Data Status & Methodology

The archived snapshot is now identified as gag1-pets-2025-08-20-r1, evidence grade D. Its 103-record coverage, field definitions, source files, byte sizes, and SHA-256 checksums are published without changing the historical numeric values.

The crop calculator uses an unofficial formula documented in September 2025:

Game patches can change base values, compatibility rules, caps, and multipliers. Calculator output is an estimate, not an official or guaranteed current sell price.

Release gag1-codes-2026-07-30-r1 compares explicit claims from five dated secondary sites. Four agreeing active reports with no explicit expired report form the consensus tier; two reports form the limited-source tier; tied active and expired reports are labeled disputed.

A source omission is not treated as an expired report. No official developer-maintained list was found, and the publisher did not redeem the codes. Original Grow a Garden and Grow a Garden 2 claims are kept separate.

Release gag2-pet-tier-consensus-2026-07-30-r1 freezes three dated GAG2 pet rankings and preserves every S–C vote. Three matching votes are unanimous; two matching votes form a majority; source omissions remain null.

A large split is not averaged into a tier no editor chose. Robin's S/C/C result therefore stays disputed. Ability roles are navigational labels, not performance tests, and trade value is outside this dataset.

Release gag2-mutation-source-matrix-2026-07-30-r1 stores each visible numeric claim by page snapshot. It does not average conflicting values, treat missing values as zero, or count two pages from one publisher as independent confirmation.

The seven mutation rows are an overlap set, not a complete roster. Source statements about one-mutation-only versus variant-plus-weather stacking are stored separately because they conflict. No multiplier or stacking mechanic was reproduced in game by this publisher.

Release gag2-seed-shop-community-snapshot-2026-07-30-r1 transcribes 25 community Seed Shop rows while preserving source lineage. Matching copies are not counted as independent confirmation when they may share the Fandom table upstream.

Pineapple and Coconut alternative prices remain explicit. Probability outputs use 1 − (1 − p)^n under a clearly labeled independence assumption. The publisher did not inspect the game shop, and the page is not live stock, a countdown, or a notifier.

Release gag2-seed-tier-consensus-2026-07-30-r1 compares five editorial rankings whose scopes range from 27 to 34 seeds. Every S–D vote is preserved, while an omitted seed remains null instead of being treated as a low rank.

The displayed tier is the unique modal vote. All available votes matching is unanimous; more than half matching is a majority; one or two votes is limited coverage. Numeric S=5 through D=1 scores are used only for sorting and never rounded into a tier no publisher selected. No ranking was reproduced through controlled in-game tests.

Release gag2-gear-shop-source-matrix-2026-07-30-r1 stores a 27-row union because the nine reviewed community pages do not publish the same roster. Every numeric price remains attached to source IDs instead of being averaged or silently replaced.

Thirteen price conflicts and two roster-only flags remain explicit; additional roster differences can coexist with a price-conflict label. Stock waits use the same independent-roll equation, but chance and quantity primarily come from one community shop table. No row was checked in game, and the page is not live stock, a countdown, a notifier, or a purchase recommendation.

The night-stealing audit compares 16 attack, defense and recovery claims across seven community pages. A claim is repeated only when multiple reviewed pages publish it; this label does not mean official or independently reproduced.

Single-source mechanics stay limited. The owner-present garden wording stays conflicting because reviewed pages alternate between “locked” and “deterred.” No exploit, automation or movement-bypass tactics are included.

The guild audit separates setup claims from weekly competition snapshots. Sixteen claims are graded across seven pages, while the latest reviewed Pet Safari reward ladder keeps every rank band attached to its source coverage.

Older Biggest Plant and Ice Serpent tables remain historical rather than being silently overwritten. A missing 1,001–5,000 row is marked as a one-page filled gap, and the live Guild panel remains authoritative after a weekly reset.

Release gag2-auction-source-audit-2026-07-30-r1 compares 18 access, mechanics, inventory, timing, stock and strategy claims across seven community pages. Twelve item mentions are a source union, not a complete current roster.

Five reviewed pages support a 30-minute rotation while one dated update page says 20 minutes. Robux availability and exact sniping thresholds also remain conflicting. No live stock, synchronized countdown or fixed price table is inferred from those sources.

Release gag2-daily-deal-source-audit-2026-07-30-r1 compares 18 Daily Deal, historical Bargain and Double or Nothing claims across eight community pages. Historical Bargain fees are never silently applied to the current tool.

The 2.2× and 5× multiplier claims, eligible sale scope, reset anchor and Sell Stock stacking remain explicit conflicts. The calculator therefore accepts player-observed inputs and does not claim a live offer, timer or verified win rate.

Release gag2-money-guide-source-audit-2026-07-30-r1 compares 20 foundation, cash-flow, crop-route, upgrade, timing and risk claims across nine money and beginner guides.

Repeated Bamboo and Mushroom recommendations remain separate from anecdotal maximum payouts. The calculator uses a player-observed cycle and fixed reserve, so no community jackpot or universal reserve percentage becomes a preset.

Release gag2-beginner-guide-source-audit-2026-07-30-r1 separates 25 claims across 11 pages into first-party boundaries, repeated guidance, limited evidence, conflicts and explicit publisher guardrails.

No named crop, pet, gear or price becomes a permanent preset. The two-cycle reserve is labeled as editable editorial policy, and public-server night exposure can override expansion.

Release gag2-seed-packs-source-audit-2026-07-30-r1 compares 11 named pack records and ten probability presets across ten pages. Roblox defines the game, seed shop and guild reward boundary but does not publish pack tables.

Every probability keeps its community verification label. The independent-roll model separates expected rolls from target confidence, assumes no pity or luck modifier and never treats a published table as proof that a TBA pack is obtainable.

Release gag2-badges-source-audit-2026-07-30-r1 preserves 23 official badge IDs, names, descriptions and a dated award-count snapshot from the Roblox Badges API, then compares three ranking pages for roster and requirement conflicts.

Official fields are never replaced by a guide's wording. Categories and award-count bands are labeled editorial; changing statistics remain tied to the release date, and an enabled historical badge is not called currently earnable.

Release gag2-npc-locations-source-audit-2026-07-30-r1 compares 24 roster, stand, zone and evidence-boundary claims across ten pages. The official Roblox listing defines the game and systems but does not supply an NPC roster or coordinates.

Names and stand labels require community agreement. Exact map positions are not inferred from vague “central hub” language, while generic secret-merchant, quest-giver and crafting claims remain visible as unsupported conflicts.

Release gag2-trade-calculator-source-audit-2026-07-30-r1 compares 18 identity, valuation, transfer and safety claims across official Roblox metadata and five community pages.

Fixed community values are not converted into a publisher price list. Players enter low, typical and high observations in one comparable unit. WFL range math and transfer safety remain separate because reviewed pages conflict on trade-window, guild, mailbox, gift and drop scope.

Release gag2-stock-tracker-source-audit-2026-07-30-r1 compares 19 timing, synchronization, live-stock, notification and prediction claims across eight current pages.

Roughly five minutes is an editable Seed Shop default, not a publisher-measured constant. The tool advances a refresh anchor supplied by the player; it has no verified inventory feed, makes no current-stock claim and excludes future-item prediction plus original-game Egg and Cosmetics timer carryovers.

Release gag2-admin-abuse-source-audit-2026-07-30-r1 compares 23 identity, schedule, history, Explorer and event-identity claims across official Roblox metadata and eight current pages.

The tool starts with no timestamp. Players can synchronize the remaining time they see in game or enter a newly verified exact local time. Explorer's Fall Harvest association remains separate from competitor Admin Party claims, while conflicting July 31, August 1 and August 2 interpretations stay visible.

Release gag2-update-log-source-audit-2026-07-30-r2 compares 29 archive claims across Roblox metadata and nine community or editorial update pages.

Reports are joined only after their content bundles overlap. Adjacent-day dates remain ranges, and incompatible labels such as 1.05.0, 1.3.0 and Update 2 stay source-specific. The latest repeated community window is not promoted to an official patch note or next-update prediction.

Release gag2-crop-values-source-comparison-2026-07-30-r1 defines one metric before comparing rows: average unmutated Sell Stand Sheckles at the listed average fruit weight. Seed prices and player-to-player trade points are excluded because they answer different questions.

The 43-row July reference is cross-checked against two narrower recent tables. Twenty-five rows match across all three. A June 15 community-tested table differs on all ten overlapping crops, so those observations remain dated conflicts instead of being averaged into a false consensus. No row was reproduced in game by this publisher.

Release gag2-egg-chances-source-audit-2026-07-30-r1 stores 44 source-level target claims without forcing their surrounding pools to total 100%. The repeated eight-pet Common table totals 89.8%; adding the Fandom Turtle row produces 107.3%. Epic and Rainbow snapshots expose separate conflicts.

Calculator outputs use 1 − (1 − p)^n only as a fixed-rate, independent-roll scenario for one selected target. They do not prove the target rate, normalize a pool, model pity, guarantee an outcome, or reproduce the game. Rainbow rows totaling 132.9% remain visible but are excluded from presets.

Release gag2-weather-source-audit-2026-07-30-r1 separates eleven cycle, weather and moon states across four community sources. Rainbow weather and Rainbow Moon remain distinct, while Lightning keeps both its five-minute and two-minute-thirty-second duration claims.

The readiness planner starts from the event a player can see in game. It does not predict weather, publish a future event time, or claim access to a live server feed. Mutation multipliers remain in the separate mutation audit.

Release gag2-prop-crates-source-audit-2026-07-30-r1 compares six community pages whose “all crates” scopes range from 14 to 19 records. The published page uses a 17-row reviewed union and labels later, limited, and release rows separately.

Item odds remain exactly as published. The Spring Crate total stays at 96%, Fence at 99.99%, and Fourth of July at 100.03%; none is normalized. Probability outputs assume independent fixed-rate openings and do not model pity, stock, guarantees, trading, or effects. No claim was reproduced in game.

Release gag2-pets-source-audit-2026-07-30-r1 compares six community pages and preserves a 17-pet reviewed union. Published roster scopes range from 12 to 17 pets, so omission is not treated as proof that a pet was removed.

Six records retain conflicting price, access, or mechanic claims. Older and newer Black Dragon and Ice Serpent descriptions remain separate rather than being merged into a false current fact. The publisher did not reproduce any price, spawn chance, ability, or obtainment route in game.

Community submissions are review candidates, not automatic proof. A submission is not incorporated into a published market range until its item specification, time, currency, and evidence have been checked.

To report an error, use the contact page and include the affected URL, game version, expected value, and supporting evidence. Do not send passwords, payment details, or unnecessary personal information.

Page materially revised: July 30, 2026.