{ "schema_version": "1.3", "table_size": 6, "id": "aces-geometric-pot-flop-vs-fish", "format": "multi_option", "free": true, "rebuild": "v2", "provenance_class": "model_spot", "updated_at": "2026-08-08T00:00:00Z", "source": { "session_id": "ComplicationofFishvsRegularsRanges", "verdict_audio": "quiz/audio/aces-geometric-pot-flop-vs-fish.m4a", "verdict_beats": "s677 (t=3304.6-3307.5) + s1126-1127 (t=5497.7-5504.7), gap-spliced via tools/cut_quiz_audio.py, 10.9 s. Beat referents: s677 is Tyler reading his typed \"QT5 vs rec ==> we bet bigger to extract more value\" Notepad line (on screen, frame t=3320); s1126-1127 is the rec-vs-reg model comparison for THIS spot, spoken moments before he names this node's own solver preference (\"the solver likes the small bet here\", s1132). Echo-checked 2026-08-08 (faster-whisper small.en): re-ASR of the shipped m4a returns \"We bet bigger to extract more value so with a recreational player we say geez like there's just not a lot of nut hands he's got a lot of one pair of combos we should bet big\" — matches verdict_text word for word except the decoder's \"one pair of combos\" vs SER s1126's \"one-pair combos\" (both are transcriptions of the same audio; the display follows SER). RE-CUT HISTORY: the previous clip (s1236-1237, \"...only one way to get three big streets of value out of aces, and that's to start with a pot on the flop\") was retired in pass 2 — it was spoken at the tail of the vs-REG exploitation passage (s1228-1235: \"you only can get two streets of value... there's nowhere near the pressing issue of betting really big on this flop\") and framed the key as a three-street stack-off, which Tyler's typed 16-SPR cap contradicts at this node's 19.6 SPR.", "lesson_slug": null, "solver_node": { "tool": "GTO Wizard (browser, app.gtowizard.com/solutions)", "file": "solution_type=gwiz · gametype=Cash6m500zGeneral · depth=100 · preflop_actions=F-R2-F-F-C-F · browser tab \"QdTh5s7c – Solutions – GTO Wizard\" at t=3600 and \"QdTh5s2c\" at t=2700 (the tab suffix is the turn card of whichever branch is open; the flop solution itself is QdTh5s)", "node": "Cash 100bb · 6max NL500 · General GTO · 3b:GTO — UTG Fold, HJ Raise 2, CO Fold, BTN Fold, SB Call, BB Fold → FLOP (pot 5) SB Check → HJ to act. HJ flop menu as read on screen: Check / Bet 1.65 (33%) / Bet 2.5 (50%) / Bet 3.75 (75%) / Bet 6.25 (125%) — the UI truncates the column with an ellipsis below 6.25, so an all-in row may exist off-view. SB's reply menu: Fold / Call / Raise 5.8 (50%) / Raise 9.95 (100%) / Allin 98 (1161%). Turn pot after the 33%-called line reads 8.3.", "video_t": [ 2700, 3600 ], "board": [ "Qd", "Th", "5s" ], "pot_chips": 5, "stacks_chips": { "HJ": 98, "SB": 98 }, "chips_per_bb": 1, "units_note": "GTO Wizard renders this solution in big blinds (header reads \"Cash 100bb\"; stacks read 100 preflop, 98 on the flop), so pot_chips/stacks_chips are BB counts read straight off the node and chips_per_bb is 1. No $↔bb conversion was performed: no currency amount appears anywhere on screen (the gametype string names NL500 but no blind/stake figures are displayed).", "board_suits_read": "4-colour deck at t=2700/t=3600: Q on a blue tile = ♦, T on a red tile = ♥, 5 on a grey tile = ♠ — matches the tab title QdTh5s.", "solver_choice_at_node": "The highlighted HJ action at this node is Bet 1.65 (33%). This quiz's keyed answer is NOT that output — it is Tyler's stated adjustment against a recreational caller (SER s1126-1127 'with a recreational player ... we should bet big'; s1132-1133 'the solver likes the small bet here ... whereas against the recreational player if we solve this out ... we didn't do the solve'). The vs-rec branch was never solved on tape; it is Tyler's reasoned counterfactual on this node.", "opponent_model": "78-VPIP REC — Tyler's typed assignment (\"EX: 78 VPIP\", Notepad, frame t=3320). The number originates as the real villain's HUD VPIP in the session's real A♣K♠ hand (Eric reads \"78 VPIP\" at s424, t=2001, with that hand history on screen). It is one assigned villain assumption, not a population measurement; the companion tight-reg model (\"2p+ 6.3%\", \"7.5 SPR with AA against Reg\") is deliberately not displayed.", "spr_cap": "Tyler's typed stack-off ceiling for AA vs this model on this board: \"16 SPR\" (Notepad, frame t=3320, directly under \"2.8% Two pair on QT5 -> AA has around 18% equity against the two pair combos\"). Derivation is on screen and on tape: calculator 0.972 ÷ 0.028 = 34.71 (frame t=2420), called the \"absolute maximum\" you could ever put in (s499-500), halved per his convention (\"we were dividing it in half and saying like we can get in like 15\" / \"Yeah, 17 17\", s507-508), typed as 16; he adds it is \"probably a little higher\" given AA's ~18% equity vs the two-pair combos (s509) and calls the get-it-all-in question \"ambiguous\" (s517). This node's SPR is 19.6 (98/5) — ABOVE the cap. Pass-2 reconciliation: the prompt no longer premises getting all 98 BB in, the keyed big bet is graded as Tyler's value-extraction adjustment (typed \"QT5 vs rec ==> we bet bigger to extract more value\", same frame), and the pass-1 \"6.25 → 21.9 → jam moves all 98 BB\" chain is no longer asserted anywhere in this fixture.", "not_this_hand": "The same session contains a real dealt hand on a DIFFERENT board (Anki front at t≈4020 / visible behind the notepad at t=3320): hero A♣K♠, flop Q♦5♠T♦, hero bets 1.88, and the SB wins with J♦5♦. That hand is unrelated to this fixture and is not depicted here. Do NOT source vs-rec sizing audio from s426-428 (\"pot this thing out and check river / pot it on the flop\", t=2004-2019): that ruling is spoken over the pasted hand history of that real hand (frame t=2010) and is about that hand, not this node." } }, "context": "SOLVER SPOT · 6-MAX 100BB · HJ OPEN, SB CALL · OPPONENT MODEL: 78-VPIP REC", "hero": { "pos": "HJ", "cards": [ "As", "Ah" ], "stack_bb": 98, "avatar": "lono_ghoul", "button": false }, "seats": [ { "pos": "CO", "avatar": "ghoul_261", "folded": true }, { "pos": "BTN", "avatar": "ghoul_365", "folded": true, "button": true }, { "pos": "SB", "avatar": "ghoul_3175", "stack_bb": 98, "villain": true }, { "pos": "BB", "avatar": "ghoul_3855", "folded": true }, { "pos": "UTG", "avatar": "ghoul_77", "folded": true } ], "streets": [ { "name": "preflop", "actions": [ { "pos": "UTG", "act": "fold" }, { "pos": "hero", "act": "open", "to_bb": 2.0 }, { "pos": "CO", "act": "fold" }, { "pos": "BTN", "act": "fold" }, { "pos": "SB", "act": "call" }, { "pos": "BB", "act": "fold" } ], "pot_bb": 5 }, { "name": "flop", "board": [ "Qd", "Th", "5s" ], "actions": [ { "pos": "SB", "act": "check" } ] } ], "prompt": "A♠A♥ · OVERPAIR ON Q-T-5 · THE 78-VPIP REC IN THE SB CHECKS INTO A 5 BB POT WITH 98 BB BEHIND — WHICH FLOP SIZE?", "pot_bb": 5, "options": [ { "id": "small", "label": "BET 1.65 — 33% POT" }, { "id": "check", "label": "CHECK" }, { "id": "big", "label": "BET 6.25 — 125% POT" } ], "correct": "big", "option_feedback": { "small": "1.65 is the size this node itself picks — Tyler: \"the solver likes the small bet here because it thinks we should be more protection focused.\" Against a regular that is the answer. Against this opponent model it is the wrong tool: bet 33% and get called three times and under 9 BB of a 98 BB stack has moved, so the chips stay in front of you.", "check": "Checking hands back the street this hand can least spare. Tyler, on this exact board against this model: \"on the flop you should bet big because right now you have a really strong hand that needs to extract value immediately before it gets worse.\" More two-pair combos arrive as the hand moves forward, and a 78-VPIP range pays flop bets with one pair.", "big": "Correct — Tyler's ruling for this exact matchup, typed into his session notes: \"QT5 vs rec ==> we bet bigger to extract more value.\" The 78-VPIP range he assigns this opponent holds only about 2.8% two-pair-or-better on Q-T-5 — one pair is most of what he has, and one pair pays the big bet. The same notes also cap the ambition: his stack-off ceiling for aces against this model is 16 SPR, and this spot is deeper than that at 19.6 — you are sizing up to extract more value, not because every chip is promised to get in." }, "verdict_text": "We bet bigger to extract more value. So with a recreational player we say, geez, like there's just not a lot of nut hands — he's got a lot of one-pair combos. We should bet big.", "principle": "Size to the range that is calling, not to the solver baseline: a 78-VPIP rec holds almost no two-pair-plus on Q-T-5, so one pair is what pays you — bet bigger to extract more value before the board gets worse. The solver's own pick here, the small 33% protection bet, is the answer against a regular, not against this range.", "pct_provenance": { "33": { "role": "sizing", "referent": "GTO Wizard's own label for the smallest HJ bet at THIS node (flop pot 5, stacks 98): \"Bet 1.65 (33%)\" — the highlighted branch. It is a sizing label, not an EV, an equity or a frequency.", "source": "solver node, own frame reads at t=2700 and t=3600 of cache/ComplicationofFishvsRegularsRanges.mp4 (flop column zoom, HJ 98 menu)" }, "125": { "role": "sizing", "referent": "GTO Wizard's own label for the largest readable HJ bet at THIS node: \"Bet 6.25 (125%)\". Used only as the big option's label. Tyler's vs-rec sizing ruling is qualitative (\"we should bet big\", \"we bet bigger to extract more value\") and the vs-rec branch was never solved on tape, so no solver output endorses 6.25 specifically — it is the big size this node's menu offers.", "source": "solver node, own frame reads at t=2700 and t=3600 of cache/ComplicationofFishvsRegularsRanges.mp4 (flop column zoom, HJ 98 menu)" }, "2.8": { "role": "range_composition", "referent": "Two-pair-or-better share of the 78-VPIP recreational range Tyler ASSIGNS this opponent on Q-T-5 — his own typed count, not a population/database measurement and not this node's solved range. Typed verbatim: \"EX: 78 VPIP / 2.8% Two pair on QT5\". His separate figure for a tight regular (6.3%, off a GTO Wizard range) is deliberately not displayed here. NOTE (pass-2 fidelity fix): Tyler's own use of this figure also sets his stack-off ceiling — calculator 0.972 ÷ 0.028 = 34.71 (frame t=2420), halved (\"we were dividing it in half\" / \"Yeah, 17 17\", s507-508), typed \"16 SPR\" directly beneath the 2.8% line (frame t=3320). That cap is why this fixture displays 2.8% as the reason to size up for value, NOT as a licence to get all 98 BB in at this node's 19.6 SPR — see source.solver_node.spr_cap.", "source": "Tyler's Notepad, own frame read at t=3320 of cache/ComplicationofFishvsRegularsRanges.mp4; restated on the Anki back \"about 2.8% of range\"" } }, "size_claims": { "small": { "street": "flop", "actor": "hero", "bet_bb": 1.65, "pct_of_pot": 33, "note": "Appears in option_feedback.small and the principle line, quoting Tyler's phrase \"the small bet\" for this node's 1.65 / 33% branch.", "source": "solver node, own frame reads at t=2700 and t=3600 of cache/ComplicationofFishvsRegularsRanges.mp4" }, "big": { "street": "flop", "actor": "hero", "bet_bb": 6.25, "pct_of_pot": 125, "note": "The keyed answer — the big button on this node's own menu. Tyler's vs-rec ruling is qualitative (\"we should bet big\" s1127, \"we bet bigger to extract more value\" s677 + the typed Notepad line); the vs-rec branch was never solved on tape, so 6.25 / 125% is cited as this menu's big size, not as a size Tyler named.", "source": "solver node, own frame reads at t=2700 and t=3600 of cache/ComplicationofFishvsRegularsRanges.mp4" }, "big (spoken, verdict_text)": { "street": "flop", "actor": "hero", "bet_bb": 6.25, "pct_of_pot": 125, "note": "Tyler's spoken size words in the verdict clip (\"bigger\", \"big\") carry no number; the fixture maps them to the 6.25 (125%) button because that is the big size on the menu the player is offered. Recorded so the spoken words are never read as naming a specific percentage.", "source": "quiz/audio/aces-geometric-pot-flop-vs-fish.m4a (10.9 s) + SER s677, s1126-1127" } }, "cosmetic_invention": [ "Hero's suits A♠A♥. The node and Tyler's analysis are class-level \"AA\"/\"aces and kings\" throughout, with no suit qualifier anywhere on the tape or in the typed notes. A♠A♥ is a legal combo on Q♦T♥5♠ and no spoken or displayed fact depends on which two aces.", "Ghoul avatars on all six seats (lono_ghoul for hero; ghoul_261, ghoul_365, ghoul_3175, ghoul_3855, ghoul_77 for the rest). A solver tree has no faces.", "The four fold branches (UTG, CO, BTN, BB) are rendered as seated players who fold. In the node they are fold branches of the preflop tree, not people; the BTN seat carries the dealer button purely so the felt reads as a legal 6-max table.", "The villain seat is deliberately unnamed (the site renders the neutral default). The only opponent description shown is the 78-VPIP recreational model Tyler typed and spoke, carried in the context line." ], "concept": "size-up-flop-vs-rec-range", "concepts_secondary": [], "topic": "bet-sizing", "difficulty": 4, "decision": { "pot_settled_bb": 5, "bets": {} } }