Everything DreamBay can do

DreamBay is a cross-chain home for digital collectibles. We started on Hedera and grew into the wider Web3 creator economy — Hedera and the XRP Ledger under one roof. Collectors trade, creators launch, players cast for prizes, and everyone earns DreamPoints along the way, no matter which chain their assets live on.

Pick the wallet that matches the chain you're trading on. You can connect more than one — switch between them with the Account dropdown.

• HashPack — primary Hedera wallet. Native HBAR + HTS support, browser extension and mobile.
• Magic — email-based Hedera wallet. No browser extension required. Good for new users; same on-chain as HashPack.
• Xaman (XRPL) — for XRP NFTs. Mobile-first signing flow. Each protected action requires opening the Xaman app.

Switch chains using the network selector to the right of the DreamBay logo.

• Hedera (HTS) — primary network. All DreamBay features live here: marketplace, MemeBay, Launchpad, DreamCast, DreamPoints.
• XRP Ledger — XRPL NFT support powered by our XRP build. Switch to XRP via the chain selector to land on the dedicated experience.

The Discover page is your home base. Featured collections, top movers, hottest sales, biggest pools, and trending memecoins all live here. Use the left sidebar to dive into:

• Collections — every collection on Hedera, sortable by floor, volume, listings, royalty %, and more.
• DreamDeals — algorithmically picked "below floor or undervalued" listings refreshed every 5 minutes.
• Activity — real-time sales feed across the entire ecosystem.
• Watchlist — your saved collections. Set price alerts and get push notifications when listings drop below your threshold.

You've got two ways to buy on DreamBay — both zero fee:

• Buy Now — click the price on any listed NFT. Sign once in your wallet; the NFT is yours when the network confirms (usually under 5 seconds on Hedera).
• Sweep — buy multiple NFTs at once. Select your items via the cart toggle on each card, then hit BUY N ITEMS in the bottom action bar. One signature, one atomic transaction.

If your wallet hasn't held this token before, Hedera requires a one-time association (~0.05 HBAR). DreamBay handles this for you in the same flow — you just sign once.

Two ways to list:

• Quick List (top right of header) — paste a token id or pick from your library, set a price, sign one allowance, done.
• From My Library — open any NFT, hit the Sell button, set price.

Single-NFT offers — bid on any specific listing or unlisted NFT. Funds are escrowed via HBAR allowance to our operator wallet so the seller can accept any time within the offer's expiry.

Collection offers — make one offer that any holder of a collection can accept. Great for collectors who want a piece but don't care which serial.

Auctions — time-bounded sales with a starting price. Highest bid at the end wins; settlement is automatic.

Hedera NFTs support native royalties enforced by HTS — every secondary sale automatically pays a percentage to the configured creator wallet. No marketplace can opt out.

• Maximum total royalty: 9.99% (Hedera network limit)
• Recommended: 5–7% for healthy secondary market activity
• Can be split across multiple recipients (e.g. team + treasury)
• Set once at collection creation; update later via the metadata key if enabled

MemeBay is a fully on-chain memecoin launchpad. It deploys a fixed-supply HTS fungible token + an immutable bonding-curve smart contract per launch. Anyone can launch; nothing is gated.

• Total supply — 1,000,000,000 (1B) per token, fixed
• Bonding curve — constant-product AMM with virtual reserves; price grows smoothly as buyers come in
• Graduation target — $10K USD market cap (priced live via Chainlink HBAR/USD)
• SaucerSwap LP — at graduation, 19% of supply + collected HBAR is paired and added; LP burned to dEaD

From the MemeBay landing page, click Launch a Token. Pick a name, symbol, description, and (optional) image + socials. Sign one transaction in your wallet — that pays the Hedera network creation fee (~15 HBAR) and deploys both the token and your bonding curve atomically.

Your token is live on the homepage seconds later.

Buys and sells go directly to the curve. The curve sets the price using the constant-product formula; your slippage tolerance is configurable per trade (default 5%).

The creator share is sent to the original creator's wallet on every single trade — instantly, in HBAR, while the curve is live. No claim flow needed.

When the curve's market cap reaches the live $10K USD target (priced via Chainlink), the curve graduates:

1. Bonding curve closes — no more buys or sells on the curve
2. 5,000,000 tokens (0.5%) → minted to the creator's wallet
3. 5,000,000 tokens (0.5%) → minted to the protocol treasury
4. ~190M tokens (19% of supply) + the curve's collected HBAR → seeded into a SaucerSwap LP
5. LP tokens burned to 0xdEaD — fully decentralized post-graduation

If you launch a token on MemeBay, you earn forever:

• While on the curve: 0.1% of every buy + sell, paid in HBAR, sent directly to your wallet on each trade
• At graduation: 5,000,000 tokens (0.5% of supply) — these become tradeable on SaucerSwap immediately
• Post-graduation: any future SaucerSwap trading fees go to LP holders; your LP share is zero (LP is burned), but you hold liquid tokens

The DreamBay Launchpad is a self-service NFT collection launcher. Every approved project gets:

• Default mint page hosted at dreambay.io/on-chain/launchpad/hedera/<your-slug>
• Multi-stage mint support (private, whitelist, public)
• Whitelist + allocation management
• Free claim or standard mint
• Real-time mint counter, social links, banner image

Buyers click Mint, choose a quantity, sign once, and the operator atomically transfers all the NFTs in a single Hedera multi-NFT transaction. Atomic = either all succeed or none — no partial mints, no stuck money.

• Per-NFT platform fee: 2% (creator keeps 98%)
• Free claim mints: 0% platform fee — you cover only the per-NFT network mint cost (~0.05 HBAR)
• Delegated minting (optional): server-side batching at ~0.25 HBAR/NFT, no per-batch wallet popup

Each launch can have unlimited stages with their own price, start/end time, and access rule. Common patterns:

• OG / Holder — gated by ownership of another collection (e.g. Wild Tigers holders)
• Whitelist — explicit wallet list managed via the Edit Mint modal
• Public — anyone can mint, optional per-wallet limit

Beyond the default mint page, we offer optional paid services:

• Custom mint page — fully custom themed mint UX (think Catizens box-tilt, Hboys mask reveal, Project 1947 vault — every collection gets its own moment)
• Website build — landing page, roadmap, team, mint links
• Art & design — collection artwork, traits, banners, branding, social media graphics

Pick any from the Create Collection wizard and our team will email you within 3 business days. Pricing is project-by-project.

DreamCast is our flagship NFT mini-game. Pool owners stake NFTs into pools with configured odds; players cast lines for HBAR and roll the dice. Real-time, fast-paced, and fully on-chain.

Pick a pool, click Cast (single) or use Multi-Cast (2–10 in one signature). The cast pays HBAR to the pool. Outcome is decided by the pool's configured odds across rarity tiers (Small Fry, Keeper, Kraken, etc.). If you catch — the NFT is transferred to your wallet.

Buyback: don't love what you caught? Sell it back to the same pool for the configured buyback price. The pool keeps it for future casts.

Buff your line before casting:

• Worms — small consumable boosts to general odds (one-cast use)
• Charms — stronger persistent effects targeting specific tiers (last across multiple casts in a chain)

Both cost HBAR; the pool owner keeps 98%, DreamBay takes 2%.

Anyone can run a DreamCast pool. You set:

• Cast price (in HBAR)
• Buyback price (sub-set of cast)
• Odds across rarity tiers — must add up correctly
• Which NFTs are stocked (you grant the operator allowance over the serials)

Owners earn 98% of every cast / charm / worm purchase. DreamBay's 2% covers operator gas + platform.

DreamPoints (DP) is the loyalty currency of DreamBay. Earn it by being active; spend it on perks. It's off-chain (database-tracked), tradeless, and tied to your wallet — no extra signup.

• Daily login — small DP every day you visit. Streaks compound.
• Trades — buy or sell on the marketplace; DP scaled by HBAR volume
• Listings — the highest-leverage daily activity. Up to 1,000 DP/day by rarity. Floor-quality multiplier rewards listings at or near floor.
• Quests — limited-time challenges (e.g. cast 5 times in DreamCast, mint a launchpad NFT)
• Referrals — DP when friends sign up using your link + a slice of their lifetime activity
• Holdings — passive 1 DP/day per held NFT, capped at 50/day. Reduced from 5/day in the May 2026 rebalance — listing replaces passive holding as the dominant non-trade earner.

Lifetime DP determines your tier. Tiers unlock larger perks and earn higher DP multipliers.

Spend DP in the perk shop:

• Mystery boxes — repeatable, randomised payouts
• Raffle entries — recurring NFT/HBAR draws
• Profile frames — animated borders around your avatar (one-time unlock)
• Private mints — exclusive launch slots reserved for high-tier holders
• VIP events — IRL meetups, AMAs with creators, sneak peeks
• Custom traits — create your own trait variant on supported collections
• Whale Vote — name a featured collection or feature flag for a week

The NFT Index 500 is a single-number readout of overall NFT marketplace activity, built the same way CoinGecko and the S&P 500 build theirs: pick a fixed set of constituent collections, sum their activity, divide by a divisor calibrated so the index starts at 1,000. Moves from there reflect the real market — not changes to our data coverage or methodology.

• Formula — Index = (7-day moving average of constituent volume) ÷ Divisor. A reading of 1,000 = the level the index opened at. 2,000 = double that activity. 500 = half.
• Constituents — the top 50 collections by lifetime trading volume on each chain. Frozen at launch, rebalanced quarterly. The current member list and each collection’s share are published alongside the index — same transparency principle the S&P 500 uses.
• Divisor — calibrated once at launch so the opening reading was exactly 1,000. On quarterly rebalances (when a collection joins or leaves the top 50), the divisor is re-calibrated so the index value RIGHT BEFORE the rebalance equals the value RIGHT AFTER. The chart stays continuous; the rebalance doesn’t cause an artificial jump.
• 7-day moving average — single-day spikes (mint launches, holiday closes, weekend bursts) don’t whip the headline number around. The average smooths them while still letting the index respond to real multi-day trend changes.
• Why this beats a rolling baseline — earlier designs pegged the index to a single historical day. That meant if our data coverage grew (e.g., we started indexing more collections), the measured volume jumped and the index printed values that misrepresented reality. The divisor approach measures market change, not coverage change.

The Token Index 500 is a composite score for the token ecosystem on each chain. It blends two things: total liquidity locked across all tokens (TVL) and the volume actively trading today. Built the same way CoinGecko builds its Global Crypto Index — the result is a single number that moves when either depth or demand changes, anchored to a launch value of 1,000 so movements are easy to read.

• Formula — Index = (0.7 × TVL + 0.3 × Volume) ÷ Divisor. TVL is weighted 70% because liquidity is the more stable backbone; 24h volume contributes the remaining 30% to keep the score responsive to live market activity.
• Divisor — calibrated once at launch so the opening reading was exactly 1,000. When the constituent token set is rebalanced quarterly, the divisor is re-calibrated using the standard S&P 500 adjustment so the index stays continuous across the change. A token joining or leaving the index doesn’t cause an artificial jump.
• Constituents — every major token on each chain (wrapped and bridged stables excluded — they don’t reflect ecosystem health). The full member list and each token’s share of the total TVL are published alongside the index.
• Daily snapshots — every point on the chart is a real recorded TVL + volume reading taken once per day. The chart fills in over time — sparse on launch day, denser each day after.
• Why this beats a price chart — earlier designs scaled a single token’s price history to fake a longer chart. That was misleading: it showed one token’s movement, not the whole ecosystem’s. The index now reflects real ecosystem health, even when that takes a few weeks of history to fill out.

Awarded automatically as you hit milestones — first trade, 10K volume, kraken catch, multi-collection holder, founder, and more. Earned badges show up next to your name across the platform.

The Creator Dashboard ( /on-chain/dashboard/creator) is mission control for creators. From here you can:

• Edit your public profile (name, avatar, bio, socials)
• Create a new NFT collection (3-step wizard)
• Import an existing HTS token you control
• Set up a launch event (mint stages, whitelists, free claim)
• See all your collections + bonding curves in one place
• Browse our fee structure
• Get help via Discord or email

The 4-step wizard:

1. Basic Information — name, symbol, max supply, description, treasury wallet
2. Creator Royalties — set up to 9.99% in any number of recipient splits
3. Token Keys — pick which keys to enable (supply, metadata, admin, fee schedule, pause, freeze, wipe). Disabled keys can't be turned on later.
4. Review & Launch — optional add-on services + Terms agreement. Submit and your collection deploys instantly.

Self-service launches go live the moment you submit. If you check any add-on box, our team emails you within 3 business days to coordinate.

Already minted a collection elsewhere? Import it via the dashboard. We verify ownership via Mirror Node + a 1-tinybar attestation transaction proving you control the treasury wallet, then it appears under Your Collections.

Your profile (visible at /on-chain/library?wallet=...) shows your username, avatar, banner, bio, socials, achievement badges, DP tier, and your full NFT inventory. Edit it from either My Library or the Creator Dashboard — it's the same record.

Refreshed every 5 minutes. Compares every active listing to its rarity rank and surfaces underpriced gems — the higher-rarity NFTs listed below their statistical fair value. Filter by collection, max price, or rarity %.

Side-by-side compare two collections. Floor, volume, holders, listings, royalty, mint date, top traits — useful for deciding which whitelist to spend on or which floor to scoop.

Platform-wide stats and charts: HBAR price, volume, marketplace activity, MemeBay launches per day, DreamCast pool flow, top creators. Profile-specific analytics live at /on-chain/profile-analytics.

Save collections to your watchlist. Set a price alert (e.g. "notify me when floor drops below 50 HBAR") and we send a push notification + email the moment it triggers. Alerts are checked every 5 minutes.

DreamBay surfaces two analytics signals on every Hedera NFT collection: Wash Pattern % (the share of trader-to-trader volume that matches coordinated wash-trade patterns) and Organic % (the share of all reported volume that happens between independent peer wallets, after excluding project mint mechanics and detected frequent-trader clusters). Both are descriptive analytics computed from public on-chain data. The two metrics answer different questions with different denominators — they are not complementary and they should not sum to 100%.

Click any wash %% chip or organic %% chip on a collection page to open the in-line popover. The popover lists the contributing signals, shows the algorithmic safeguards that fired, surfaces the volume-decomposition bars (Project mechanic / Frequent-trader cluster / Independent peer), links here to the full methodology, and offers a one-click dispute path. The numbers shown on the chip are computed from public on-chain data — every flagged trade is a real nft_sale row visible on HashScan.

For every NFT sale that lands on Hedera, we compute a pattern-confidence score from 0–100 based on the statistical structure of the surrounding sales history. Scores are then aggregated across four parallel windows (24h, 7d, 30d, all-time) into per-collection wash percentages: the share of total HBAR volume in each window attributable to sales whose surrounding patterns crossed our confidence threshold (60+). The window matched to the period you're viewing is what the chip displays; narrower windows cascade to wider ones only when there isn't enough data to rate confidently. The score measures pattern fit — it does not measure or claim to measure intent.

The chip appears in three places: the "Wash Pattern %" column on the on-chain rankings table, the "Wash Pattern" tile on every collection's page header, and a per-sale flag in the NFT activity feed. All three link back to this page.

• Algorithm version: v1.0 (any chip computed under an older version is flagged in its popover)
• Window selection: matches the period you're viewing (1d → 24h window, 7d → 7d window, 30d → 30d window, all → lifetime). If the preferred window has too few sales to band confidently, the chip cascades outward to the next-widest window and shows a small ·7d / ·30d / ·all suffix so you know it's a fallback.
• Minimum sample size: 10 sales in the window. Below this, the window is treated as too thin to rate, and the algorithm cascades. If every wider window also falls below 10, the chip shows Unrated rather than a noisy percentage.
• Refresh cadence: every 3 minutes incrementally (re-scores any collection that had a new sale in the last 30 minutes), plus a top-100 force-refresh every 30 minutes that re-validates the visible homepage rollups even when no new sales happened. Single-collection live recompute available on demand via the API.

Every rated collection lands in one of five bands, picked from its wash percentage:

• ● Low — under 5%. Healthy market signal.
• ● Medium — 5–20%. Worth a closer look.
• ● High — 20–50%. Take headline volume with a pinch of salt.
• ● Severe — over 50%. Trade carefully; published volume is unreliable.
• ● Unrated — fewer than 10 clean sales in every available window (24h, 7d, 30d, all-time). "Clean" excludes sales touching the project's treasury, mint distributor, or detected buyback bots — see below. Thin samples are where false positives bite hardest, so we refuse to band them.

The small label after the percentage (·7d, ·30d, ·all) means the chip cascaded from the period you're viewing to a wider window. Example: on the 1D period, "2.8% ·7d" means "there weren't 10 clean sales in the last 24h to rate confidently, so we're showing the 7d-window rating". No suffix = the rating is for the selected period directly. This is intentional: a 50% wash chip computed from 4 sales is statistical noise; the cascade keeps small samples from telling false stories.

Each sale is evaluated against eight independent detectors. A single strong signal can flag a sale on its own; weaker signals only flag in combination. Final confidence = MAX of all signal scores, plus a +15 combo bonus when two or more signals fire at meaningful strength (capped at 100). A trade is flagged as wash when its confidence score crosses the 60 threshold. This aggregation favours specificity — we don't combine a bunch of weak hints into a false positive.

The seven signals we evaluate (descriptions only — the per-tier point weights are intentionally not published, see the note at the end of this section):

• Self-trade. Buyer and seller account IDs are identical. Rare on Hedera (fees discourage it) but unambiguous when it happens — the strongest single-signal flag.
• Wash loop (A→B→A). The NFT was sold by the seller, bought by the buyer, then sold back. Scored higher for tighter time windows; loops older than ~7 days score below the combo-bonus floor and only contribute when other signals also fire.
• Rapid churn. Same NFT changes hands repeatedly in a 7-day window. Scored higher as the trade count grows; two trades alone is below the wash-flag threshold and only contributes via the combo bonus.
• Pair cadence. Same buyer/seller pair trading multiple distinct NFTs within five minutes. Scored higher as the number of NFTs in the burst grows. Repeated trades of the same NFT are handled by wash-loop instead.
• Round-trip, no delta. A loop where the final price is within a tight band of the entry price. Manufactured volume with no economic exit — one of the stronger single-signal flags.
• Stair-step pricing. Same NFT changes hands three or more times at near-identical upward price-step ratios within 7 days. This is a pattern associated with non-organic price discovery — the data is consistent with the price being walked up between cooperating wallets, though the pattern alone does not establish intent. Downward staircases aren't flagged here; they show up via wash-loop or round-trip when the loop closes.
• Pair rotation. Same buyer/seller pair churning many distinct NFTs in a 24-hour window. Catches the "different serials every 10 minutes" evasion that side-steps wash-loop and rapid-churn. Scored higher as the distinct-NFT count grows.
• Self-financed pair. Buyer's wallet was originally funded by the seller (or both wallets share a common upstream funder one hop back). This is Chainalysis's primary "habitual wash trader" signal — buying from your own sock-puppet wallet. Known onboarding operators (e.g. SentX onboarding accounts) are exempted so users funded by a common operator aren't mis-flagged.

Why we don't publish the per-tier point weights: the exact tier values for each signal (e.g. "wash-loop within 24h scores X, within 7d scores Y") are intentionally held back. Publishing them would let a motivated wash trader engineer trades that score 59 by design — just below the 60-threshold — and convert this page into an evasion playbook. The aggregation math, the threshold, the safeguards, and the signal categories themselves are all public so any flagged trade can be audited; the per-tier weights live alongside the cluster-detection internals as the only operational specifics we keep private. This is the same disclosure model Chainalysis, CryptoSlam, and FICO use for their fraud-detection scoring.

The system is honest about its limits. These are the explicit gaps:

• Intent. We detect patterns, not motives. A signal can fire on legitimate market-making, a creator buying back for an airdrop, or two wallets owned by the same human rebalancing.
• Bot vs human. "Wash trade" is precise; "bot" is broader. Plenty of bot activity is legitimate (market-making, arbitrage), and plenty of wash trading is hand-clicked. We call this wash-trade detection because that's what we can defensibly measure.
• Off-chain coordination. Sales lined up in Discord DMs or OTC deals show up as ordinary trades.
• Slow pair rotation. A pair that rotates serials at a slower-than-24h cadence and never trades the same serial twice will trip neither wash-loop nor pair rotation. We catch the bursty case; patient adversaries are still beyond what we currently measure.
• Cross-chain. We only score Hedera nft_sale rows (which include SentX / Kabila / Zuse / Hashport / DreamBay sales). XRP rolls in once the XRP indexer ships matching data.
• Algorithm staleness. Once detection logic is public, adversaries adapt. Future revisions are version-tagged and chips from older revs are flagged in their popover.

Because every signal works from public on-chain trades, a determined party can deliberately produce wash-pattern trades on a competitor's collection to drag its rating down. This is real and we treat it as a known threat, not a hypothetical.

The cheapest version: an attacker controlling two wallets buys an NFT, churns it A→B→A within seven days, and adds one more pair trade within five minutes of either leg. That sale clears our 60-confidence threshold for roughly the cost of three marketplace fees and slippage. Repeated 50 times on a thinly-traded collection, the attacker can push it from Low into Medium for a few hundred dollars of inventory churn.

Defamation-proofing safeguards active in v1.0:

• Treasury account exclusion. We fetch each token's treasury account from the Hedera mirror node and filter sales touching it on either side. Mints, creator buybacks, airdrops, and burns can look algorithmically identical to wash trading — but they aren't market activity, so they don't feed scoring. Auto-detected per collection and cached.
• Severe-band qualifier. A public "Severe" rating requires three independent corroborating conditions: sample size ≥ 100 sales, at least 2 distinct flagged signals firing across the window, and a top wallet cluster of ≥ 3 wallets (no 2-wallet grief). If any condition fails, the chip shows "High" with a note explaining the downshift — never a Severe accusation on thin evidence.
• Confidence-aware downshift. When the concentration cap fired AND the sample is below 100, the displayed band drops one level. Single-pair-dominated small collections never get strong public ratings.
• Rating gate. Any collection whose source data has NULL buyer/seller rows (mint events misindexed as sales by the upstream pipeline) is displayed as "Unrated — limited evidence" regardless of computed percentage. The audit row still stores the raw band for transparency.
• Concentration audit (two triggers).We cluster every flagged sale by connected wallet component — each sale is a graph edge between buyer and seller; transitively-connected wallets become a single cluster. Two triggers fire the algorithmic cap:
  • Single-cluster trigger. When one cluster drives more than 50% of flagged volume, the chip popover surfaces a "single-cluster concentration" note and the algorithm caps that cluster's contribution at 50% of the flagged bucket. This blocks the cheapest grief attack (two wallets, or two wallets plus a relay).
  • Top-three trigger. When the three largest clusters together drive more than 75% of flagged volume — even if no single cluster crosses 50% — the cap fires on their combined contribution. This blocks the slightly-pricier grief attack (three disjoint wallet groups at ~33% each).
  Either way, a coordinated grief attempt can't single-handedly push a collection past a band the rest of the flagged activity wouldn't support on its own.
• Sudden-jump trigger. Any collection whose 24-hour wash percentage rises by more than five points between refreshes lands on the manual-review queue (metadata.wash_review_queue). The popover shows a "rose by Xpp — flagged for review" badge; the chip continues to render the algorithmic rating while triage is pending.

Short answer: yes, the pattern can be generated by anyone. Wash-trade patterns are a property of the structure of the trade sequence, not the identity of the trader. Two wallets producing trades that fit the pattern generate the corresponding signature on that collection's data — regardless of whether those wallets belong to the project team, an outside flipper, a friend group, or a grief attacker. The buyer who pays based on price discovery influenced by that pattern loses the same money either way. That is precisely why the metric measures market-quality patterns, not project conduct.

This is the most common misunderstanding about the wash %% metric, and it deserves a direct answer. A common rebuttal sounds like: "Two random wallets washing this collection don't represent the project — calling it ‘our wash %’ is unfair." The framing is incorrect on two counts. First, the chip has never accused the project — or any party — of wash trading. It is a pattern-detection signal: it reports the share of trader-to-trader volume whose data matches coordinated wash-trade signatures. Second, independent flippers, friend groups, and grief attackers can all generate that pattern on a collection they have zero formal affiliation with — so labelling the chip with the collection's name is descriptive of the data on that collection, not an attribution of conduct to its creators.

At least eight well-documented motivations exist for non-team wallets to wash trade a collection. None of them require any project involvement:

• Token / point incentive farming. When a marketplace rewards traders for volume (LooksRare's $LOOKS token program is the textbook case), self-trading harvests rewards with minimal risk. The Chainalysis 2022 NFT wash trading report found this was the dominant source of NFT wash trading on Ethereum — outside flippers, not project teams.
• Ranking manipulation for flips. Pump a collection into a "trending" slot to attract real buyers, then dump into the artificial demand at the top.
• Floor manipulation. Wash a few sales at high prices to make the floor look higher, then offload to a buyer who got tricked by the inflated price discovery.
• Airdrop / reputation farming. Chains and platforms airdrop tokens or status based on "active wallet" criteria. Wash trading qualifies a wallet without any economic risk.
• Friend-group coordination. A small group of holders pump each other's trades to attract attention to a collection they all hold. Often informal, no central coordinator, but produces wash-pattern volume.
• Cover for market making. A market maker accumulating inventory may wash trade to obscure their entry price from competitors who would front-run them.
• Tax loss harvesting. Fake sales between wallets you control to realize losses on paper. Legally dubious in most jurisdictions, but it happens.
• Grief attacks. An attacker deliberately wash-trades a competitor's collection to give it a high wash score and hurt its reputation. The cheapest version costs only a few hundred dollars and is exactly why DreamBay ships the concentration safeguards documented above in why a clean collection might appear flagged.

The receipts: when Chainalysis published its 2022 NFT wash trading analysis, it identified clusters of 100+ self-funded wallets controlled by individuals churning hundreds of NFTs across dozens of collections — collections those individuals had no formal affiliation with. The report was striking precisely because it killed the "it must be the team" framing. Wash trading on a collection is dominated by independent actors, not by project teams.

What this means for how to read the chip:

• A wash %% chip describes the collection's market quality — "is the price discovery here trustworthy?" — not the project team's conduct.
• A clean team can still have a polluted wash %% if outside flippers or grief attackers are active on the collection.
• The dispute path is available for projects that believe specific flagged wallets were misclassified (e.g. legitimate market makers, treasury rebalancing, etc.); the algorithm can carve out allowlists when evidence supports it. See disputing a rating.
• Restricting detection to only labeled project wallets would miss the dominant share of real wash trading — which is exactly what every serious anti-fraud system (Chainalysis, CryptoSlam, OpenSea's internal filtering, FICO in adjacent fraud detection) refuses to do, for the same reason.

Every rating is computed from public on-chain data. If you believe a collection has been mis-rated:

1. Open the rating chip's popover and note which signals are contributing most.
2. Look up the flagged sales on HashScan using the collection's token id. Every sale we score is a real on-chain row.
3. If a signal is firing on benign activity (coordinated airdrops, treasury rebalancing, known market-maker wallets, etc.), let us know — the next algorithm revision can carve out allowlists or refine the heuristic.
4. Fastest path: open a ticket in Discord — the team responds in real time most days. Or email dreambay@possestudios.io with the collection address and the specific sales you believe are mis-classified.

Important context up front: the Organic % metric (previously labelled “Independent %” in earlier builds — same number, clearer name) describes volume composition based on observable on-chain patterns. It is not an accusation of wrongdoing. Many projects run legitimate mint mechanics, rarity lotteries, and instant-sell buyback features. The metric simply tells users what share of reported volume comes from those mechanics versus secondary peer trading, so people can read marketplace volume figures with eyes open.

Marketplaces report "volume" as the sum of HBAR paid across every recorded sale of a collection. That number doesn't distinguish between two structurally different kinds of activity:

• Independent peer trades — one wallet buying an NFT from another wallet because they want it. Secondary-market liquidity.
• Project mechanic flow — sales where the project's own wallets are on at least one side. Examples: a perpetual mint distributor selling fresh mints to participants, an instant-sell buyback bot purchasing NFTs back at a preset rebate price. Real HBAR moves on-chain, but the project is on one side of the trade — it represents primary-mint demand or operator market-making, not secondary peer demand.

The Organic % chip answers a specific question: of the reported volume in this window, how much happens between independent wallets? It is computed per collection per window (24h / 7d / 30d / all-time) and refreshed by the same poller jobs that maintain the wash-trade chip: an incremental re-score every 3 minutes for any collection that had a new sale in the last 30 minutes, plus a top-100 force-refresh every 30 minutes that re-validates the visible homepage rollups even when no new sales happened. Like the wash chip, the Organic % cascades from the period you're viewing to a wider window when the narrower one has fewer than 10 clean sales — the small ·7d / ·30d suffix on the chip indicates a fallback, and the tooltip explains how many clean sales the preferred window had.

Worked example — an illustrative 30-day window for a collection with an active perpetual mint mechanic:

• Reported marketplace volume: 16 M HBAR
• Sales involving the project's mint distributor or buyback bot on one side: ~13 M HBAR (~83%)
• Sales involving a detected frequent-trader cluster wallet on one side (but neither side a project wallet): ~1.4 M HBAR (~9%)
• Sales between independent wallets, no mechanic / no cluster: ~1.3 M HBAR (~8%)

Across an audit of the top 25 Hedera NFT collections by 30-day volume, the median Organic % was roughly 10%. Most reported marketplace volume on Hedera NFT collections involves project mechanic wallets on at least one side of the trade. This is not unique to any specific project; it reflects the structural reality of NFT markets where mint mechanics are the dominant activity pattern. The chip surfaces that reality so users can interpret raw volume figures accurately for whichever collection they are evaluating.

Every collection with at least 5 sales in the window lands in one of five descriptive bands. The band describes the volume mix, not the project's behaviour:

• ● Independent — over 80% of reported volume happens between independent wallets. No significant project mechanic or frequent-trader cluster footprint.
• ● Mostly Independent — 50% to 80% independent. Small share of mechanic / cluster activity.
• ● Mixed — 20% to 50% independent. Meaningful share of mechanic or cluster activity in the mix.
• ● Project-led — under 20% independent. Most reported volume involves the project's mint distributor, buyback bot, or a detected frequent-trader cluster on at least one side. This is common for collections running active mint mechanics — legitimate by design, just structurally different from secondary peer demand.
• ● Unrated — under 10 clean sales in the preferred window AND every wider fallback window. "Clean" means after excluding sales that touched the project's treasury, mint distributor, or detected buyback bot. Thin samples are noisy, so we refuse to characterize them.

What the chip shows: the Mechanic %% chip on the rankings table reports the share of reported HBAR volume where at least one side of the trade is a project-operated wallet (mint distributor, instant-sell buyback bot, or HTS treasury). A high number means the collection runs an active mint mechanic and most reported volume flows through that mechanic rather than secondary peer trading. It is not a warning sign — perpetual mints with instant-sell rebates are a legitimate product pattern. The metric just helps users interpret raw marketplace volume figures with the structural reality of the collection in mind. Mechanic + Cycler + Organic always sum to 100%% of reported volume.

We auto-detect three categories of project-operated wallet per collection. None of these are accusations — they are legitimate operational accounts that many Hedera NFT projects intentionally run as part of their product. Labelling them lets the chip tell users which trades involve the project itself on one side vs which trades are pure secondary peer activity.

• HTS treasury. The Hedera-native token treasury account for the collection, sourced from the Hedera mirror node. Mints, creator buybacks, and burns flow through this wallet. Always excluded from independent peer volume.
• Mint distributor (heuristic). A wallet that receives NFTs from the treasury at 0 HBAR and re-sells them to mint participants at the project's set price. Detected when a single seller accounts for ≥10% of priced sales AND holds <1,000 HBAR balance (consistent with a pass-through proxy used as the project's mint mechanic).
• Instant-sell buyback bot (heuristic). A wallet that buys NFTs back from holders, typically at a discount to the mint price as part of the project's "instant sell" or rarity-lottery rebate. Detected when a single buyer accounts for ≥10% of priced buys AND ≥50% of its trading activity is on a single collection.

When at least one mint distributor or buyback bot is detected, the chip popover surfaces a "Project mechanic wallets detected" note. The note describes what was observed; it does NOT characterize the project's motive or imply wrongdoing. A perpetual mint with an instant-sell rebate is a well-known product pattern; the metric just lets users interpret marketplace volume accurately given that pattern is active.

What the chip shows: the Cycler %% chip on the rankings table reports the share of reported HBAR volume where at least one side of the trade is a wallet in a detected frequent-trader cluster (heavy mint-mechanic users who also trade directly with other heavy mint-mechanic users). It is a statistical co-occurrence signal, not proof of coordination. A high number means the collection's peer market is dominated by a small group of inter-trading wallets; a low number means independent peer traders dominate. Cluster wallets could be a friend group, a power-user community, or a paid amplification network — the chip can't tell which from on-chain data alone. Mechanic + Cycler + Organic always sum to 100%% of reported volume.

A "frequent-trader cluster" is a behavioural pattern, not an accusation about individuals. A wallet enters the cluster when, in the same 30-day window, it (a) bought ≥10 NFTs from a project's mint distributor, (b) sold ≥10 NFTs to a buyback bot, AND (c) directly traded with at least two other wallets matching the same pattern. The first two conditions describe heavy use of the mint mechanic; the third is the statistical co-occurrence signal that distinguishes a coordinated cluster from independent heavy users.

The cluster label is statistical — it reflects observed trade-graph density, not proven coordination intent. A wallet meeting the threshold could be:

• A power-user degen who happens to also trade with other power-user degens
• A member of a friend group / trading club
• An organized cabal / paid amplification network
• Project-affiliated

We do not publish individual wallet labels in the public UI. The chip aggregates: "X% of volume involves wallets in a detected frequent-trader cluster." The underlying detection is internal tooling used for audit and review purposes only. If you believe your wallet is misclassified, the dispute path is documented in Disputing a rating below.

Any peer-to-peer sale where at least one cluster wallet sits on either side is excluded from the "independent peer" bucket. We split it into two sub-buckets internally:

• Cluster ↔ cluster direct trades — both sides are flagged. These are mathematically inherited into the wash-trade rating (DA-006) because two coordinated wallets trading with each other matches a wash-trade signature.
• Cluster ↔ outsider trades — one side is a cluster wallet, the other is unflagged. Counted as cluster-involved (so it doesn't inflate Organic %) but NOT counted as wash, since the outsider may be a genuine retail buyer.

The threshold is intentionally conservative — a wallet has to show heavy mint mechanic activity AND direct trades with multiple other matching wallets before it's included. Solo heavy gamblers without that coordination signal are NOT included.

The two metrics answer different questions and catch different failure modes. Reporting only one would leave a blind spot.

Both metrics ship with defamation-safety guardrails:

• Wash Pattern % is a coordinated-pattern detector. It has band-qualifier gates, concentration safeguards, and a dispute path. It flags patterns; it does not claim intent.
• Organic % is descriptive. It says "X% of reported volume happens between independent wallets". It does not say "the rest is fake" — mechanic flow is real on-chain HBAR; cluster activity is real trading. The metric describes structure.
• Neither chip publicly identifies individual wallets as bad actors. Disputes are handled through Discord and email — see Disputing a rating for the process.

The chip is honest about its limits. Specifically:

• We don't prove intent. A project can show "Project-led" volume composition because it runs an actively-used mint mechanic and that activity is the dominant trade pattern. That doesn't mean the project is doing anything wrong. The chip describes; it does not accuse.
• We don't prove cluster wallets are coordinated. The cluster label is statistical — it flags wallets whose trade-graph density and behaviour match a coordinated-cluster signature. The matching wallets might be friends, an alpha trading group, organic power users, or something more concerning. From chain data alone, we cannot distinguish.
• We don't prove off-chain coordination. If a project is paying participants off-chain (Discord rewards, OTC deals, fiat payments), that is invisible to chain data. We catch the on-chain pattern only.
• We don't name individuals publicly. Even with high-confidence cluster detection, we never publish wallet-level labels in the marketplace UI. The chip is always aggregate.
• Heuristics can mis-classify. A legitimate single-collection whale might trip the buyback-bot heuristic. A real friend-group of degens might trip the cluster heuristic. Open a dispute if you believe a wallet was misclassified — we will review and update.
• Perpetual mints and rarity lotteries are legitimate. Many Hedera projects ship them transparently. Labelling that activity as "project mechanic" doesn't accuse the project — it just lets buyers and sellers read marketplace volume figures accurately given that pattern is active.

Dispute path — same as the wash chip: open a ticket in Discord or email dreambay@possestudios.io. Include the token id and any specific wallet you believe is misclassified. We respond promptly and adjust the heuristics or allowlist when appropriate.

The full fee structure across DreamBay surfaces:

XRP NFT fees

DreamBay's XRP marketplace mirrors the Hedera fee model exactly. Every DreamBay XRP listing is broker-locked to our operator wallet (matches xrp.cafe's pattern) — the lock keeps fulfillment on DreamBay so we can mediate / refund / track, but the buyer-paid marketplace fee is still 0% on direct buy.

The full creator-facing fee table with detailed descriptions lives in the Creator Dashboard → Fees tab.

DreamBay smart contracts and platform code have been through multiple audit passes:

• Hashlock — third-party audit; badge in the header links to the report
• MemeBay V2 security audit — multi-tool replication of a CertiK-level audit on the current MemeBay contracts (Slither, Foundry fuzz, mainnet-fork, three independent adversarial reviewers, 2026 exploit pattern catalog cross-check)
• Auth + DreamCast separate audits — wallet-signature verification, replay protection, atomic swap correctness, RNG hardening

Found something? Email dreambay@possestudios.io with "security" in the subject. We pay bounties.

What's being built right now (subject to community input):

• XRP Ledger — full feature parity for XRPL NFTs (browse, buy, sell, auctions). Switch via the chain selector.
• Off-chain marketplace — Web2-flavoured digital goods marketplace alongside Web3
• Multi-chain DreamCast — pools that span multiple chains, unified leaderboards
• Creator multisig admin — split admin keys for higher-risk collections

The fastest path is Discord — open a ticket and the team responds in real time most days.

• Discord: discord.gg/possestudios
• Email: dreambay@possestudios.io
• X / Twitter: @DreamBayIO
• Legal: Terms of Service