{ "schema_version": "1.3", "table_size": 5, "id": "flat-straight-draw-multiway", "format": "multi_option", "free": true, "rebuild": "v2", "provenance_class": "dealt_hand", "updated_at": "2026-08-09T00:00:00Z", "source": { "session_id": "MoreHandReviewwithCurrentHands", "verdict_audio": "quiz/audio/flat-straight-draw-multiway.m4a", "lesson_slug": null, "master_slug": "MoreHandReviewwithCurrentHands_master", "master_t": 6058.3, "src_t": 6902.7, "ser_sentences": [ 767, 769, 770, 771, 772, 773, 779, 780, 781, 782, 783 ], "hand_record": { "lane": "FRAME_READ", "played": "A hand from the student's own play, replayed on tape. Eric rewinds to it explicitly (s767 'actually kind of want to go back… so why was the flop raise bad') and then steps the PokerTracker 4 replayer BACKWARDS through the flop, action by action, so the exact decision node is on screen. Hero really raised the flop to 10 BB; the quiz stops one action earlier, at hero's decision, and grades Tyler's review verdict (flat) rather than the line taken. Solver note: at the literal verdict words (t=6902.7) Tyler has swapped to two Windows Scientific Calculators to sketch his EV arithmetic by hand — that is a calculator, not a solve, and the configuration being judged is the replayer's dealt hand, so this is dealt_hand, not model_spot.", "hand_id": "replayer title bar 'Playback for … Hand #4xxxxxxxxx (2) - $1 NL (6 max) - Holdem' (operator name omitted per policy rule 2). The hand-number digits sit below the legibility floor of this 1324x720 capture: a 70-frame temporal average of the title bar (t=6775-6810 at 2 fps) still resolves the word shapes but not the digits.", "stake": { "source": "TITLE_BAR (provenance only)", "usd_per_bb": null, "note": "No conversion was performed and no dollar figure ships. The replayer is running in DISPLAY-IN-BB mode: every felt readout and log line is already BB-native ('raises to 2 BB', 'bets 3 BB', 'calls 1.50 BB', 'Pot: 9 BB', '96.50 BB'), so the fixture's bb figures are transcribed, not computed. For the record only, the title bar reads '$1 NL (6 max)' and the table overlay reads 'Avg Stack: $110' against stacks of 98-120 BB, consistent with $1 = 1 BB — but nothing in this fixture depends on that.", "screening_note": "The dp1/t1 screening record read this title bar as '31 NL'. My own 70-frame average reads '$1 NL', and the Avg Stack overlay corroborates ~$1/BB. Either way it is provenance only." }, "frames": [ "RE-DERIVED 2026-08-09 from independently extracted frames (scratchpad fix1/frames/n_6756.7.png + zooms z_pot/z_hero_cards/z_board/z_p4/z_p5.png, and the 13-frame log average log_avg_raw.png): 'Pot: 9 BB', 'Odds: 3.00:1 - 25%', hero 9 on BLUE with ♦ + 7 on GREEN with ♣ over a GREEN '98 BB', board 6 GREY ♠ / 8 GREEN ♣ / Q GREEN ♣, P5-3370e8d6d tagged BET with a '3 BB' chip and a '6 BB' centre chip, P4-6edd5eb03 '96.50 BB' STILL HOLDING TWO CARDS with no fold tag. Every load-bearing number below re-read and confirmed.", "cache/MoreHandReviewwithCurrentHands.mp4 t=6756.7 — THE DECISION NODE (scratchpad dh3/flat-straight-draw-multiway/frames/DECISION_6756.7.png). Verbatim felt: board 6♠ 8♣ Q♣; hero 9♦7♣ 'Hero - 38%' with a GREEN (to-act) stack readout '98 BB' and NOTHING in front; P5-3370e8d6d tagged 'BET' with a '3 BB' chip in front; P4-6edd5eb03 '96.50 BB' STILL HOLDING CARDS with the dealer button beside him (no fold tag — this is the pixel proof that the opener is still to act behind hero); centre pot chip '6 BB'; readout 'Pot: 9 BB' and 'Odds: 3.00:1 - 25%' (the replayer's own statement of hero's price: 3 to call into 9). P3 and P1 show stacks but no cards.", "t=6756.1-6756.5 frames/pot_sheet.png — the same rewind one step earlier: 'Pot: 6 BB' with no bets out, i.e. the flop dealt into the 6 BB preflop pot", "t=6757 / t=6758 frames/pot_sheet.png, felt_6758.png — the two steps AFTER the decision: hero's '10 BB' raise chip, 'Pot: 19 BB' with 'Odds: 1.90:1 - 34%' (P4 facing 10 into 19), then P4 tagged 'FOLD' with 'Odds: 2.71:1 - 27%' (P5 facing 7 into 19). Three independent odds readouts triangulate the 3 BB lead.", "t=6800 frames/f_6800.png (+ zooms hero_cards_10x.png, board_8x.png, hero_plate.png, p5_plate.png, bottom_seats.png, right_seats.png) — the turn state: 'Pot: 26 BB', turn 8♥, Hero 88 BB / HUD 'Hero / 100 BB / (11)', P5 96.51 BB / HUD '109 BB / (2)', P4 96.50 BB / HUD '98 BB / (2)', P3 119.75 BB, P1 115.03 BB, and the bare name-less EMPTY seat plate on the right rail. Card suits read off PokerTracker's 4-colour tiles: 9 on BLUE with a ♦ glyph, 7 on GREEN with a ♣ glyph; 6 GREY ♠, 8 GREEN ♣, Q GREEN ♣, 8 RED ♥.", "t=6772 / t=6800 frames/log_6772.png, log_6800_6x.png, lg/lines16.png (60-frame temporal average, per-line strips) — the action log", "t=6742 frames/log_6742.png — log scrolled to its end, the full runout through 'Action has ended' (used only to count total log lines, see log_lines_audit)", "t=6746 frames/felt_6746.png — NEGATIVE/boundary frame: a DIFFERENT hand (hero T♣J♠, 'Pot: 4 BB', P3 tagged RAISE) is briefly on screen while Eric loads 'one more', before he says 'actually kind of want to go back' (s767) and reloads this hand at ~t=6750. Nothing from that hand is imported." ], "log_verbatim": [ " Hand #… begins at 2021/12/31 5:13 am' — operator name and hand digits omitted per policy rule 2 / hand_id>", "Seat 4 (P4-6edd5eb03f9c2355b07b9f4edfa5574a) is the dealer", "P5-3370e8d6d… posts the small blind 0.50 BB", "Hero posts the big blind 1 BB", "P1-a4bf2019d… folds", "P3-b026f44b1… folds", "P4-6edd5eb03… raises to 2 BB", "P5-3370e8d6d… calls 1.50 BB", "Hero calls 1 BB", "Flop: [6s 8c Qc]", "P5-3370e8d6d… bets 3 BB", "<<< THE QUIZ'S DECISION POINT (felt at t=6756.7): hero faces 3 into a 9 BB pot with P4/BTN still holding cards behind him >>>", "Hero raises to 10 BB", "P4-6edd5eb03… folds", "P5-3370e8d6d… calls 7 BB", "Turn: [8h]", "P5-3370e8d6d… checks", "Hero bets 21 BB / P5 calls 21 BB / River: [Jh] / P5 checks / Hero bets 67 BB / P5 calls 67 BB / Action has ended" ], "price_audit": "The pre-audit fixture and the t1/dp1 screening record BOTH had this price as a 1 BB lead into 6 (pot 7). That is wrong: the log line is 'bets 3 BB' and five independent on-screen numbers agree — the '3 BB' chip in front of P5, the 6 BB centre chip, 'Pot: 9 BB', 'Odds: 3.00:1 - 25%' (3 to call into 9) at the decision, and 'calls 7 BB' after hero's raise to 10 (10 − 3 = 7). The downstream arithmetic closes: 6 + 3 + 10 = 19 = the 'Pot: 19 BB' readout after the raise, and 6 + 10 + 10 = 26 = the 'Pot: 26 BB' turn readout. Corrected here.", "log_lines_audit": "CORRECTED 2026-08-09 (re-derived from own frames): NOTHING in this hand's preflop log is hidden or reconstructed. At the decision node itself the action-log widget is scrolled to the TOP and the whole hand-to-date fits in it, 11 lines, read directly off a 13-frame temporal average of t=6756.58-6756.82 (the interval over which the felt is bit-identical; the widget changes at 6756.58 when 'bets 3 BB' posts and again at 6756.86 when hero's raise posts) — scratchpad fix1/frames/log_avg_raw.png + per-line strips A0-A9.png, source frame n_6756.7.png: (1) ' Hand #… begins at 2021/12/31 5:13 am', (2) 'Seat 4 (P4-6edd5eb03f9c2355b07b9f4edfa5574a) is the dealer', (3) 'P5-3370e8d6d… posts the small blind 0.50 BB', (4) 'Hero posts the big blind 1 BB', (5) 'P1-a4bf2019d… folds', (6) 'P3-b026f44b1… folds', (7) 'P4-6edd5eb03… raises to 2 BB', (8) 'P5-3370e8d6d… calls 1.50 BB', (9) 'Hero calls 1 BB', (10) 'Flop: [6s 8c Qc]', (11) 'P5-3370e8d6d… bets 3 BB'. The pre-fix version of this field inferred ONE hidden line from a scrollbar-thumb ratio measured at t=6772 (a LATER, scrolled state) and its stated calibration did not even close (22/57 = 0.386, not the 11/19 = 0.579 it claimed). The conclusion it reached — UTG's fold is the missing first action — happens to be right, but it is now READ, not reconstructed, so the inference is retired rather than repaired.", "seat_map": "Felt geometry (frames/f_6800.png, DECISION_6756.7.png), clockwise: Hero top-centre → P1-a4bf2019d top-right → EMPTY plate right rail → P3-b026f44b1 bottom-right → P4-6edd5eb03 bottom-left → P5-3370e8d6d left. Positions are forced by the log, not guessed: 'P5 calls 1.50 BB' can only be a small blind completing to 2 (0.50 posted) and 'Hero calls 1 BB' can only be the big blind (1 posted), so P5 = SB and Hero = BB; the raiser immediately before the SB is therefore the BTN = P4 — and the log STATES it outright in its own header block, 'Seat 4 (P4-6edd5eb03f9c2355b07b9f4edfa5574a) is the dealer' (read directly at the decision node, 2026-08-09), corroborated by the dealer-button disc sitting beside P4's plate. Walking back from the button: P3 = CO, the empty plate = the vacant HJ/MP seat, P1 = UTG. That also matches the log order (P1 acts before P3), and the postflop order the replayer actually executes (P5 → Hero → P4).", "stacks": "Hero's and the BTN's stacks at the decision are READ DIRECTLY off the node frame (t=6756.7): Hero '98 BB', P4/BTN '96.50 BB'. P5/SB's plate is covered by his 'BET' tag at that instant, so his is derived from the turn frame: 96.51 BB shown after investing 12 (2 preflop + 3 lead + 7 call) = 108.51 start, minus the 5 he has in at the decision = 103.51 behind. Starting stacks: Hero 100.00 (HUD 'Hero / 100 BB' agrees), SB 108.51 (HUD rounds to 109), BTN 98.50 (HUD rounds to 98), CO 119.75 and UTG 115.03 (never invest). No single depth is declared in the context line because the three live stacks genuinely differ.", "result": "In game hero RAISED to 10 BB, the BTN folded, the SB called 7; the turn was 8♥, the river J♥, and hero barrelled 21 then 67 into a 202 BB pot and was called with nine-high. None of that is displayed or aired — the quiz ends at the flop decision and grades Tyler's verdict on it. Eric's own framing (s767 'why was the flop raise bad') is what re-opens the spot.", "render_convention": "Per-street pot_bb is the POST-action pot (preflop 6). The flop street carries the SB's live 3 BB lead and no pot_bb, so the site renderer shows 3 in front against a settled pot of 6 — decision.pot_settled_bb 6.0 / bets {SB: 3} / top-level pot_bb 9, which is the replayer's own 'Pot: 9 BB'." } }, "context": "5-HANDED CASH · BTN OPEN · 3-WAY FLOP", "hero": { "pos": "BB", "cards": [ "9d", "7c" ], "stack_bb": 98, "avatar": "lono_ghoul" }, "seats": [ { "pos": "UTG", "avatar": "ghoul_942", "folded": true }, { "pos": "CO", "avatar": "ghoul_355", "folded": true }, { "pos": "BTN", "avatar": "ghoul_5170", "name": "OPENER", "stack_bb": 96.5, "button": true }, { "pos": "SB", "avatar": "ghoul_4504", "stack_bb": 103.51, "villain": true } ], "streets": [ { "name": "preflop", "actions": [ { "pos": "UTG", "act": "fold" }, { "pos": "CO", "act": "fold" }, { "pos": "BTN", "act": "open", "to_bb": 2 }, { "pos": "SB", "act": "call" }, { "pos": "hero", "act": "call" } ], "pot_bb": 6 }, { "name": "flop", "board": [ "6s", "8c", "Qc" ], "actions": [ { "pos": "SB", "act": "bet", "to_bb": 3 } ] } ], "prompt": "9♦7♣ · OPEN-ENDER · 3-WAY · THE SB LEADS HALF-POT AND THE BTN OPENER IS STILL BEHIND YOU — RAISE OR CALL?", "pot_bb": 9, "options": [ { "id": "raise", "label": "RAISE" }, { "id": "call", "label": "CALL" } ], "correct": "call", "option_feedback": { "raise": "This is the raise Tyler says he has made about never. Work his two cases. Case one, the pot stays two-handed: the raise is worth about a big blind at most, because the player leading into you doesn't tend to fold, so you rarely just take it down. Case two, the opener behind you wakes up and three-bets or flats — and that case costs enough to take that whole big blind back. Add them together and the raise is worth basically nothing.", "call": "Correct — flat. Everything the raise could be worth sits in the case where the pot stays two-handed, and the player still to act behind you is exactly what takes it away. Flatting turns that same player into an asset: if the opener comes along too, your draw is aimed at a pot with two stacks in it — spike a five and two people can pay you, not one." }, "verdict_text": "The case where the opener behind us wakes up is essentially just taking a big blind away from us. So the primary decision with that sort of hand — one that doesn't have great card removal or great equity — is to flat. If we call and then the other guy calls, the value of our hand goes up: if we spike a five, we've got two people we can stack rather than just one.", "principle": "A straight draw is a flat, not a raise, when a player is still to act behind you: everything the raise is worth sits in the case where the pot stays two-handed, and the case where the player behind wakes up takes it straight back — while flatting invites a second stack into a pot your draw can win.", "concept": "flat-dont-raise-draw-multiway", "concepts_secondary": [], "topic": "multiway-pots", "difficulty": 4, "aired_transcript": "It's essentially just taking a big blind away from us. And so the primary decision here with that sort of hand that doesn't have great card removal or great equity is to flat. First, if we call and then the other guy calls, the value of our hand goes up. Yeah. Because if we spike a five, we've got two people who you can stack rather than just one.", "aired_transcript_note": "Verbatim faster-whisper small.en re-ASR of the RE-CUT quiz/audio/flat-straight-draw-multiway.m4a (22.17s; beats w16931-16963 + w17002-17035 per tools/cut_quiz_audio.py, re-cut + re-ASR 2026-08-08). RE-CUT rationale: the previous 35.3s splice (sentences 780-782 whole) carried Tyler's solver-jargon aside mid-clip — 'what we're trying to do is we're trying to basically lower the value of this node… and then the other thing here is the call node' — which is opaque to a first-time viewer and says nothing the verdict does not already say. Nothing was added: the new clip is a strict SUBSET of the old one, same tape, same order. ASR notes: the cutter's handles pull in the tape's own 'It's' (w16930) at the head and 'First' (w17001) at the seam, both Tyler's words in place; 'two people who you can stack' is an ASR variant of the tape's 'two people who can stack' (w17027-17031). Every aired claim is true of the frame-read hand — a five completes 9♦7♣ on 6♠8♣Q♣, and there really are two other stacks in the pot at the decision. No cards, positions, streets, sizes, prices or percentages are aired. RE-VERIFIED 2026-08-09 (independent re-ASR of the shipped .m4a, faster-whisper small.en, 22.17s, same words; and of the raw tape t=6891.5-6916.5): the clip's opening pronoun — 'IT'S essentially just taking a big blind away from us' — has its antecedent one sentence earlier on the tape, in s779, which the clip does not carry: 'the fact that we've got the secondary case where that other player enters the pot has a giant impact of the value of the raise.' The antecedent is therefore supplied by verdict_text, whose first clause names it ('The case where the opener behind us wakes up…'). s779 is deliberately NOT re-cut into the audio: its ASR/diction is mangled on tape ('answers the pot' for 'enters the pot', 'the race' for 'the raise') and airing it would cost 9.5s of muddle to say what the displayed line says cleanly. DISPLAY/AUDIO POLARITY (fix 2026-08-09): the payoff clause is Tyler saying hero can win two stacks. Tape words are 'two people who can stack rather than just one' (w17027-17035, s782, spk tyler/confident); the shipped m4a re-ASRs as 'two people who YOU can stack'. verdict_text used to render this as 'two people who can stack US', which inverts it — the display told the viewer he was the one being stacked while the same sentence claimed his hand's value went UP. Corrected to 'two people we can stack'.", "pct_provenance": {}, "size_claims": { "half pot": { "street": "flop", "actor": "SB", "pct_of_pot": 50.0, "source": "Log line 'P5-3370e8d6d… bets 3 BB' into the 6 BB preflop pot = 50.0%, confirmed on the felt at the decision node (t=6756.7): a '3 BB' chip in front of P5, a '6 BB' centre chip, 'Pot: 9 BB' and 'Odds: 3.00:1 - 25%'. The adjective is the fixture's, describing exactly that bet." } }, "claims": [ { "text": "preflop pot 6 bb", "lhs": "BTN 2 + SB 2 + hero 2", "op": "=", "rhs": "6 bb", "source": "log 'P4 raises to 2 BB / P5 calls 1.50 BB / Hero calls 1 BB' (0.50 + 1.50 = 2 for the SB, 1 + 1 = 2 for the BB); shown directly as the '6 BB' centre chip and as 'Pot: 6 BB' before the lead (frames/DECISION_6756.7.png, pot_sheet.png)" }, { "text": "the price at the decision: 3 to call into 9", "lhs": "SB's flop lead", "op": "=", "rhs": "3 BB into a settled 6 BB pot; hero is 3 to win 9, which the replayer states as 'Odds: 3.00:1 - 25%'", "source": "frames/DECISION_6756.7.png (t=6756.7): 'Pot: 9 BB', '3 BB' chip in front of P5 tagged BET, '6 BB' centre chip, hero's stack green with nothing in front. Log line 'P5-3370e8d6d… bets 3 BB'. BB-native readouts; no conversion." }, { "text": "the opener is still to act behind hero", "lhs": "flop action order", "op": "=", "rhs": "SB (bets) → Hero → BTN", "source": "at t=6756.7 P4-6edd5eb03 still HOLDS CARDS and shows no fold tag while hero is the player to act; the log then reads 'Hero raises to 10 BB' followed by 'P4-6edd5eb03… folds', and the intermediate frame t=6757 shows 'Odds: 1.90:1 - 34%' = P4 facing 10 into 19. P4 = BTN per seat_map (dealer button beside his plate)." }, { "text": "hero holds an open-ended straight draw and a five completes it", "lhs": "9♦7♣ on 6♠8♣Q♣", "op": "⇒", "rhs": "5 or T makes the straight; 5-6-7-8-9 is the one Tyler names on tape", "source": "pixels frames/hero_cards_10x.png + frames/board_8x.png (4-colour tiles: blue=♦, green=♣, grey=♠); spoken s781 'if we spike a five'" }, { "text": "three players are live at the decision", "lhs": "preflop continuers", "op": "=", "rhs": "BTN, SB, hero — UTG and CO folded preflop", "source": "log lines 1-5; at t=6756.7 only P5, P4 and hero hold cards (frames/DECISION_6756.7.png)" }, { "text": "stacks behind at the decision (displayed)", "lhs": "hero / BTN read directly; SB derived", "op": "=", "rhs": "98 / 96.50 / 103.51 bb", "source": "frames/DECISION_6756.7.png shows 'Hero … 98 BB' and 'P4-6edd5eb03… 96.50 BB' at the node itself; the SB's plate is covered by his BET tag, so 103.51 = 108.51 start (96.51 at the turn + 12 invested, frames/p5_plate.png) − 5 already in at the decision" }, { "text": "'the raise is worth about a big blind at most' / 'the raise is worth basically nothing' (option_feedback.raise), 'takes that whole big blind back' (option_feedback.raise) and 'taking a big blind away from us' (verdict_text)", "lhs": "Tyler's own EV sketch for THIS raise", "op": "=", "rhs": "an explicitly hypothetical two-branch sum he types into two Windows Scientific Calculators at t≈6842-6891 — 'let's say hypothetical… win a blind here when I raise flop if we're heads up' against 'let's say hypothetically I'm gonna lose something like eight big blinds when the other guy gets involved', concluding 'this raise that was once worth one big blind is now worth basically zero'; s779-780 then names what did it: 'the fact that we've got the secondary case where that other player enters the pot has a giant impact on the value of the raise — it's essentially just taking a big blind away from us'", "source": "s774, s775, s776, s778, s779 (tyler, spk_confident). The branch FREQUENCIES he pairs with those figures (s772/s773/s775/s776) are deliberately NOT displayed and NOT aired — see unverifiable_claims. COHERENCE NOTE (fix 2026-08-09): the display now runs ONE baseline end to end — the absolute value of the raise: case one is worth about a big blind AT MOST (his hypothetical, displayed as the upper bound he chose), case two takes that same big blind back, so the sum is 'basically nothing' and the aired 'taking a big blind away from us' is that subtraction. Tyler's looser phrasing at s770-s772 ('roughly neutral BETWEEN raising and calling') measures a different baseline (raise-vs-call), and mixing the two baselines is exactly what made the pre-fix text self-contradicting (neutral branch that something else could still 'kill'), so it is not displayed." }, { "text": "'the player leading into you doesn't tend to fold' (option_feedback.raise) — direction only, no magnitude", "lhs": "Tyler's read on the seat leading into hero", "op": "=", "rhs": "'it felt really hard to win the pot because when they don't, they don't tend to fold here'", "source": "s770 (tyler, spk_confident), words w16680-16699. Stated as a direction; no frequency is displayed or aired. Phrased as 'the player leading into you' (fix 2026-08-09) so the referent is the seat the viewer can see betting — the earlier wording 'that player type' pointed at a player-type label this fixture deliberately does not render (the SB ships as the bare VILLAIN fallback)." }, { "text": "'the raise Tyler says he has made about never' (option_feedback.raise)", "lhs": "Tyler on raising this hand class here", "op": "=", "rhs": "'heuristically here, I've raised this hand about never… maybe in 2006 I raised it'", "source": "s769 (tyler, spk_confident)" }, { "text": "the second case Tyler adds is the third player entering", "lhs": "the branch that kills the raise", "op": "=", "rhs": "'the other case is where player three enters the pot' / 'this dude wakes up and he three-bets us or flats us' — i.e. the BTN behind", "source": "s770, s773 (tyler, spk_confident)" } ], "unverifiable_claims": [ "RETIRED 2026-08-09 — this entry used to disclose UTG's fold as a reconstructed, hidden log line. It is not hidden: at the decision node the log widget sits at the top of the track and 'P1-a4bf2019d… folds' is directly legible, as is 'Seat 4 (P4-6edd5eb03…) is the dealer'. See hand_record.log_lines_audit. Nothing in the preflop history is inferred.", "The table is a 6-max table (title bar '(6 max)') with one seat visibly EMPTY, so five players were dealt in and the fixture renders 5-handed. Whether that seat was sitting out or between hands is not on screen.", "Villain profiling was deliberately NOT imported: the HUD samples are 2 hands for both P4 and P5 (frames/p5_plate.png, bottom_seats.png), and Tyler's 'these guys are oftentimes a lot more erratic with the small bets' (s763) belongs to the DIFFERENT hand under discussion before Eric rewinds at s767. Neither live seat carries a player-type name.", "Tyler's hypothetical branch frequencies for the EV sketch (s772, s773, s775, s776) and the arithmetic on his two on-screen calculators are recorded in claims for provenance only. They are not displayed and not aired: they are population-behaviour frequencies, and rule 7 gates the magnitude while the quiz teaches the direction (a player behind you can wake up, and that branch is what kills the raise).", "The replayer's hand number could not be resolved at this capture resolution (see hand_record.hand_id); identity rests on the frame sequence, the seat map, the HUD strips and the complete action log." ], "cosmetic_invention": [ "Avatars are art only: hero lono_ghoul; UTG ghoul_942, CO ghoul_355, BTN ghoul_5170, SB ghoul_4504 (carried over from the pre-audit fixture).", "The vacant 6-max seat is not rendered — the felt shows the five players who were dealt in, in their true clockwise order." ], "sourced_not_invented": [ "Hero's 9♦7♣ and the flop 6♠8♣Q♣ are read off PokerTracker's 4-colour tiles (blue+♦, green+♣, grey+♠) — frames/hero_cards_10x.png, board_8x.png. The suits are LOAD-BEARING (the two clubs are why Tyler's raising region here is club straight-flush draws and flush draws rather than bare straight draws, s783-s786), so they are sourced, not cosmetic.", "Every position is forced by the action log, not assigned: 'calls 1.50 BB' = SB, 'Hero calls 1 BB' = BB, raiser-before-the-SB = BTN (corroborated by the dealer-button disc beside P4's plate). The BTN's seat is the load-bearing one — the whole lesson is that he acts after hero, and at the decision frame he is still holding cards.", "All amounts (2 BB open, 1.50 call, 3 BB lead, 6 BB pot, 9 BB at the decision) are BB-native log/felt text, and the stacks (98 / 96.50 behind) are read off the decision frame itself. No $-to-BB conversion was performed anywhere.", "The BTN's seat label 'OPENER' is his role in the log ('P4-6edd5eb03… raises to 2 BB'), not a player-type read. The SB seat carries no name at all: the site renders its bare VILLAIN fallback, because Tyler states no read on that specific player and the HUD sample is 2 hands. The pre-audit fixture's 'FISH' tag is removed.", "The 5-handed structure is pixels: five occupied plates plus one bare, name-less, stack-less plate on the right rail (frames/right_seats.png). The pre-audit fixture's 6-max ring with a folding HJ is removed — that seat was never dealt a hand." ], "decision": { "pot_settled_bb": 6.0, "bets": { "SB": 3 } } }