 or get bluff raised. So on this trip sport here, what I really wanna do is I wanna make sure I'm betting a sizing here that keeps all of his 10s in as many a size as I can and hopefully gets raised.", "aired_transcript_note": "Verbatim faster-whisper small.en re-ASR of the re-cut quiz/audio/trips-block-bet-small-1088.m4a (19.89s, 2026-08-08). Two artefacts of the tape, both harmless and both shipped unchanged: the clip's 0.12s pre-handle catches the tail of the trimmed 'I think', which the model renders as a clean 'I think' lead-in; and 'trips board' / 'for as many streets as I can' come back as 'trip sport' / 'as many a size as I can' (a second decode of the same file returned 'as many ace highs as I can') — the audio is genuinely mushy there. No card, no bet size, no percentage, no population figure and no street the fixture does not show is aired.", "principle": "On a paired board the card in your hand decides who your customer is: hold one of the trips cards and he has far more top pair than the fourth card — so size for the top pair. Big enough that a raise is possible, never big enough to fold him.", "concept": "trips-card-removal-sizing", "concepts_secondary": [], "topic": "bet-sizing", "difficulty": 4, "decision": { "pot_settled_bb": 6.76, "bets": {} }, "pct_provenance": {}, "pct_provenance_note": "Deliberately empty: this rebuild displays NO percentage anywhere. The figures the tape offers at or near this node are all attached to something other than the graded decision — 42.69% / 57.30% (HoldEq, on an invented board), '51 percent against eights or better' (s243, same hypothetical), 'nine big blinds' from J-T (s200-201, a whole-hand value estimate annotated 'JT ==> 9bb'), and the s320-330 call frequencies (RIVER sizing work). Every one of them would be a figure from another node wearing this node's label, so none is shown. The one figure that IS this node's — the 6.76 BB pot — is a pot, not a percentage, and it is rendered on the felt.", "size_claims": { "small": { "street": "flop", "actor": "hero", "bet_bb": 1.13, "pot_before_bb": 4.5, "pct_of_pot": 25.1, "note": "The word 'small' survives only in the quiz's URL slug, and it names the FLOP bet, which really is small: 1.13 BB into the 4.50 BB pot = 25.1%. Tyler's own words for it: 'I go small on ten eight eight with King eight' (s183). It is NOT the graded turn size — the shipped fixture's answer label 'BET SMALL' has been retired, along with the invented 3 BB flop bet (55% of a wrong 5.5 BB pot) that flatly contradicted him.", "source": "own frames t=676/t=684 (log 'Hero bets 1.13 BB'; 4.50 BB centre chips; 'Pot: 5.63 BB'); SER NutsandBetsizing s183" }, "half pot": { "street": "turn", "actor": "hero", "bet_bb": 3.38, "pot_before_bb": 6.76, "pct_of_pot": 50.0, "note": "OPTION SIZE, not a bet anyone made: the graded answer 'BET ABOUT HALF POT' modelled at the midpoint of the pot at this node (3.38 into 6.76). Tyler's spoken target for this turn: 'it probably works out to 50 or 60... you could consider half. You could consider 40' (s264-266), after Eric offers 'the one third bet, 40% pot at most' (s263).", "source": "own frame t=702 ('Pot: 6.76 BB'); SER NutsandBetsizing s263-s266" }, "pot-sized": { "street": "turn", "actor": "hero", "bet_bb": 6.76, "pot_before_bb": 6.76, "pct_of_pot": 100.0, "note": "OPTION SIZE, not a bet anyone made: the 'BET POT' distractor modelled at exactly the 6.76 BB pot. Tyler's objection covers it — 'structurally on the river, he's not going to call many 10 X's to the over bet or a big bet' (s196).", "source": "own frame t=702 ('Pot: 6.76 BB'); SER NutsandBetsizing s196" }, "pot-size": { "street": "turn", "actor": "hero", "bet_bb": 6.76, "pot_before_bb": 6.76, "pct_of_pot": 100.0, "note": "Alias of the 'pot-sized' entry above — the validator matches 'pot-size' and 'pot-sized' as two separate size words, and the word 'Pot-sized' in option_feedback.pot contains both. Same modelled option size, same referent.", "source": "own frame t=702 ('Pot: 6.76 BB'); SER NutsandBetsizing s196" }, "overbet": { "street": "turn", "actor": "hero", "bet_bb": 10.14, "pot_before_bb": 6.76, "pct_of_pot": 150.0, "note": "OPTION SIZE, not a bet anyone made: the 'OVERBET' distractor modelled at 1.5x the 6.76 BB pot. This is the line Eric proposes on tape — 'this one I would also over bet because it's pretty connected and wet' (s191) — and the one Tyler spends the whole passage refusing.", "source": "own frame t=702 ('Pot: 6.76 BB'); SER NutsandBetsizing s191-s196" } }, "claims": [ { "text": "the decision node: K♠8♦ on 8♣T♣8♠-2♥, villain has checked, 6.76 BB in the middle", "basis": "Log verbatim through 'Turn: [2h] / P2-... checks'; felt prints 'Pot: 6.76 BB' with 6.76 BB of centre chips and the villain tagged CHECK; hero is last to act because 'Seat 5 (Hero) is the dealer'. Pot rebuilds exactly: 0.50 dead SB + 2 + 2 = 4.50 preflop, + 1.13 + 1.13 = 6.76.", "source": "own frames t=676 (log head), t=684 (flop node), t=702 (turn node)" }, { "text": "hero holds K♠8♦ and the flop is TWO-TONE (two clubs), so live club draws are part of the villain's calling range", "basis": "Log 'Flop: [8c Tc 8s]'; the 4-colour felt shows 8 and T on GREEN (club) cards and the second 8 on a WHITE (spade) card, hero's K on a WHITE card with a spade pip and his 8 on a BLUE card with a diamond pip. Tyler names the hand out loud: 'Okay, King eight' (s182). The shipped fixture's rainbow 'Th 8s 8c' deleted exactly the draws the argument leans on (Eric s191 'it's pretty connected and wet'; Tyler s194 'you probably lose all his draws').", "source": "own frames t=684 (hero684.png 3x, log684_a.png), t=702" }, { "text": "he has far more tens than eights — 'at least three to one' — and the eight in hero's hand is why", "basis": "Tyler asks 'So how many eights does he have here?' (s205), Eric answers 'Not nearly as many as like 10' (s206), Tyler: 'Yeah, it's at least three to one again, right?' (s207) and then supplies the removal: 'I'd point out that you have an eight in your hand. We have an eight.' (s210-211), 'So you've got a whole bunch of trailing one combo' (s213). The pixels agree with the direction: the board shows 8♣ and 8♠ and hero holds 8♦, so exactly one eight (8♥) is left in the deck while both remaining tens are live. The fixture ships Tyler's stated ratio, not a computed one.", "source": "SER NutsandBetsizing s205-s213; own frames t=684/t=702 for the cards" }, { "text": "the villain is an unknown the coach reads as a recreational player", "basis": "'It's an unknown player. He's like he's more likely to be a rec based on stack size.' (s184-185) — the seat tag 'LIKELY REC' is those words and nothing more. The stack corroborates the read: he sat down for 95.62 BB rather than a round 100, and the replayer shows him as an anonymized hash (P2-3055527b6...) with no HUD history ('VP 0 / PR 0 / 3B 0 / CB -', sample '(2)').", "source": "SER NutsandBetsizing s184-s185; own frames t=684/t=702 (seat plate + HUD line)" }, { "text": "the overbet is Eric's proposal and Tyler's refusal of it is the substance of the quiz", "basis": "Eric: 'this one I would also over bet because it's pretty connected and wet' + 'they call all their tens even for a large size' (s191-192). Tyler: 'you probably get 10 X to call, but you probably lose all his draws right or a lot of them' (s194) and 'just structurally on the river, he's not going to call many 10 X's to the over bet or a big bet' (s196). The aired verdict is the positive form of the same ruling.", "source": "SER NutsandBetsizing s191-s196" }, { "text": "the table is five-handed at a six-seat ring and hero is the button", "basis": "Log line 2 verbatim: 'Seat 5 (Hero) is the dealer'. Five seat plates are occupied (P6 395.83, P2 93.62/92.49, hero 96.37, P4 100, P3 71.93) and the top-right seat is drawn as an empty plate. The shipped fixture invented a sixth player (an HJ that folds) and equal 94/94 stacks.", "source": "own frames t=676 (log), t=684, t=702, emptyseat702.png" }, { "text": "every amount in this fixture is a BB figure printed by the replayer; no currency conversion was performed", "basis": "The replayer log and felt are in BB display mode: 'posts the small blind 0.50 BB', 'Hero raises to 2 BB', 'Hero bets 1.13 BB', 'Pot: 6.76 BB', '96.37 BB'. The two dollar figures in the window ('$39.22' in the hand selector, 'Avg Stack: $830' in the left HUD margin) have no on-screen mapping to this hand and are used for nothing; '$5 NL (6 max)' is quoted as the table's name.", "source": "own frames t=676, t=684 (title684.png), t=702, t=1006" }, { "text": "what actually happened after the graded node (recorded, not displayed, not graded)", "basis": "Log verbatim: 'Hero bets 4.82 BB / P2-... calls 4.82 BB / River: [Kh]'; felt 'Pot: 16.39 BB', hero 91.56 BB, villain 87.67 BB. 4.82 into 6.76 is 71.3% of pot. The quiz stops at the villain's turn check and grades the review verdict, not the line taken.", "source": "own frame t=1006 (log1006.png 4x, f_1006.png)" } ], "unverifiable_claims": [ "'So I went 60. 60 [BB?] here.' (s267-268) does not match the log: hero bet 4.82 BB into 6.76 = 71.3%, not 60%. His red note on the screenshot near hero's seat is a two-glyph scribble that reads as either '6%' or '60' (own zoom ann_left.png, 8x) and cannot settle it. Neither figure is displayed; the fixture grades the sizing he argues for out loud (s263-266, 'you could consider half', 'it probably works out to 50 or 60') and records the executed 71.3% in hand_record.result.", "'if he has Jack 10 here, you make nine big blinds. I mean, that's roughly the GTO, maybe a little bit less.' (s200-201) — an estimate of the whole-hand value of J-T, hand-annotated on the screenshot as 'JT ==> 9bb'. It is not this decision's price and is attached to no sim on screen; not displayed.", "TWO HoldEq readouts, on TWO different hand-typed boards — both hypothetical, both excluded, neither aired. (i) 'It's only 51 percent against eights or better on a blank' (s243) is read off board 'Tc8c8d4h2s', Range 1 'K8o' = 51.027% vs a weighted 8x list 48.973% (own frames t=899, t=902, zoom eq_902.png 4x). That board is this hand's flop+turn with the suits RE-TYPED (8d for the real 8s, 2s for the real 2h) plus a BLANK 4h standing in for a river nobody had seen at the graded node — a hypothetical runout, not a figure at the turn decision. (ii) Only afterwards — 'if you move this card up ... pick something a little more connecting' (s247), 'It's now behind' (s249) — does he retype it to 'JsTc8c8d2s', carrying a jack this hand never dealt, and K8o drops to 42.693% vs 57.307% (own frame t=916, zoom eq916b.png 4x). Both boards were built while he says 'I don't know what we want the river to be' (s232) and 'I guess I need to get the suits right' (s240); excluded from the fixture and from the re-cut audio. [Corrected 2026-08-09 fidelity gate: the previous wording attached the 51% to the jack board — my own frames show the jack was not on the board until ~t=912.]", "Everything from s277 on ('against 8x we probably maximize with 2x pot', '10x... three quarters pot', 'King X... 1.25x pot', the 2%/38 BB/54% arithmetic) is spoken over a RIVER screenshot (board 8-T-8-2-K, 'Pot: 16.39 BB', annotated '8x ==> 2x pot / Tx ==> 3/4 pot / Kx ==> 1.25x pot', own frame t=1160). It is river sizing for a different node and is excluded — none of it is aired or displayed.", "Tyler's 'he's clearly like not good, so let's take advantage of that' (s180-181) sits ~14s before this hand is opened at s182, with no frame confirming which player it refers to; it is not used and the villain tag draws only on s184-185." ], "cosmetic_invention": [ "Ghoul avatars on every seat (lono_ghoul for hero) — art only; the replayer shows anonymized hashes (P2-3055527b6..., P6-9d69ed76f..., P3-f45672775..., P4-c1ef03b73...). The four seat avatars are carried over unchanged from the shipped fixture; the fifth (ghoul_2344) is dropped with the invented HJ seat.", "Seat LABELS SB / BB / UTG / CO are the standard five-handed names for the four occupied non-hero seats, derived from the log rather than shown by it: 'Seat 5 (Hero) is the dealer' fixes the button, P6 and P2 are named by their posts, and P3/P4 take the two remaining five-handed seats in the order they act preflop.", "The empty sixth seat of the six-seat ring is not rendered; the fixture declares table_size 5 and the context line says FIVE-HANDED.", "Option sizes: 'BET ABOUT HALF POT', 'BET POT' and 'OVERBET' are modelled in size_claims at 3.38 / 6.76 / 10.14 BB into the 6.76 BB pot so the size words are checkable. Only the pot they are measured against is real; the three amounts are the options' own definitions, not bets anyone made." ], "sourced_not_invented": [ "Hero's cards K♠8♦ — 4-colour card faces at 3x (white card + spade pip, blue card + diamond pip), corroborated by Tyler saying 'Okay, King eight' (s182). Load-bearing: the 8♦ IS the blocker the whole lesson turns on.", "The board 8♣ T♣ 8♠ / 2♥ with suits — log verbatim 'Flop: [8c Tc 8s]' and 'Turn: [2h]', confirmed by the felt's 4-colour cards. Load-bearing twice over: the two clubs are the draws the overbet-feedback charges, and the third 8 is what makes the hand trips.", "Hero on the BUTTON, villain in the BIG BLIND — log 'Seat 5 (Hero) is the dealer' and 'P2-... posts the big blind 1 BB'. Load-bearing: hero acts last, which is why the graded decision exists at all.", "Every size in the tree — 0.50 / 1 blinds, open to 2 BB, flop bet 1.13 BB, flop call 1.13 BB — all log-verbatim; and the pots 4.50 BB (centre chips at t=684) and 6.76 BB (top-line readout at t=702).", "Stacks: hero 96.37 BB and villain 92.49 BB behind at the decision, SB 395.83, CO 100, UTG 71.93 — all read off seat plates in own frames.", "Five players dealt at a six-seat ring, no straddle, no ante — five occupied plates plus one empty, and a log whose only forced bets are the two blinds.", "The villain read 'LIKELY REC' — Tyler's words at s184-185 ('It's an unknown player... more likely to be a rec based on stack size'). The shipped fixture's tag 'FISH' is removed: he never says it.", "The stake string '$5 NL (6 max)' and hand id #4652230422 — title bar verbatim at 3x." ] }