Tens or Better

{
  "game": {
    "name": "Tens or Better",
    "in_game_wordmark": "TENS OR BETTER",
    "wordmark_note": "The catalogue name and the in-game wordmark are the same phrase. Below the logo the game additionally prints a hand-count subtitle ('50 HAND', '1 HAND', etc.) that is NOT part of the game's name -- it is the variant label for whichever hand-count build was entered from the splash screen. This slot's demo_url/provider_ref (keyname=TensorBetter50Hand) is the 50-hand variant.",
    "provider": "Habanero",
    "provider_ref": "TensorBetter50Hand",
    "studio_evidence": "The only place the provider's name is printed anywhere in this build's own UI is the 'More Games From Habanero!' header inside the top-left grid-icon overlay (studio_evidence.png). The splash screen, base game and Help screens carry no visible 'Powered by Habanero' credit or wordmark of their own.",
    "game_kind": "table",
    "game_subtype": "poker",
    "game_subtype_note": "This is a draw-poker video poker machine, not a reel slot: there are no reels, paylines, ways, scatters or wilds. The capturer's manifest originally wrote game_subtype='video_poker', which does not exist in the consumer enum (GAME_SUBTYPES = blackjack|roulette|baccarat|poker|craps|sic_bo|other); the TL corrected it to 'poker' at GATE-1, following the Joker Poker (InOut, 3ac8aa06) precedent.",
    "hand_count_variants": {
      "observed_on_splash": [1, 5, 10, 50, 100],
      "this_record": 50,
      "how": "preview.png: pre-entry splash 'SELECT NUMBER OF HANDS TO PLAY' with five buttons in a 3-then-2 grid ('1 HAND' / '5 HAND' / '10 HAND' on the top row, '50 HAND' / '100 HAND' on the second row). The chosen count fixes the build for the whole session (it is not a per-round option once inside the game) and multiplies the total stake (see bet_model)."
    },
    "art_style": "Flat 2D vector casino-machine UI, not a painterly or 3D-rendered scene. Thick black outlines around every button, card and panel; solid saturated fills (royal blue felt background, marigold/orange bottom bar) with a light top-to-bottom gloss gradient on buttons only -- no photographic textures, no realistic lighting or shadow, no isometric or 3D perspective anywhere. The logo and card-back artwork use a hot-rod/starburst motif (five-point red stars, radial yellow-to-orange gradient, thick red-to-yellow outlined lettering with a drop shadow) reminiscent of a physical arcade/bar-top video poker cabinet rather than a modern mobile slot. Card faces themselves are plain, high-contrast, flat playing-card iconography (white card, black or red pips, serif rank letters) with no illustration or ornamentation. Overall mood is bright, static, daytime/interior-cabinet lighting (no night/neon palette), front-on flat camera angle (no tilt, no depth-of-field)."
  },

  "color_palette": {
    "background_playfield": "solid royal blue (mid-dark blue, no gradient) behind the 50-hand grid and the 1-hand card row",
    "background_splash": "solid black behind the hand-count picker",
    "bottom_bar": "warm marigold/orange-yellow flat fill, framed top and bottom by a thin darker orange rule",
    "value_readouts_balance_bet_coins_win": "black rounded rectangles with a thin light-grey border, white bold numerals, prefixed with a stylised '$'-like currency glyph",
    "primary_action_buttons_deal_draw_auto_play_bet_one_bet_max": "medium blue fill with a lighter blue top gloss, white bold caps text, dark navy border",
    "active_toggle_auto_hold_on": "same blue button body as the others, but the label text is rendered in yellow AND the button gets an added orange/red outer outline -- an on-state is therefore coded by BOTH colour and an extra border shape, not colour alone",
    "disabled_buttons_bet_one_bet_max_after_deal": "desaturated olive-grey/sage green fill with tan/khaki-coloured text, visibly duller and flatter than the glossy blue of enabled buttons -- distinguishable even in greyscale by the loss of the gloss highlight, not colour alone",
    "top_corner_icon_buttons_more_games_options_sound": "dark charcoal/near-black rounded squares with a gold/amber metallic frame",
    "card_faces": "off-white/light-grey card body with a soft rounded corner; hearts and diamonds are printed in a saturated red, spades and clubs in solid black; rank letter/number top-left in the same colour as the suit pip",
    "card_backs_and_mini_logo_cards": "hot orange-to-yellow radial gradient field with a repeating dark-red five-point star motif and thin gold rays; the game's own wordmark card-backs additionally carry the red/yellow outlined 'TENS OR BETTER' lettering",
    "paytable_gradient_bars_high_tier_left_panel": "each hand rank has its own flat-to-gradient coloured pill, running crimson red (Royal Flush) through orange/amber (mid tiers) into yellow-green and teal/cyan (Straight, low tiers) -- colour encodes RANK ORDER, and the same order is redundantly given by the printed rank name and by row position (top = best hand)",
    "current_coin_column_indicator": "on the 1-hand full 5-column paytable, the column matching the currently selected COINS value is filled solid bright red, overriding the row's own gradient colour for that cell only",
    "winning_row_indicator": "the row LABEL text (e.g. 'FLUSH') for the hand rank that was just paid is rendered in a dulled, desaturated olive/khaki-yellow, visibly flatter than the bright saturated yellow used for every other (non-won) row label; TL correction: an earlier capturer note described this as the row turning 'white', which this documentation corrects after a direct zoomed re-check of win_screen_1hand.png -- the row never turns white, it turns a muted version of its normal yellow",
    "help_gear_overlay_and_autoplay_submenu": "near-black semi-opaque panel with a thin light-grey/white rounded border; body text white, headings white/bold, buttons rendered as light-grey pill shapes with black text"
  },

  "hand_paytable": {
    "unit": "coins (NOT currency) -- this is the native pay unit of the machine; a payout in currency = coins_won x current_coin_denomination, independent of how many coins were staked per hand",
    "source_frame": "paytable_p1.png (captured on the 1-hand build, which is the only build state that prints the full 9-row x 5-coin-column matrix at once; identical paytable confirmed in rules_p4.png text: 'The game and paytable are the same as the one-handed version')",
    "rows": [
      {"rank": "Royal Flush",     "coins_1": 500, "coins_2": 1000, "coins_3": 1500, "coins_4": 2000, "coins_5": 4000, "note": "the only rank with a non-linear jump at 5 coins (2000->4000, i.e. double the linear step) -- the classic 5-coin jackpot bonus of Jacks-or-Better-family video poker"},
      {"rank": "Straight Flush",  "coins_1": 50,  "coins_2": 100,  "coins_3": 150,  "coins_4": 200,  "coins_5": 250},
      {"rank": "4 of a Kind",     "coins_1": 25,  "coins_2": 50,   "coins_3": 75,   "coins_4": 100,  "coins_5": 125},
      {"rank": "Full House",      "coins_1": 5,   "coins_2": 10,   "coins_3": 15,   "coins_4": 20,   "coins_5": 25},
      {"rank": "Flush",           "coins_1": 4,   "coins_2": 8,    "coins_3": 12,   "coins_4": 16,   "coins_5": 20},
      {"rank": "Straight",        "coins_1": 4,   "coins_2": 8,    "coins_3": 12,   "coins_4": 16,   "coins_5": 20},
      {"rank": "3 of a Kind",     "coins_1": 3,   "coins_2": 6,    "coins_3": 9,    "coins_4": 12,   "coins_5": 15},
      {"rank": "2 Pair",          "coins_1": 2,   "coins_2": 4,    "coins_3": 6,    "coins_4": 8,    "coins_5": 10},
      {"rank": "Tens or Better",  "coins_1": 1,   "coins_2": 2,    "coins_3": 3,    "coins_4": 4,    "coins_5": 5, "note": "the game's namesake rule: a pair pays ONLY if its rank is Ten or higher (10, J, Q, K, A); pairs of 2-9 do not pay at all, confirmed by rules_p3/p4.png's 'Overall Hand Ranks' list, which defines 'Tens Or Better' as 'Two cards paired that are 10 or ranked better'"}
    ],
    "hand_rank_definitions_verbatim_from_help": {
      "card_ranks": "A, K, Q, J, 10, 9, 8, 7, 6, 5, 4, 3, and 2 (A can be high or low for straights)",
      "royal_flush": "A, K, Q, J and 10 of the same suit",
      "straight_flush": "Five cards of same suit in sequence - A ranked lowest or highest",
      "four_of_a_kind": "Four cards of the same rank, and 1 other card of different rank",
      "full_house": "Three of a Kind and a Pair",
      "flush": "Five cards all the same suit",
      "straight": "Five cards in sequence - A ranked lowest or highest",
      "three_of_a_kind": "Three cards of same rank, and 2 other mixed ranked cards",
      "two_pair": "Two different Pairs, with one other card of different rank as single kicker",
      "tens_or_better": "Two cards paired that are 10 or ranked better, and 3 other mixed ranked cards",
      "tie_break_rule": "Only the highest rank pays for any hand (a hand that qualifies for several ranks at once, e.g. a flush that is also a straight, is paid once at the higher rank only)",
      "deck_rule": "A hand is played with 1 deck - a card with the same rank and suit cannot appear more than once"
    },
    "50_hand_build_display_quirk": "In the 50-hand build (base_game.png, win_screen.png, max_bet_evidence.png) the on-screen paytable does NOT show the full 5-column matrix; it splits into two flanking panels (6 high-value ranks on the left, 3 low-value ranks on the right) and shows a SINGLE value per rank -- the value for whichever COINS count (1-5) is currently selected. Confirmed dynamic: at COINS=1 the panel prints Royal Flush 500 (base_game.png); at COINS=5 it prints Royal Flush 4000 (max_bet_evidence.png), matching the 1-hand build's own 1-coin and 5-coin columns exactly."
  },

  "bet_model": {
    "formula": "total_bet (currency) = coin_denomination x coins_per_hand (1-5, stepped by BET ONE or the COINS +/- buttons) x number_of_hands (fixed per build: 1, 5, 10, 50 or 100, chosen once on the splash screen and not changeable afterwards)",
    "verified_examples": [
      {"denom": 0.02, "coins": 1, "hands": 50, "bet": 1.00, "frame": "base_game.png"},
      {"denom": 0.5,  "coins": 5, "hands": 50, "bet": 125.00, "frame": "max_bet_evidence.png"},
      {"denom": 0.2,  "coins": 1, "hands": 1,  "bet": 0.20, "frame": "win_screen_1hand.png"}
    ],
    "denomination_ladder": "floor 0.02, ceiling 0.5 observed via the +/- stepper next to the coin-chip icon (plateaus after repeated + clicks; ladder rungs not individually enumerated beyond floor/ceiling since Help does not print the full list and help_facts.py refused on this build's HTML template)",
    "bet_one_button": "cycles COINS between 1 and 5 (per Help step 2: 'Click Bet One to cycle between 1 and 5 coins bet per deal')",
    "bet_max_button": "sets COINS to 5 AND immediately deals (per Help step 3), i.e. it is a combined 'max stake + deal' action, not a stake-only control",
    "payout_formula": "payout (currency, per hand) = coin_value_from_paytable(hand_rank, coins_per_hand) x coin_denomination. The number of hands in play affects total exposure ONLY through the stake (total_bet) and through how many hands can each independently hit a rank -- it never multiplies a single hand's own coin payout.",
    "worked_reconciliation": "win_screen.png: 50-hand round, denom 0.02, 1 coin/hand -> total_bet $1.00. Of the 50 hands, 7 resolved to 3 OF A KIND (3 coins each) and 8 resolved to 2 PAIR (2 coins each); the other 35 hands paid nothing. Coins won = 7x3 + 8x2 = 21 + 16 = 37 coins. Currency won = 37 x $0.02 = $0.74, which is exactly the WIN field printed on that frame ($0.74). This is the single most load-bearing number in this slot's documentation: it independently proves both the coins-then-denomination payout formula and that the right-hand mini paytable panel doubles as a per-combination HIT COUNTER once a round resolves (the small numbers '7' and '8' printed to the left of the '3 OF A KIND' / '2 PAIR' rows on that frame are hand counts, not coin values -- they only appear after a round completes, and are absent at rest, e.g. max_bet_evidence.png)."
  },

  "bet_limits_verified": {
    "min_bet": 1.00,
    "max_bet": 125.00,
    "currency_symbol_in_game": "a stylised '$'-like glyph (rendered as a superscript currency mark before the numeral in all balance/bet/win fields); this is a demo-balance token, not confirmed as a real-money currency code -- do not read it as a currency identifier",
    "how": "min_bet = 1 coin x $0.02 (denomination floor) x 50 hands = $1.00, read at rest on base_game.png. max_bet = 5 coins x $0.50 (denomination ceiling, reached after repeated + clicks on the denomination stepper, then plateaus) x 50 hands = $125.00, read on max_bet_evidence.png. This is the FULL ceiling for the 50-hand variant specifically; the 1-hand variant of the same underlying paytable/assembly runs $0.02-$2.50 (tested only to confirm the paytable and Gamble text, not part of this record's own bet range)."
  },

  "max_multiplier": {
    "value_x_of_total_bet": 800,
    "derivation": "A literal SINGLE royal flush on the 50-hand build is only 4000 coins against a 250-coin total bet (5 coins x 50 hands) = 16x -- not the headline number. The 800x ceiling requires the multi-hand inheritance mechanic: Help (rules_p4.png) states 'You are dealt the first hand, which determines your overall cards for the draw that follows for the rest of the hands', so a dealt/held royal flush is drawn into ALL 50 hands simultaneously: 50 x 4000 = 200,000 coins against the 250-coin total stake = 800x of the total bet. The provider's own game page independently prints 'Max Multiplier: x800', agreeing with this derivation.",
    "frame": "max_bet_evidence.png (paytable values) + rules_p4.png (multi-hand inheritance rule)"
  },

  "multi_hand_mechanic": {
    "name": "Multi-Handed Play (5, 10, 50, 100 Hands)",
    "verbatim_help_text": "The game and paytable are the same as the one-handed version. You are dealt the first hand, which determines your overall cards for the draw that follows for the rest of the hands. The draw of cards dealt repeats multiple times for the multiple hands. When playing multiple hands, winning hands are highlighted in the game.",
    "plain_explanation": "Only ONE hand is actually dealt and played by the player (the 5 centre cards, with HOLD choices). Every one of the other 49 (or 4, 9, 99) hand-grids on screen is a face-down placeholder that, on Draw, independently redraws its own non-held card positions from a fresh shuffle while re-using the player's own held cards -- so 50 different final hands emerge from one starting hand and one hold decision, each scored and paid against the paytable independently.",
    "observed_on": ["base_game.png (50 face-down hand grids at rest)", "base_game_deal.png (grids still face-down right after Deal, before Draw)", "win_screen.png (resolved 50-hand grid, each cell showing its own final 5 cards and, where it paid, a coloured banner naming its hand rank)"]
  },

  "rtp": {
    "value_percent": 99.14,
    "condition": "with optimal strategy",
    "source": "rules_p4.png, section 'Return to Player': \"With optimal strategy, the theoretical RTP of this game is 99.14%.\" This is the assembly's own printed text (read directly off the Help screen), not derived from help_facts.py -- that script refused on this build because Video Poker's Help screen uses a different HTML template than the reel-slot Help template it was written against (fewer than its required 4 anchor lines were found).",
    "cross_check": "The provider's game-catalogue page lists a rounded 'Available RTPs: 99%', which agrees with 99.14% after rounding. Per the Habanero playbook's general caution (P-H03), that field is usually an operator-configuration menu rather than this build's own RTP and should not normally be trusted -- it is only noted here because it happens to agree, not used as the source of truth."
  },

  "build_stamps": {
    "file_date": "2026/09/10 06:11:10 (UTC)",
    "client_version": "1.1.23503.626",
    "server_version": "5.1.10548.421",
    "rng_version": "5.1.4478.308",
    "frame": "rules_p6.png",
    "note": "Per the Habanero playbook (P-H04), these are build/deployment stamps, NOT a release date, and are recorded here purely as provenance/versioning information."
  },

  "gamble_feature_discrepancy": {
    "claim_in_help_text": "Help screen step 6 (rules_p2.png): \"You may gamble a winning game with a 'Double or Nothing' round. Press Gamble. In the Double or Nothing game - select a card that is higher than the dealer's card to win.\" Step 7 adds: \"Gamble is disabled for the Royal Flush jackpot win.\"",
    "observed_in_build": "No Gamble control exists anywhere after a winning round, in either the 50-hand or the 1-hand build.",
    "evidence": [
      "pixels: the full button row after a resolved win shows only AUTO PLAY / AUTO HOLD / BET ONE / BET MAX / DEAL -- checked on two separate wins (the 50-hand $0.74 win on win_screen.png and the 1-hand $0.80 flush win on win_screen_1hand.png)",
      "DOM: document.body.innerText read immediately after the 1-hand flush win contains no 'GAMBLE' string anywhere on the page"
    ],
    "conclusion": "This is a documented text/build mismatch, not a missed feature: the Help text describes a Gamble/Double-or-Nothing round that this specific build does not expose. Recorded as a not_captured claim with a stated reason (feature not present in build) rather than as a feature of the game."
  },

  "screens": [
    {
      "file": "preview.png",
      "type": "preview",
      "description": "Entry/splash screen, solid black background. Centred game logo: 'TENS OR' stacked above 'BETTER' in bold red-outlined yellow-orange lettering with a drop shadow, flanked by two five-point star outlines and a soft radial glow. Below it, white bold caps heading 'SELECT NUMBER OF HANDS TO PLAY', then five orange-to-gold gradient rounded-rectangle buttons with black bold text: a top row of three ('1 HAND', '5 HAND', '10 HAND') and a second row of two, horizontally centred ('50 HAND', '100 HAND'). This IS the entry screen -- there is no separate promotional/max-win splash before it."
    },
    {
      "file": "base_game.png",
      "type": "base_game",
      "description": "50-hand table at rest, denomination 0.02, 1 coin, BET $1.00, BALANCE $2,000.00. Playfield is a 6-column x 7-row grid (42 visible cells, more scroll off-screen implicitly as part of the 50) of face-down 'card' tiles, each tiled with the orange/star card-back art and, on cells near the grid edges, a fragment of the red/yellow 'TENS OR BETTER' wordmark bleeding through the tile artwork. Below the grid: a left paytable panel (Royal Flush 500 down to Straight 4, coloured red-to-yellow-green gradient pills) and a right paytable panel (3 of a Kind 3, 2 Pair 2, Tens or Better 1, teal gradient pills), flanking a small centred mini-logo reading 'TENS OR BETTER / 50 HAND' and a row of 5 face-down mini 'TENS OR BETTER'-branded cards (decorative, representing the un-dealt hand). Bottom bar: BALANCE box, denomination chip selector with +/- steppers, then COINS/BET/WIN readouts, then AUTO PLAY / AUTO HOLD (yellow text + orange outline, i.e. currently ON) / BET ONE / BET MAX / DEAL buttons. Top-left: gold-framed grid icon (More Games). Top-right: gold-framed list icon (Options) and speaker icon (sound)."
    },
    {
      "file": "base_game_deal.png",
      "type": "base_game",
      "description": "Same 50-hand layout immediately after pressing DEAL. The 50 hand-grids are still all face-down (the multi-hand draw has not resolved yet). The centre 5-card row now shows the player's actual dealt hand: 6 of Hearts, 6 of Spades, Jack of Hearts, 9 of Clubs, Ace of Spades. Auto Hold has pre-selected the pair of 6s: both 6 cards show a yellow 'HOLD' banner across their lower third; the Jack, 9 and Ace are not held. BALANCE has already dropped to $1,999.00 (bet deducted up front). A round id '#5340600000094015' appears top-left, absent at rest. BET ONE and BET MAX are now visually disabled (dulled olive-green fill, tan text); the DEAL button has become DRAW (same active blue styling)."
    },
    {
      "file": "win_screen.png",
      "type": "win_screen",
      "description": "Resolved 50-hand round. Every one of the 50 grids now shows its own final 5 cards; cells that paid display a small coloured banner across the middle naming the hand ('3 OF A KIND' in a green banner, '2 PAIR' in a teal banner) -- 7 grids show a 3-of-a-kind banner and 8 show a 2-pair banner, all built from the same held pair of 6s carried into every hand. WIN $0.74 is printed bottom-right, BALANCE has updated from $1,999.00 to $1,999.74. The right-hand paytable panel now additionally shows small hit-count numerals (7, 8) to the left of the '3 OF A KIND' / '2 PAIR' rows -- present only because this round paid; compare to base_game.png/max_bet_evidence.png where these numerals are absent."
    },
    {
      "file": "win_screen_1hand.png",
      "type": "win_screen",
      "description": "1-hand build result, denom 0.20, 1 coin, BET $0.20. Dealt hand was 6C 6D 3D 4D JD; the four diamonds (6,3,4,J) were held (yellow 'HOLD' banners) and the drawn replacement was Queen of Diamonds, completing a FLUSH (all five cards diamonds). WIN $0.80 (legible via OCR on the bottom bar). BALANCE reads $2,000.60 by eye (tesseract garbles this stylised font on the full frame; confirmed by zoom). The full 9-row paytable is visible top-centre since the 1-hand build has room to show it permanently; the 'FLUSH' row label is rendered in a dulled/desaturated olive-yellow versus the bright saturated yellow of every other row label -- this is the build's win indicator for which rank just paid. (Documentation correction: an earlier working note described this row as turning 'white'; a direct zoomed re-check shows it never turns white, only a muted version of its normal yellow -- see color_palette.winning_row_indicator.)"
    },
    {
      "file": "paytable_p1.png",
      "type": "paytable",
      "description": "1-hand build at rest, no hand dealt yet, all 9 ranks x 5 coin-columns visible simultaneously (this build's paytable panel has room for the full matrix; the 50-hand build does not, see hand_paytable.50_hand_build_display_quirk). Gradient bars run red (Royal Flush) through orange/amber/yellow to green and teal (Tens or Better), each row a horizontal pill shape. Five face-down mini 'TENS OR BETTER'-branded cards sit below the table as decoration; no cards are dealt yet (WIN $0.00)."
    },
    {
      "file": "rules_p1.png",
      "type": "rules",
      "page": 1,
      "description": "Help overlay, top of a single continuous scrolling document (near-black semi-transparent panel over the dimmed playfield). Heading 'How To Play', a 10-step numbered list covering coin denomination selection, Bet One cycling 1-5 coins, Bet Max, Deal, clicking cards to hold, Auto Hold defaulting to optimal strategy, Draw, and the winning-payline flash. Back-arrow control at the bottom returns to the game (does not paginate to a next page -- this Help document scrolls, it does not page)."
    },
    {
      "file": "rules_p2.png",
      "type": "rules",
      "page": 2,
      "description": "Scrolled further: 'Game Options' 1-7 covering Auto Play hand counts, Stop Auto, Auto Hold on/off and manual override, 'Auto Hold appears in yellow font when active', the Gamble/Double-or-Nothing description and its Royal Flush exception, followed by the start of 'Card Ranks' (A high to 2 low) and the beginning of 'Overall Hand Ranks'."
    },
    {
      "file": "rules_p3.png",
      "type": "rules",
      "page": 3,
      "description": "Scrolled further: the complete 'Card Ranks' line and the full 'Overall Hand Ranks' definitions list (Royal Flush down to Tens Or Better, each with its exact qualifying condition), plus 'Only the highest rank pays for any hand' and 'A hand is played with 1 deck - a card with the same rank and suit cannot appear more than once'."
    },
    {
      "file": "rules_p4.png",
      "type": "rules",
      "page": 4,
      "description": "Scrolled further: 'Payouts' ('The paytable displayed gives the payout(s) for different hands in coins'), the full 'Playing Multi-Handed Versions (5, 10, 50, 100 Hands)' text (first-hand-determines-the-draw rule), and 'Return to Player' with the 99.14% RTP line."
    },
    {
      "file": "rules_p5.png",
      "type": "rules",
      "page": 5,
      "description": "Scrolled further: repeats the tail of 'Playing Multi-Handed Versions' and 'Return to Player' visible at the top edge, then the full 'General Rules' ('Malfunction voids all plays and pays.') and the full 'Disconnection Policy' (three points: server-side processing on disconnect, resumable game state, 15-day forfeiture window). NOTE: this page was assembled by the TL at GATE-1 from a promoted scroll capture because the capturer's own p4/p5 pair had clipped this single policy paragraph across the seam between two screenshots, leaving it incomplete on both (thinning/seam repair is TL work per defect D-50, not a capturer error)."
    },
    {
      "file": "rules_p6.png",
      "type": "rules",
      "page": 6,
      "description": "Bottom of the Help document: end of 'Disconnection Policy', then the build stamps block -- 'File Date 2026/09/10 06:11:10 (UTC)', 'Client Version 1.1.23503.626', 'Server Version 5.1.10548.421', 'RNG Version 5.1.4478.308'. No further content below this; confirmed as the true bottom of the scroll."
    },
    {
      "file": "max_bet_evidence.png",
      "type": "other",
      "description": "50-hand build with denomination pushed to its ceiling (0.5) and COINS pushed to 5 via repeated +-clicks: BET $125.00, BALANCE untouched at $2,000.00 (no hand dealt yet). Paytable panels now show the 5-coin column values (Royal Flush 4000, Straight Flush 250, 4 of a Kind 125, Full House 25, Flush 20, Straight 20 / 3 of a Kind 15, 2 Pair 10, Tens or Better 5), confirming the panel re-renders per the selected COINS value rather than always showing the 1-coin column."
    },
    {
      "file": "autoplay.png",
      "type": "autoplay",
      "description": "Auto Play submenu overlay (opened via the AUTO PLAY button, not started): a near-black rounded panel with three light-grey pill buttons stacked vertically -- 'AUTO PLAY 50', 'AUTO PLAY 100', and a plain 'AUTO PLAY' (unlimited, stoppable at any time via 'Stop Auto' per the Help text), plus a circular back-arrow to close the panel. Captured over the 1-hand build's win_screen_1hand.png state (FLUSH win still visible dimmed behind the overlay)."
    },
    {
      "file": "settings.png",
      "type": "settings",
      "description": "Options overlay opened via the top-right list icon: heading 'OPTIONS', a 'FULLSCREEN' row with a toggle switch (shown OFF -- white circular knob on the left of a dark track), a horizontal rule, a large light-grey 'HELP' pill button, and the circular back-arrow. No separate sound-effects/music toggles live here -- the speaker icon on the main bar is a standalone mute control outside this panel."
    },
    {
      "file": "studio_evidence.png",
      "type": "other",
      "description": "'More Games' overlay (opened via the top-left grid icon): header 'More Games From Habanero!' with a close (X) button, a 'Featured' strip of four game tiles (Battleship Strike, Battle The Beast, Rhythm Nights, Galaxy Babies), a genre dropdown defaulted to 'Slots' and a search box, then a further grid of game tiles (Rhythm Nights, Mystic Shaman, Safari Rumble, Gladiator Royal). This is the only place in the entire build where the studio name 'Habanero' is printed."
    }
  ],

  "feature_coverage": [
    {"feature": "base game: single-hand deal / hold / draw", "gameplay": "win_screen_1hand.png", "win": "win_screen_1hand.png"},
    {"feature": "multi-hand deal (50 hands) / auto-hold / draw", "gameplay": "base_game_deal.png", "win": "win_screen.png"},
    {"feature": "gamble / double-or-nothing", "status": "described in Help text but NOT present in this build -- see gamble_feature_discrepancy", "gameplay": null, "win": null}
  ],

  "not_captured": [
    {
      "claim": "The 'Gamble' (Double or Nothing) feature described in the game's own Help text does not appear as an actionable control after a winning round, in either the 50-hand or the 1-hand build.",
      "methods": [
        "pixels: zoomed crop of the full button row after a resolved win shows only AUTO PLAY / AUTO HOLD / BET ONE / BET MAX / DEAL on two separate wins",
        "DOM: document.body.innerText immediately after the 1-hand flush win contains no 'GAMBLE' string"
      ]
    },
    {
      "claim": "Whether the 'AUTO HOLD' button's label reverts from yellow-with-outline to a plain colour when toggled OFF is unconfirmed -- no frame with Auto Hold switched off was captured.",
      "methods": ["not tested"]
    }
  ],

  "anomalies": [
    "Contrary to the general Habanero playbook note ('MODALS FUN PLAY MODE NO LONGER EXIST', current as of reel-slot titles), this Video Poker title (served through the PrePlayApp path with a hand-count selector) DOES show a 'FUN PLAY MODE / You are playing for fun.' modal on every fresh entry, in both the 50-hand and 1-hand sessions. Flagged for the playbook maintainer: the 'no modal' finding may be scoped to reel-slots specifically rather than the whole provider.",
    "help_facts.py (the provider's standard Help-screen scraper) refuses on this assembly because Video Poker's Help/How-To-Play screen uses a different HTML template than the reel-slot Help template the script expects. All numbers in this document (RTP, paytable, bet limits, build stamps) were instead read directly from the same on-screen Help panel by eye/zoom.",
    "The visible coin-denomination chip (e.g. '0.02', '0.05', '0.5') is NOT itself the per-coin cost of a hand: total_bet = denomination x coins_per_hand x number_of_hands. At denom 0.02, 1 coin, 50 hands the displayed '0.02' actually corresponds to a $1.00 total bet -- readers should not assume the chip value is the stake."
  ],

  "accessibility_observations": {
    "color_redundancy_good_practice": "The one state that matters most for correct play -- whether Auto Hold is currently active -- is coded THREE ways at once: label font colour (yellow vs white), an added outline around the button (present vs absent), and the Help text itself explicitly warns 'Auto Hold appears in yellow font when active'. A colourblind player who cannot distinguish yellow from white still has the outline as a second channel.",
    "color_redundancy_good_practice_2": "Disabled buttons (BET ONE / BET MAX once a hand is dealt) change BOTH hue (blue to olive-green) and the loss of the glossy highlight/border contrast, not colour alone -- distinguishable in greyscale.",
    "color_only_concern": "The paytable's rank-order colour gradient (red=best down to teal=worst) is decorative reinforcement only; rank is also given redundantly by row position (fixed, best-to-worst top-to-bottom) and by the printed rank name itself, so no information is colour-only there.",
    "color_only_concern_2": "The 'which rank just won' indicator (dimmed/desaturated row label after a win) IS a colour-only signal in isolation, but it is redundant with a much stronger cue: the specific cards in the resolved hand and their suit/rank pips are plainly legible and unambiguously identify the hand rank without needing to read the paytable highlight at all.",
    "text_legibility": "All static rule text (Help screens, paytable numerals, button labels) is high-contrast bold sans/serif on a plain background and reads cleanly. The BALANCE/BET/WIN numerals use a stylised, slightly compressed font that is legible to a human eye but defeats plain tesseract OCR on the full frame (values were confirmed by zoomed crops instead, see anomalies and bet_model).",
    "motor_accessibility": "Every action (Bet One, Bet Max, Deal/Draw, hold-card selection, Auto Play counts) is a single discrete click/tap on a clearly bounded button or card; no drag, no timed reflex input, no simultaneous multi-key combination is required anywhere in the captured flow."
  },

  "unreadable_or_uncertain": [
    "The full coin-denomination ladder between its observed floor (0.02) and ceiling (0.5) was not individually enumerated -- Help does not print the list and help_facts.py refused on this build's HTML template.",
    "Whether the AUTO HOLD button reverts to a plain (non-yellow, non-outlined) style when switched off was not observed (no frame captured with it toggled off).",
    "No networked/ticking jackpot of any kind (the MINOR/GRAND or Jackpot Race(TM) mechanics documented elsewhere in the Habanero playbook for reel-slot titles) was seen on any of this game's captured frames; this title appears to have no jackpot feature at all, but that is an absence-of-evidence observation on this demo build, not a claim verified against provider copy."
  ]
}