Sweet Keno

{
  "game": {
    "name": "Sweet Keno",
    "provider": "InOut Games",
    "wordmark_discrepancy": "The in-game logo (top-left of every screen) reads 'KENO' only -- a rounded, chunky, candy-styled wordmark where the letters K-E-N are solid white with a thin dark-red outline and the final 'O' is replaced by a red-and-white spiral lollipop icon (matching the 'hit' candy used on the number grid). This in-game wordmark does NOT include the word 'Sweet'. The catalog/marketing name used for this documentation is 'Sweet Keno'. Both names are reported here without asserting which is 'more correct' -- the discrepancy is simply a fact of the build.",
    "game_kind": "instant_win numbers-keno game (NOT a reel slot: no reels, no paylines, no slot-style symbol paytable). Round resolves instantly and fully on a single BET click; there is no multi-step reveal, no cash-out decision, and no shared/timer-driven round.",
    "art_style": "Flat-shaded 2D vector illustration with a soft, rounded, toy-like 3D bevel on every UI element (buttons, number tiles, panels look like glossy candy/plastic, with visible highlight-and-shadow bevels rather than flat material-design flatness). No painterly gradients on the background scenery -- it is built from simple rounded blob shapes (hills, mushroom-topped trees, swirling ice-cream cones, gumdrops, jellybeans) in saturated pastel colors, lit as if by a bright, shadowless midday sun (no directional shadows, no night/neon lighting anywhere). The overall mood is cheerful, sugary, child-friendly casual-mobile-game -- a candy-land theme (frosting-pink scalloped border dripping with icing and sprinkles around the main play panel, ice-cream-cone hill, striped candy-cane accents). Viewpoint is a static flat frontal 2D layout (no isometric or 3D-perspective camera; the only '3D' quality is the bevel/glossy shading applied to 2D shapes, sometimes called 'candy-crush style' UI). This is a bright-daylight, low-drama, low-tension visual register -- the opposite of a moody/cinematic dark casino aesthetic."
  },
  "documentation_purpose": "Exhaustive accessibility documentation (blind/low-vision screen-reader users, deaf/hard-of-hearing, color-blind users of all types, motor-impaired users) built from 59 desktop screenshots (1280x800) + OCR ground truth + zoomed crops + pixel-level grid-cell verification. Numbers are taken verbatim from on-screen text. Anything that stayed unclear, contradictory, or could not be verified is listed in 'unreadable_or_uncertain' and is NOT invented.",
  "language": "English",
  "color_palette": {
    "background_scenery": "Bright pastel candy-land: sky-blue-to-turquoise gradient sky, lime-green rolling hills, a large pink-and-cream soft-serve ice-cream cone on the left, pastel pink/teal/yellow rock-candy and gumdrop shapes scattered on both sides, a red-and-white spiral lollipop mushroom-lamp on the left hill.",
    "main_panel_frame": "Pale mint/aqua rounded rectangle panel with a thick pink 'frosting' border dripping down the top edge, decorated with small multicolored sprinkles (red, green, blue, yellow, purple dots/rods) and red-and-white peppermint-swirl accents at the top corners.",
    "number_grid_unselected_cell": "Pale cream / light butter-yellow rounded-square tile with a soft brown-tan embossed bevel edge; number printed in warm brown.",
    "number_grid_selected_cell": "Burnt-orange / copper rounded-square tile (chosen-but-not-yet-drawn or chosen-no-match state); number printed in dark brown.",
    "number_grid_drawn_not_chosen": "Plain lavender/purple glossy spiral-candy circular icon (no number visible) sitting on top of a cream tile -- represents a number the game drew that the player did NOT pick.",
    "number_grid_hit_cell": "Bright yellow/gold-bordered square badge containing a red-and-white spiral lollipop icon -- represents a number that was BOTH picked by the player AND drawn by the game (a match/hit). This is the single most important state for a color-blind or screen-reader user: it is distinguished from 'drawn, not chosen' by SHAPE (square gold badge vs. plain circle) and by an added red/white candy color, not by hue alone.",
    "tier_bar_tabs": "Pale cream/butter-yellow rounded pill tabs (inactive) vs. bright golden-yellow pill tabs with a darker gold outline (the tier(s) reached/highlighted during or after a resolved round); tab text is dark brown, values in bold dark brown.",
    "risk_selector_bar": "Dark olive-brown/charcoal-brown bar; the ACTIVE risk level (EASY / MEDIUM / HIGH) is a bright golden-yellow rounded tab with black bold text; inactive levels are plain white text on the dark brown bar.",
    "bet_panel": "Pale cream rounded bar for the MIN/bet-amount/MAX stepper; smaller cream pill buttons below it for preset stake chips (e.g. 2$/3$/8$/20$).",
    "bet_button": "Bright green rounded rectangle with a thick darker-green bottom edge (glossy 3D button look) and bold white 'BET' text with a dark outline.",
    "top_bar": "Solid teal/turquoise background band. 'How to play?' is a dark teal pill button with a white circular 'i' info icon and white text. The balance readout is a dark navy-teal pill with white text and a white '$' coin icon. Fullscreen and hamburger-menu icons are plain white glyphs on the teal band.",
    "burger_menu_panel": "Warm brown/terracotta rounded panel (semi-opaque) sliding in from the top-right, listing Sound / Music toggles (white circular toggle switches, off = grey dot at left, would be green/right when on) and Provably fair settings / Game rules / My bet history menu rows in white text with matching white line icons; 'Powered by iNOUT' logo (orange 'iN' + white 'OUT') sits at the bottom of the panel.",
    "modal_panels": "'How to play?', 'Game rules', 'Provably fair settings', 'My bet history' all use the same warm terracotta/copper rounded modal over a darkened/blurred game background, white 'X' close icon top-right, white heading text, lighter tan sub-text.",
    "colorblind_note": "Every state that matters for outcome (chosen vs. unchosen vs. drawn vs. hit) is redundantly coded by SHAPE and by the presence/absence of printed digits, not by hue alone: unchosen/chosen are both plain rounded tiles that differ mainly in printed-number tone plus a border-color shift; drawn-not-hit is a circular candy icon (no digit); hit is a square gold-bordered badge with a red/white candy (visually the most distinct of all four states). The active risk level and the active/resolved payout tier are both marked by a filled pill shape in addition to the gold color, so a color-blind user can find them by shape/fill, not color alone."
  },
  "top_bar_and_global_ui": {
    "logo": "Top-left: 'KENO' wordmark (see game.wordmark_discrepancy).",
    "how_to_play_button": "Top bar, left-of-center: dark teal pill button, white circular '(i)' icon + white text 'How to play?'. Opens the how_to_play modal (see screens.rules_how_to_play).",
    "balance_display": "Top bar, right-of-center: dark navy-teal pill showing the current play-money balance with a white '$' coin icon, e.g. '999 998.5' (six-figure demo credit visible across all frames, decreasing by the bet amount after each round and increasing on wins).",
    "fullscreen_button": "Small square icon (four corner arrows) to the right of the balance -- toggles fullscreen.",
    "hamburger_menu_button": "Three horizontal white lines in a square, far top-right -- opens the burger_menu_panel (avatar/name, Sound toggle, Music toggle, Provably fair settings, Game rules, My bet history, Powered by iNOUT footer)."
  },
  "base_game_screen": {
    "layout": "Centered pink-frosted panel containing, top to bottom: (1) the payout-tier carousel bar, (2) the 5-row x 8-column (40-cell) number grid, (3) the bet-amount stepper row with preset stake chips, (4) a green BET button, (5) the EASY/MEDIUM/HIGH risk-level selector bar, and (6) two small round action buttons (shuffle/Quick-Pick and broom/Clear).",
    "idle_state_prompt": "Before any numbers are chosen, the tier bar is replaced by a single banner reading 'Choose from 1 to 10 numbers to see the odds' -- the payout tiers only appear once at least 1 number is selected (base_game.png).",
    "number_grid": "40 numbered cells (1-40) arranged in 5 rows of 8, each a rounded square. Clicking a cell toggles it selected (max 10 selected at once, per the idle prompt). Selected/unselected states described in color_palette.",
    "tier_carousel": "A horizontal row of up to 6 pill tabs, each reading 'N hits' above its multiplier ('xM'), one tab per possible hit-count from 0 to the number of cells picked. Left/right chevron arrows ('<' / '>') appear at the ends of the bar once there are more than 6 tiers (i.e. once 6 or more numbers are picked), letting the player scroll the tier window left (toward 0 hits) or right (toward max hits). CONFIRMED CAPTURE DEFECT: this carousel does NOT reset its scroll position when the player changes the pick-count -- see 'data_integrity_findings' below for the concrete case this caused.",
    "bet_stepper": "A cream bar with a 'MIN' button (left), the numeric bet amount (center), and a 'MAX' button (right) -- clicking MIN/MAX jumps the stake straight to the game's floor/ceiling (verified, see bet_limits). Below it, four preset stake chip buttons showing fixed amounts with a '$' coin icon (seen as 2$ / 3$ / 8$ / 20$ in the captured frames; the demo's default stake is $3).",
    "action_icons": "Two round buttons bottom-right of the panel: a 'shuffle'/crossed-arrows icon (QUICK PICK -- randomly selects the currently-configured number of cells for the player, demonstrated in other_quickpick.png) and a paintbrush/broom icon (CLEAR -- deselects all picked cells; rendered visually dimmed/inactive when no cells are selected, e.g. on the very first idle frame).",
    "risk_selector": "Three-way tab bar: EASY / MEDIUM / HIGH. Switching it re-computes the ENTIRE payout table for every pick-count (see 'risk_levels' and 'payout_tables'). Only one tab is active at a time; the active tab is bright gold with black text.",
    "keyboard_or_alternate_input": "Not established from the available frames -- no on-screen text mentions a keyboard shortcut for BET/selecting cells; not asserted either way (see unreadable_or_uncertain)."
  },
  "risk_levels": {
    "how_to_play_wording": "From the in-game 'How to play?' modal, item 1 (verbatim OCR): 'Choose the risk level. At the easy level, the risk is lower, but if you guess all the cells — the payout rate is not as high. At the hard level, you take more risk, but the payout rates are significantly higher.'",
    "headline_comparison_of_the_top_tier": "Choosing 10 numbers and hitting all 10 pays a DIFFERENT top multiplier depending on risk level: EASY = x500, MEDIUM = x1 000, HIGH (confirmed max win of the whole game) = x10 000. That is a 20x difference in ceiling multiplier between EASY and HIGH for the identical 10-hit outcome -- risk level is not a cosmetic label, it is a full re-pricing of the paytable.",
    "general_pattern_low_picks": "At low pick-counts (1-4 numbers), HIGH risk pays noticeably more per hit than EASY/MEDIUM (e.g. picking 1 number and hitting it: EASY x1.52, MEDIUM x2.6, HIGH x3.92) but HIGH also zeroes out more of the low tiers (e.g. at 5-9 picks, HIGH pays x0 for 0 AND 1 hits, where EASY/MEDIUM still pay something for 0 hits at low pick counts, or MEDIUM pays x0 only from 1-2 hits at higher pick counts) -- i.e. HIGH risk trades away small/consolation wins for a much bigger top prize."
  },
  "payout_tables": {
    "note_on_verification": "Every table below was cross-checked by TWO independent readings: (1) the printed 'N hits / xM' text on the tier-carousel tabs (read directly from full-resolution screenshots, not thumbnails), and (2) a pixel-level automated count of how many of the 40 grid cells are rendered in the 'selected' (copper/orange) color on that exact frame, which must equal the pick-count N implied by the file name. Every table listed as 'complete' below passed both checks: the grid's selected-cell count matches N, and the tier list runs unbroken from 0 hits to N hits. One table (HIGH risk, 10 picks) FAILED the pixel check for its 0-4 hit tiers and is reported as partial -- see 'data_integrity_findings'.",
    "self_check_rule": "For N numbers picked, a complete table must have exactly N+1 tiers (0 hits through N hits). This was verified for every table below.",
    "easy": {
      "1_pick": {
        "tiers": {
          "0": "x0.8",
          "1": "x1.52"
        },
        "complete": true
      },
      "2_picks": {
        "tiers": {
          "0": "x0.4",
          "1": "x1.52",
          "2": "x3"
        },
        "complete": true
      },
      "3_picks": {
        "tiers": {
          "0": "x0",
          "1": "x1.3",
          "2": "x2.54",
          "3": "x5"
        },
        "complete": true
      },
      "4_picks": {
        "tiers": {
          "0": "x0",
          "1": "x1.1",
          "2": "x1.72",
          "3": "x2.55",
          "4": "x10"
        },
        "complete": true
      },
      "5_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.25",
          "2": "x1.36",
          "3": "x5",
          "4": "x10",
          "5": "x15"
        },
        "complete": true
      },
      "6_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.2",
          "2": "x1.36",
          "3": "x2.51",
          "4": "x5",
          "5": "x15",
          "6": "x25"
        },
        "complete": true
      },
      "7_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.1",
          "2": "x0.75",
          "3": "x2",
          "4": "x5.25",
          "5": "x15",
          "6": "x25",
          "7": "x50"
        },
        "complete": true
      },
      "8_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.1",
          "2": "x0.5",
          "3": "x1.62",
          "4": "x2.5",
          "5": "x15",
          "6": "x25",
          "7": "x50",
          "8": "x100"
        },
        "complete": true
      },
      "9_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.1",
          "2": "x0.25",
          "3": "x1.25",
          "4": "x2",
          "5": "x10",
          "6": "x22",
          "7": "x50",
          "8": "x100",
          "9": "x250"
        },
        "complete": true
      },
      "10_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.1",
          "2": "x0.25",
          "3": "x0.75",
          "4": "x1.75",
          "5": "x5",
          "6": "x25",
          "7": "x50",
          "8": "x100",
          "9": "x250",
          "10": "x500"
        },
        "complete": true,
        "note": "Top prize of EASY risk level: x500 at 10/10 hits."
      }
    },
    "medium": {
      "1_pick": {
        "tiers": {
          "0": "x0.44",
          "1": "x2.6"
        },
        "complete": true
      },
      "2_picks": {
        "tiers": {
          "0": "x0",
          "1": "x1.8",
          "2": "x5"
        },
        "complete": true
      },
      "3_picks": {
        "tiers": {
          "0": "x0",
          "1": "x1.2",
          "2": "x2.42",
          "3": "x10"
        },
        "complete": true
      },
      "4_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.6",
          "2": "x2.2",
          "3": "x5",
          "4": "x20"
        },
        "complete": true
      },
      "5_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.5",
          "2": "x1.4",
          "3": "x3.45",
          "4": "x10",
          "5": "x35"
        },
        "complete": true
      },
      "6_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.25",
          "2": "x1.4",
          "3": "x2.1",
          "4": "x5",
          "5": "x25",
          "6": "x50"
        },
        "complete": true
      },
      "7_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x0.75",
          "3": "x2.41",
          "4": "x5",
          "5": "x10",
          "6": "x25",
          "7": "x100"
        },
        "complete": true
      },
      "8_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x0.5",
          "3": "x1.2",
          "4": "x5",
          "5": "x10",
          "6": "x25",
          "7": "x50",
          "8": "x250"
        },
        "complete": true
      },
      "9_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x0.25",
          "3": "x1.1",
          "4": "x2.45",
          "5": "x10",
          "6": "x25",
          "7": "x50",
          "8": "x250",
          "9": "x500"
        },
        "complete": true
      },
      "10_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x0.2",
          "3": "x0.6",
          "4": "x2.45",
          "5": "x5.15",
          "6": "x20",
          "7": "x50",
          "8": "x100",
          "9": "x500",
          "10": "x1000"
        },
        "complete": true,
        "note": "Top prize of MEDIUM risk level: x1 000 at 10/10 hits."
      }
    },
    "high": {
      "1_pick": {
        "tiers": {
          "0": "x0",
          "1": "x3.92"
        },
        "complete": true
      },
      "2_picks": {
        "tiers": {
          "0": "x0",
          "1": "x1.53",
          "2": "x6.8"
        },
        "complete": true
      },
      "3_picks": {
        "tiers": {
          "0": "x0",
          "1": "x1",
          "2": "x2.62",
          "3": "x15"
        },
        "complete": true
      },
      "4_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0.5",
          "2": "x2",
          "3": "x6.32",
          "4": "x35"
        },
        "complete": true
      },
      "5_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x2",
          "3": "x3.3",
          "4": "x15",
          "5": "x50"
        },
        "complete": true
      },
      "6_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x1.1",
          "3": "x2.25",
          "4": "x10",
          "5": "x50",
          "6": "x100"
        },
        "complete": true
      },
      "7_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x0.2",
          "3": "x2",
          "4": "x5",
          "5": "x50",
          "6": "x100",
          "7": "x250"
        },
        "complete": true
      },
      "8_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x0",
          "3": "x1.5",
          "4": "x3.33",
          "5": "x20",
          "6": "x100",
          "7": "x250",
          "8": "x500"
        },
        "complete": true
      },
      "9_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x0",
          "3": "x0.98",
          "4": "x2.7",
          "5": "x10",
          "6": "x50",
          "7": "x100",
          "8": "x500",
          "9": "x1000"
        },
        "complete": true
      },
      "10_picks": {
        "tiers": {
          "0": "x0",
          "1": "x0",
          "2": "x0",
          "3": "x0",
          "4": "x1",
          "5": "x10",
          "6": "x50",
          "7": "x100",
          "8": "x500",
          "9": "x1000",
          "10": "x10000"
        },
        "complete": true,
        "note": "COMPLETE. Stitched from two carousel views captured by the TL after the original capture was found unreliable: paytable_high_p10_low_real.png (tiers 0-5) and paytable_high_p10_top.png (tiers 5-10). The two views overlap on '5 hits x10' and agree, which is what makes the stitch safe. Tier count 11 = picks + 1, so the self-check passes.",
        "source_frames": [
          "paytable_high_p10_low_real.png",
          "paytable_high_p10_top.png"
        ]
      }
    }
  },
  "data_integrity_findings": {
    "purpose": "This project's brief specifically warned that an incomplete-but-plausible-looking payout table for this exact tier (HIGH risk / 10 picks / max multiplier) could silently understate the game's true max win by 10x if the wrong frame were trusted. This section documents the verification work done to avoid that, using a reproducible pixel measurement rather than eyeballing the grid.",
    "method": "For every paytable_*.png frame, the 40 number-grid cells were sampled at their pixel centers and clustered into 'selected' (darker copper/orange) vs. 'unselected' (pale cream) by luma, using the largest brightness gap in the 40 samples as the split threshold. The resulting selected-cell COUNT was compared against the pick-count N implied by the frame's file name (e.g. 'paytable_high_p8.png' should show exactly 8 selected cells).",
    "result_all_clean": "EASY and MEDIUM risk levels: all 26 relevant frames (defaults + _low + medium's one _top) show a selected-cell count that exactly matches their file name's pick-count, for every N from 1 to 10. HIGH risk levels 1 through 9: all defaults/_low frames also match their file name's pick-count exactly.",
    "result_anomaly": "HIGH risk, pick-count 10: 'paytable_high_p10.png' (default view) and 'paytable_high_p10_low.png' both show only 8 selected cells, not 10 -- and their printed tier text (tiers 0 through 5/8) is IDENTICAL, digit for digit, to 'paytable_high_p8.png' and 'paytable_high_p8_low.png'. In other words, these two 'p10' frames are a duplicate/stuck capture of the 8-picks grid state, not a genuine 10-picks round. This matches (and sharpens) an anomaly the capturer had already flagged in the manifest as 'p8 and p10 look pixel-identical' -- the capturer's own manual eye-check of that anomaly ('p8 shows 8 highlighted cells 1-8, p10 shows 10 highlighted cells 1-10') was itself re-checked here with the pixel method and found to be WRONG on both counts: neither frame's selected set actually starts at cell 1 (both start at cell 2, because this game's Quick-Pick/manual selection does not have to be a contiguous 1..N block), and 'p10' does not actually contain 10 selected cells at all -- it contains 8, same as p8. The TL therefore treated 'paytable_high_p10.png' and 'paytable_high_p10_low.png' as NOT valid evidence for the 10-picks table and did not use their printed values for the 0-4 hit tiers of 10 picks (see payout_tables.high.10_picks).",
    "recovery": "The TL personally re-opened the live game, selected exactly 10 numbers on HIGH risk, and captured a fresh frame ('paytable_high_p10_top.png', already present in the manifest as a promoted frame) which the pixel check confirms shows exactly 10 selected cells (1 through 10) and whose tier window covers hits 5 through 10 (x10, x50, x100, x500, x1 000, x10 000). This is corroborated by a second, independently-drawn genuine 10-picks frame -- 'other_quickpick.png', captured via the Quick-Pick shuffle button with a different random set of 10 cells (3, 6, 16, 17, 18, 23, 26, 31, 37, 39) -- which shows the SAME tier window and SAME values (x10/x50/x100/x500/x1 000/x10 000), giving two independent confirmations of the game's true maximum win tier. The 0-4 hit tiers for 10 picks on HIGH risk were never captured with a verified-genuine 10-picks grid state and are honestly left blank rather than filled in from the contaminated 8-picks frames.",
    "conclusion": "The game's overall maximum payout is x10 000 of the total stake (HIGH risk, 10 numbers picked, all 10 hit) -- confirmed by two independently-drawn genuine 10-pick frames, pixel-verified. This is 10x higher than what a naive reading of the delivered 'p10' default/low frames would have produced (those, being an 8-picks duplicate, top out at x500).",
    "resolution_by_tl": "RESOLVED. The TL confirmed the finding by comparing paytable_high_p10.png with paytable_high_p8.png (identical printed tiers) and with a genuine 10-picks frame (tiers differ: 5 hits x10 vs x20). Both mislabelled frames were removed from the manifest so they cannot reach the rig as page=10, and the missing low tiers were re-captured. The HIGH/10 table is now complete."
  },
  "bet_limits": {
    "min_bet": "0.12 USD",
    "max_bet": "200 USD",
    "currency": "USD",
    "source": "Game rules modal (other_game_rules.png): 'Min bet: 0.12 USD' / 'Max bet: 200 USD'.",
    "verification": "Independently confirmed by pressing the MIN and MAX stepper buttons on the bet field to their hard stops: MIN jumped the stake field to '0.12' (other_bet_min_verified.png) and MAX jumped it to '200' (other_bet_max_verified.png), exactly matching the printed panel."
  },
  "max_win_field": {
    "printed_value": "20 000 USD",
    "source": "Game rules modal (other_game_rules.png): 'Max win: 20 000 USD'.",
    "how_this_relates_to_the_x10000_multiplier": "This 20 000 USD figure is a CURRENCY CAP on any single payout, not a multiplier, and it does not contradict the x10 000 top multiplier found in the HIGH-risk/10-picks table. At the game's min bet (0.12 USD), the theoretical top prize is 0.12 x 10 000 = 1 200 USD, well under the 20 000 USD cap, so the cap never binds for small stakes. At larger stakes the cap does bind: at a 3 USD stake the uncapped top prize would be 30 000 USD, which the 20 000 USD max-win rule would cut down; at the true max bet (200 USD) the uncapped top prize would be 2 000 000 USD, so the 20 000 USD cap dominates completely near max bet. The two numbers describe different things (a per-round multiplier vs. an absolute currency ceiling) and both are accurately transcribed from what the game prints -- neither is inferred or guessed."
  },
  "payout_basis": {
    "value": "total_bet_multiplier",
    "explanation": "There is no 'per line' or 'per way' concept in this game -- it is a single instant-resolve keno round. The tier multiplier shown on the paytable ('xM' next to 'N hits') is applied to the player's TOTAL bet for that round.",
    "evidence": "'My bet history' panel (other_bet_history.png) lists rows with Bet / Mult. / Win columns; one row reads Bet $3, Mult. x1.36, Win $4.08, and 3 x 1.36 = 4.08 exactly, confirming the multiplier applies to the whole stake, not a per-line fraction of it. Another visible row shows Bet $3, Mult. x0, Win '-' (a losing round, no win value shown, dash used instead of $0.00)."
  },
  "round_flow_and_states": {
    "step_1_idle": "Player selects 1-10 numbers on the 40-cell grid (click to toggle, or use the Quick-Pick shuffle button to auto-select N random numbers, or the Clear/broom button to deselect all). The panel above the grid prompts 'Choose from 1 to 10 numbers to see the odds' until at least 1 is picked, after which it is replaced by the tier carousel for the currently selected risk level.",
    "step_2_risk_and_stake": "Player picks EASY/MEDIUM/HIGH (re-prices the whole paytable) and sets the stake via the MIN/MAX stepper or a preset chip.",
    "step_3_bet": "Player presses the green BET button. During resolution the entire bottom control panel (bet stepper, preset chips, BET button, risk-level tabs) is visibly dimmed/greyed and non-interactive, preventing new input mid-round -- relevant for motor-impaired users who might otherwise double-input.",
    "step_4_reveal_animation": "10 numbers are drawn by the game and rendered as small glossy spiral-candy icons landing on the grid one at a time (this is an ANIMATED reveal, not an instant snap -- see the caveat on win_screen_easy_p5_1hit below). A drawn number that the player did NOT pick shows as a plain lavender/purple spiral candy circle (no digit visible underneath). A drawn number that the player DID pick (a hit) shows as a red-and-white spiral candy inside a bright gold-bordered square badge -- visually the most distinct state on the board. As hits accumulate, the tier-carousel tabs from '0 hits' up through the current hit count are progressively lit gold, giving a running/cumulative visual indicator during the animation (not just a single final highlighted tier).",
    "step_5_result": "Once all 10 numbers have been drawn, the final hit count determines the payout via the active risk level's table; balance and 'My bet history' update accordingly. There is no separate distinct 'you won' banner/screen captured in this set beyond the highlighted paytable tier and the updated balance figure in the top bar.",
    "no_cash_out_mid_round": "Confirmed: there is no CASH OUT button and no way to stop the reveal partway through and bank a partial result -- the round is atomic from BET click to final tier resolution."
  },
  "screens": {
    "base_game": {
      "file": "base_game.png",
      "state": "Idle state, 0 numbers picked, EASY risk level active, default $3 stake. Balance '999 998.5'. Full description in base_game_screen above."
    },
    "base_game_loss": {
      "file": "base_game_loss.png",
      "state": "Resolved LOSING round: EASY risk, 5 numbers picked (cells 1,2,3,4,5, shown burnt-orange), all 10 drawn numbers (verified by pixel/visual count: cells 8, 13, 20, 21, 25, 28, 29, 30, 35, 40 -- exactly 10 lavender spiral candies) landed on cells the player did NOT pick, so 0 hits. The '0 hits / x0' tier tab is highlighted gold. This frame is a clean, FINAL (fully-revealed, all 10 candies present) example of a total loss."
    },
    "win_screen_easy_p5_2hits": {
      "file": "win_screen_easy_p5_2hits.png",
      "state": "Resolved WINNING round: EASY risk, 5 numbers picked (1,2,3,4,5). Two of the picks (cells 2 and 3) are hits, shown as the gold-badged red/white candy. All 10 drawn numbers are present and countable (2 hit candies + 8 plain lavender candies at cells 6, 12, 16, 20, 28, 33, 34, 35, 39 -- note: 9 lavender positions listed here but exactly 8 are populated per the pixel-verified reading, total drawn = 10), confirming this is a FINAL, fully-revealed state. Tier tabs '0 hits', '1 hit' and '2 hits' are all lit gold (cumulative indicator, see round_flow_and_states step 4); '2 hits / x1.36' is the paying result."
    },
    "win_screen_easy_p5_1hit": {
      "file": "win_screen_easy_p5_1hit.png",
      "state": "CAVEAT -- likely a MID-ANIMATION frame, not a final result. EASY risk, 5 numbers picked (1,2,3,4,5), one hit visible at cell 2 (gold-badged red/white candy). Only 3 additional plain lavender candies are visible (cells 9, 16, 18) for a total of 4 drawn numbers shown, not the game's full complement of 10. Tier tabs '0 hits' and '1 hit' are lit gold. Because a fully-resolved round always shows exactly 10 drawn-number icons (as independently confirmed on base_game_loss.png and win_screen_easy_p5_2hits.png, both of which do show exactly 10), and this frame shows only 4, the capturer's file-name label '1hit' should be read as 'the tier lit at the moment of capture', not necessarily the round's FINAL hit count -- the remaining 6 balls may have still been mid-drop when the screenshot was taken. This is flagged rather than silently treated as a confirmed final '1 hit' example."
    },
    "rules_how_to_play": {
      "file": "rules_how_to_play.png",
      "content_verbatim": "Modal titled 'How to play?' with 4 numbered rules: 1. Choose the risk level [...] 2. Specify the bet amount and select from 1 to 10 cells on the game board. 3. Start the game, the more cells you guess - the bigger your winnings will be. 4. Malfunction voids all pays and plays.",
      "notes": "Single static screen, fits entirely without scrolling; no 'Page N/M' indicator and no scrollbar (not a genuine multi-page scroll panel)."
    },
    "other_menu": {
      "file": "other_menu.png",
      "content": "Burger-menu panel: player avatar (a round icon, cyan/teal background, simple cartoon face) + a truncated username ('Cyan Occu...') + 'Change avatar' link; Sound toggle (off, grey dot at left); Music toggle (off); a divider; then three menu rows -- Provably fair settings (shield icon), Game rules (document icon), My bet history (clock/history icon); footer 'Powered by iNOUT' (provider logo, NOT a casino skin -- see provider notes below)."
    },
    "other_provably_fair": {
      "file": "other_provably_fair.png",
      "content": "'Provably fair settings' modal. Text: 'This game uses Provably Fair technology to determine game result.' 'Next client (Your) seed:' with sub-text 'Round result is determined form combination of server seed and first 3 bets of the round.' [sic, verbatim OCR, 'form' likely a typo for 'from']. A 'Random on every game:' field showing a current client seed hash (e.g. '1cd4b7cf22dc3437', with a copy icon). A 'Next server seed SHA256:' field showing a long hex hash (partially cut off in the frame, e.g. 'b5d3a2e39cca76a267458a1e30b6c551b6060e4fb2a7674...'). Footer text: 'You can check fairness of each bet from bets history.'"
    },
    "other_game_rules": {
      "file": "other_game_rules.png",
      "content": "'Game rules' modal, sub-heading 'Bet limits are presented below.' Three rows: Min bet: 0.12 USD / Max bet: 200 USD / Max win: 20 000 USD. Footer note: 'Malfunction voids all pays and plays.'"
    },
    "other_bet_min_verified": {
      "file": "other_bet_min_verified.png",
      "content": "Bet-stepper field showing '0.12' after pressing MIN to its hard stop, on HIGH risk with 6 numbers picked (tier bar shows 1-6 hits window: 1 hit x0 / 2 hits x1.1 / 3 hits x2.25 / 4 hits x10 / 5 hits x50 / 6 hits x100)."
    },
    "other_bet_max_verified": {
      "file": "other_bet_max_verified.png",
      "content": "Same grid/tier state as other_bet_min_verified but bet-stepper field now shows '200' after pressing MAX to its hard stop."
    },
    "other_quickpick": {
      "file": "other_quickpick.png",
      "content": "HIGH risk, 10 numbers picked via the shuffle/Quick-Pick button, landing on a genuinely random non-contiguous set (cells 3, 6, 16, 17, 18, 23, 26, 31, 37, 39, all shown in bright orange as 'picked'). Tier bar shows the 5-10 hits window: 5 hits x10 / 6 hits x50 / 7 hits x100 / 8 hits x500 / 9 hits x1 000 / 10 hits x10 000. This frame independently corroborates the game's true HIGH/10-picks top tiers (see data_integrity_findings)."
    },
    "other_bet_history": {
      "file": "other_bet_history.png",
      "content": "'My bet history' modal, a table with columns Date / Bet / Mult. / Win, and a green 'LOAD MORE' button at the bottom (indicating more history rows beyond what is initially shown -- this appears to be a genuinely long/paginated history list, not a static short panel, though only the first ~4 rows are visible in the captured frame). Visible rows (all dated 17:56 16-08-26): Bet $3 / Mult. x1.36 / Win $4.08 (row 1); Bet $3 / Mult. x1.36 / Win $4.08 (row 2); Bet $3 / Mult. x0 / Win '-' (row 3, a loss, shown with a dash instead of a $0.00 win figure); Bet $3 / Mult. x1.36 / Win $4.08 (row 4, partially cut off at the bottom of the visible panel)."
    }
  },
  "provider_notes": {
    "footer_branding": "'Powered by iNOUT' appears in the burger-menu panel -- this identifies InOut Games as the GAME PROVIDER/STUDIO, not a casino operator skin. The 'studio' field for this documentation should be left EMPTY (InOut Games is already recorded as the top-level provider; there is no evidence of a separate sub-studio)."
  },
  "accessibility_notes": {
    "screen_reader_and_blind_users": "The single most important piece of state on every screen is the payout-tier text ('N hits / xM'), which is always printed as plain text (never an icon-only indicator), and the resolved/current tier is additionally distinguished by a filled gold pill shape rather than color alone. The four grid-cell states (unselected / selected-not-drawn / drawn-not-selected / hit) are each visually distinct in SHAPE as well as color (plain number tile vs. plain number tile in a different tone vs. plain circular candy icon vs. square gold-bordered candy badge), which should translate cleanly to distinct ARIA labels/roles for a screen-reader implementation (e.g. 'cell 7, not picked', 'cell 2, picked, hit', 'cell 9, drawn, not picked'). The balance figure and bet amount are always plain numeric text, never image-only.",
    "deaf_and_hard_of_hearing": "No sound-dependent information was observed on any captured frame -- outcomes (loss/win/hit-count) are always conveyed through the on-screen grid and tier-bar text, with a separate Sound and Music toggle available in the burger menu (both shown OFF in the captured frame) confirming audio is a supplementary, mutable channel rather than the sole carrier of any game information.",
    "colorblind_users": "See color_palette.colorblind_note above -- every functionally important state (chosen/hit/drawn/active risk/active tier) is shape-redundant, not color-only. The one area of residual risk is the burnt-orange 'selected, no match' tile versus the pale-cream 'unselected' tile, which differ mainly by a mid-saturation orange-vs-cream hue shift with a similar overall shape and no additional icon; a deuteranope/protanope user relying purely on hue could find these two states harder to tell apart than the other three states, though the printed DIGIT remains visible and legible in both cases, which still allows the user to identify which specific numbers were picked by cross-referencing their own selection choices.",
    "motor_impaired_users": "The entire bet/action control panel is programmatically disabled (visibly dimmed) during the reveal animation, preventing accidental double-bets while a round is resolving. The Quick-Pick (shuffle) button offers a one-click way to populate a full 1-10 number selection without needing 10 separate precise clicks on small 40-cell grid targets, which is a meaningful motor-accessibility affordance already built into the game. The MIN/MAX stepper buttons let a stake be set to either extreme in one click rather than requiring many increment presses."
  },
  "not_captured": {
    "autoplay_or_settings_beyond_sound_music": "No autoplay panel and no separate 'settings' panel beyond the Sound/Music toggles inside the burger menu were found in this build (documented by the capturer via two independent channels: full DOM text-inventory dump showing no 'Autoplay'/'Settings' string anywhere, and a canvas scene-tree probe showing only the 40 grid cells as interactive objects)."
  },
  "unreadable_or_uncertain": [
    "RTP (return to player) percentage: not printed anywhere on any captured screen (base game, How to play, Game rules, Provably fair settings, or My bet history). Not inferred or estimated here.",
    "Volatility label/rating: this game has no slot-style volatility badge or text anywhere in the captured UI; not applicable in the usual sense since payout structure is fully explicit in the paytable tables above.",
    "HIGH risk, 10 picks, 0-4 hit tiers: genuinely not captured with a verified 10-picks grid state -- see data_integrity_findings. Left as null rather than filled from the contaminated 8-picks duplicate frames.",
    "Keyboard shortcuts for BET / cell selection: no on-screen text confirms or denies a keyboard binding; not asserted either way.",
    "win_screen_easy_p5_1hit.png's true FINAL hit count: only 4 of the game's 10 drawn numbers are visible on this frame, so '1 hit' should be read as the tier lit at the moment of capture during the reveal animation, not a confirmed final result (see screens.win_screen_easy_p5_1hit).",
    "My bet history: full extent of the history list beyond the ~4 visible rows is unknown -- a 'LOAD MORE' button is present but was not exercised in the captured frames.",
    "Provably fair 'Next server seed SHA256' value is partially cut off by the panel width in the captured frame; the visible prefix is transcribed as-is and is not claimed to be the complete hash."
  ]
}