Compiled 2026-05-19. Cross-chain research on NFT interaction mechanics in Web3 games (Ethereum, Polygon, Ronin, Immutable, Solana, Avalanche, Base, TON, etc.) plus a prioritised list of patterns for Cube Worlds.
Companion docs: MARKET_RESEARCH.md (Telegram crypto-game landscape), TOKEN_INTERACTIONS.md (token economy patterns and Cube Worlds sink plan).
⚠️ Pre-pivot research (2026-06-14 note). Point-in-time NFT-mechanics catalog. Since it was compiled, Cube Worlds shipped NFT collections for the entry pass plus castle / hero / equipment items (all DB-canonical with deploy-gated on-chain mint, gated behind*_COLLECTION_ADDRESS), and made $CUBE DB-only (no on-chain jetton). The pattern catalog still applies; map any "jetton" reference to a DB ledger. Ground truth:../CLAUDE.md.
The Web3 gaming industry converged on a small set of NFT verbs that actually work. Fusion / burn-to-upgrade is the single most universal mechanic — every long-lived game has it (Gods Unchained Forge, Splinterlands BCX, Illuvium 3:1, Guild of Guardians Altar, Champions Tactics Forge, Shrapnel fragments). The 2025–2026 trend is decisively PFP → functional NFTs with on-chain progression, oracle-driven dynamic state, and bonded in-game wrappers that decouple emission from open-market token price (Axie's bAXS is the canonical reform).
A reference document covering the design space of what players can do with NFTs in modern blockchain games. Each entry covers the verb, on/off-chain effect, token-economy role, best example, pros/cons, and enabling standards.
Verb: Two parent NFTs are paired; after a cooldown, a child NFT is minted with traits derived from parents (often pseudo-random Mendelian inheritance).
Effect: New NFT minted to the breeder's wallet; both parents get incremented "breed count" stored on-chain, capping their fertility.
Token role: Heavy sink. Each breed consumes a utility token plus the governance/scarce token. In Axie Infinity (2025), each breed costs 0.5 AXS plus 300–5,100 SLP depending on breed count (max 7 breeds, 5-day egg maturation). This burns supply on both axes.
Best example: Axie Infinity — the canonical breeding economy. CryptoKitties pioneered the pattern in 2017 but collapsed under oversupply because there was no cap on breeds and no sinks for the resulting kittens.
Pros: Endless content from finite seed assets; emergent rarity from trait combinations; deep secondary-market dynamics.
Cons / failure modes: Without strict supply throttling and meaningful sinks for offspring, oversupply destroys floor prices. CryptoKitties saw single-kitten value collapse precisely because breeding compounded supply faster than demand. Requires fierce balance work on cooldowns, breed-count caps, and offspring utility.
Standard: ERC-721 with custom breeding contract holding lineage state. No dedicated EIP.
Sources: Axie Breeding Whitepaper, Frontiers — CryptoKitties Rise and Fall
Verb: Burn N copies of an NFT (and optionally pay a token) to mint 1 higher-tier NFT.
Effect: Permanent supply reduction at the input tier; new mint at output tier. On-chain burn is verifiable.
Token role: Strong supply sink for low-tier assets, mild token sink for the fusion fee. Gods Unchained Shine Fusing: 5 Meteorite → 1 Shadow → 1 Gold → 1 Diamond, consuming Flux (earned by playing Ranked) plus $GODS. Splinterlands combines two copies into a single card with higher BCX, retiring the merged Card ID forever; burn value is also the card's "collection power."
Best example: Gods Unchained's Forge (stair-step rarity model) and Splinterlands (BCX leveling tied to DEC burn value).
Pros: Self-balancing economy — supply reduces automatically as players chase upgrades; clear, satisfying progression loop; minimal engineering (one burn-and-mint contract function).
Cons: If output tier has insufficient additional utility, players hoard inputs and the sink dries up. Whales can monopolize top tiers.
Standard: ERC-721 or ERC-1155 with a burn function; ERC-1155 is more efficient when many input copies exist.
Sources: Gods Unchained Shine Fusing, Splinterlands Combining Cards
Verb: Use, train, or hold an NFT over time; its on-chain attributes and visual metadata change in response.
Effect: Metadata URI is dynamic (often via a tokenURI() that reads on-chain state, or via off-chain renderer keyed to chain events). Stats, art, or both update.
Token role: Often gated behind a token-paid "evolution" action; sometimes free (XP earned by play). The token sink is the evolution fee.
Best example: Sorare's seasonal cards (real-world player stats refreshed via Chainlink oracles), and Pokemon-style evolution chains in many gacha-NFT games.
Pros: Solves the "JPEG that never changes" problem; ties NFT value to play time, deepening retention.
Cons: Front-running and farming risk (bots level NFTs cheaply); metadata stability concerns for marketplaces displaying stale snapshots.
Standard: ERC-721 with on-chain metadata, or ERC-721 + Chainlink Functions/Automation for oracle-driven updates. ERC-4906 (MetadataUpdate event) signals marketplaces to refresh.
Sources: Chainlink Dynamic NFT Use Cases
Verb: Slot child NFTs (weapons, armor, pets) into a parent NFT (character, mech), with the parent visually reflecting equipped children.
Effect: Parent NFT owns the children on-chain; transferring the parent transfers the loadout atomically.
Token role: Indirect — drives demand for both parents and a long tail of compatible children; marketplace fees on each item.
Best example: Wreck League mechs (RMRK-standard equippable parts), and BAYC + Token Bound Account experiments where the ape owns its sidekicks and gear.
Pros: Massive content surface from a small set of base characters; secondary markets for every slot; cross-collection collaborations become trivial.
Cons: Significant engineering — must define equip rules, slot compatibility, render pipeline; marketplace UX for nested NFTs is still immature.
Standard: ERC-7401 (Parent-Governed Nestable, evolved from ERC-6059) for nesting; ERC-5773 for multi-asset; ERC-6220 for equippable. Alternative: ERC-6551 Token-Bound Accounts give each NFT a smart-contract wallet that can hold any other tokens — more flexible than ERC-998, which only held ERC-20/721.
Sources: RMRK ERC-7401 Docs, ERC-6551 vs ERC-998
Verb: Owner delegates use rights (not ownership) of an NFT to a renter for a fixed period; renter plays and earns, owner takes a cut.
Effect: A separate user address gains in-game permissions while owner retains transfer rights and recovers full control on expiry.
Token role: Rental fees create yield for owners and lower the entry barrier for new players, expanding the active user base without diluting NFT supply.
Best example: Axie scholarship program (off-chain, custodial, pre-standard) onboarded millions in 2021. Fanton (TON fantasy football) bakes renting into the protocol — Legendary/Epic/Rare cards rent for matchdays and auto-return.
Pros: Solves the "high entry cost" complaint; converts dormant NFTs into yield-generating assets; powerful funnel for new players.
Cons: Trust/custodial risk in pre-EIP-4907 implementations; renter exploits (asset abuse) need rate limits.
Standard: ERC-4907 adds a user role with expires field, finalized June 2022. ERC-5006 extends to ERC-1155. TON has no formal rental standard yet — implementations are app-level.
Sources: EIP-4907, Fanton TON Innovators
Verb: Lock an NFT into a vault and mint fungible tokens representing shares; many wallets co-own one asset.
Effect: ERC-20 (or ERC-1155) shares trade freely; buyout auctions reassemble full ownership.
Token role: Liquidity unlock for high-value NFTs; secondary-market trading of fractions generates fees.
Best example: Fractional.art (later Tessera) until it shut down May 2023 amid bear-market conditions and weak product-market fit.
Pros: Democratizes blue-chip ownership; price discovery for illiquid assets.
Cons: Regulatory grey zone (often resembles a security); Tessera's shutdown shows the model struggles outside speculative peaks. Rarely useful in games specifically.
Standard: Custom vault contracts producing ERC-20 or ERC-1155 shares. No accepted EIP.
Sources: Tessera Shutdown
Verb: Receive a non-transferable NFT for an achievement, attestation, or identity claim.
Effect: Token is permanently bound to the receiving wallet; transfers revert. Acts as an on-chain credential.
Token role: Not directly economic — but unlocks gated drops, whitelists, and reputation systems. Cheap to mint, zero supply pressure.
Best example: Galxe OATs (on-chain achievement tokens) issue millions of credentials; POAP uses transferable proofs for attendance (30M+ minted) but is moving SBT-flavored variants. On TON, First Force minted 10,000 SBTs for the TAC community.
Pros: Cheap, fraud-resistant, perfect for retention/loyalty loops; the simplest pattern to ship.
Cons: No secondary market = no speculator pull; recoverability problematic if user loses wallet.
Standard: ERC-5192 (Minimal Soulbound NFTs, finalized September 2022) — adds a locked() interface to ERC-721. On TON: SBT collections built atop TEP-62 with transfer disabled.
Sources: EIP-5192, Galxe SBT Build
Verb: Own a geographic plot; build on it, generate resources, earn yield from neighbors or sharecroppers.
Effect: Plot NFT acts as a location key; on-chain it stores building state, yield multipliers, traits. Adjacent plots can synergize.
Token role: Plots are premium-tier sinks (high mint cost) and yield-bearing assets that print game tokens — must be carefully throttled or they become inflation engines.
Best example: Pixels on Ronin (5,000 NFT lands with Coop/Windmill/Tree-density traits; owners earn a share of crops grown by renters/Specks). Otherside's Otherdeeds (200K plots across 5 regions, harvestable Anima/Ore/Shard/Root). Big Time's SPACE attaches Armory/Forge/TimeWarden utility NFTs to plots.
Pros: Strongest retention mechanic in Web3 gaming — players treat plots like real estate; daily-login behavior; powerful social/neighbor mechanics.
Cons: Inflation risk if yield isn't capped; "land barons" hoard supply; engineering cost is highest of any pattern (map data, building state, multiplayer sync).
Standard: ERC-721 with grid coordinate metadata. TON has no canonical land standard; would be a TEP-62 collection.
Sources: Pixels Land Guide, Big Time SPACE Wiki
Verb: NFT changes state based on external real-world data — sports results, weather, prices, AI outputs.
Effect: Smart contract calls an oracle (Chainlink Functions, Automation) and rewrites token state/metadata. Marketplaces re-fetch via ERC-4906.
Token role: Indirect — drives engagement spikes around real-world events; no direct sink.
Best example: Sorare's seasonal cards refresh stats from real soccer/NBA games via Chainlink-sourced data feeds (TheRundown, SportsMonks, SportsDataIO).
Pros: Strong narrative hook (your card becomes a living asset); event-driven engagement spikes.
Cons: Oracle costs and reliability; centralization risk in the data feed; complex testing.
Standard: ERC-721/1155 + Chainlink (Functions, Automation, Data Feeds). VRF for randomness.
Sources: Chainlink Sports Data
Verb: Lock NFT in a contract for a period; earn fungible token yield.
Effect: NFT is escrowed (or flagged as staked); contract emits rewards proportional to time/tier.
Token role: Distributes tokens (inflation), but in exchange creates lockup — removing NFTs from secondary market also tightens floor.
Best example: ApeStake.io — BAYC pool allowed staking up to 10,094 APE per ape; the highest-yield tier in the protocol. Pudgy Penguins evolved staking into utility/IP-licensing rights rather than token emissions. MOBOX (MOMO NFTs on BNB Chain) emits MBOX per staked MOMO based on hash power.
Pros: Simple to ship; locks up float and tightens floor; aligns long-term holders.
Cons: Pure emissions staking is inflationary and Ponzi-adjacent — when emissions stop, demand collapses. Better used in tandem with sinks.
Standard: Custom staking contract on top of ERC-721/1155.
Sources: ApeStake.io
Verb: Combine multiple NFT inputs + token costs according to a recipe to produce a new NFT.
Effect: Inputs burned (or returned), token burned, output minted. Recipes are usually on-chain.
Token role: Crafting is a primary token sink — the engine of healthy game economies. Big Time's Forge/Armory burn $BIGTIME to craft Cosmetic Collectibles, then those Cosmetics get sold to players who lack utility NFTs, recycling demand.
Best example: Big Time — the most engineered crafting chain in Web3 (Time Wardens craft Hourglasses → Suppliers generate $BIGTIME → Forge owners craft cosmetics → players buy them). Diablo-like loot-fusion games on Immutable adopt similar loops.
Pros: Best-in-class token sink design; deep economic interdependence between player roles.
Cons: Requires multi-token, multi-NFT engineering; balancing recipes is an ongoing live-ops job.
Standard: ERC-1155 ideal for stackable inputs; custom recipe registry contract.
Sources: Big Time Token Sinks
Verb: Buy/earn purely visual items; equip them on an avatar.
Effect: Avatar metadata references the wearable's contract+ID; render pipelines fetch and composite.
Token role: Pure consumer-discretionary spend; no stat impact = no pay-to-win complaints. Highest-margin category in F2P games translates directly.
Best example: Decentraland wearables (ERC-721/1155 with the "Linked Wearables" feature, letting external collections — your BAYC, your Pudgy — appear as in-world wearables after DAO approval).
Pros: Pay-to-look, not pay-to-win; cross-game potential; AAA-game-tier monetization model.
Cons: Requires asset pipeline; visual quality bar is high.
Standard: ERC-1155 for stackable cosmetics; Decentraland uses a registry pattern for cross-collection wearables.
Sources: Decentraland Linked Wearables
Verb: Hold NFT → vote on proposals; one NFT = one vote (or delegated).
Effect: NFT contract integrates with a governor (Compound-style); votes attach to current holder, transfer on resale.
Token role: No direct sink — but governance utility supports floor price.
Best example: Nouns DAO — one Noun auctioned every 24 hours forever, 100% proceeds to a community treasury, one Noun = one vote, delegatable but not separable from the NFT. The cleanest implementation of NFT-as-membership.
Pros: Aligns holders with project; treasury accumulation; transferable membership.
Cons: Plutocracy (whales dominate); voter apathy; legal ambiguity around DAO liability.
Standard: ERC-721 + ERC-721Votes extension; Compound Bravo governor fork.
Sources: Nouns Governance Explained
Verb: Hold NFT → unlock event, future drop, whitelist, or token allocation.
Effect: Gating logic reads NFT ownership; pass may be consumed (burned) or persistent.
Token role: Pass sales front-load revenue; consumed passes act as one-shot sinks; persistent passes are loyalty anchors.
Best example: Catizen Airdrop Pass — 90-day seasons distribute 19% of CATI supply through pass-gated tasks. Notcoin and X Empire minted NFT vouchers on TON to let users speculate on pre-listing token allocations; the X Empire collection minted 570,000 vouchers tradable on Getgems. Adidas Originals' Into the Metaverse pass unlocked physical and virtual drops.
Pros: Simplest ROI pattern after SBTs; immediate revenue; powerful retention via FOMO on future drops.
Cons: Pass utility must keep delivering or floor collapses post-event.
Standard: ERC-721 (collectible pass), ERC-1155 (stackable tickets); on TON, NFT vouchers via TEP-62.
Sources: Catizen Airdrop Pass, X Empire NFT Vouchers
Verb: Complete an in-game or off-chain task → claim a proof NFT.
Effect: NFT minted (free or gas-only) to the completer's wallet. May be transferable (POAP-style) or soulbound (achievement-style).
Token role: Discovery and retention tool, not a direct sink. Frequently used as eligibility input for later airdrops.
Best example: POAP (30M+ event-attendance NFTs minted) and Galxe (campaign-driven OATs that brands stack into reputation graphs).
Pros: Cheapest possible pattern; gigantic retention/marketing impact; integrates trivially with quest systems.
Cons: Wallet bloat for users; the line between "credential" and "spam" is thin.
Standard: ERC-721 (transferable, POAP-style) or ERC-5192 (soulbound, Galxe Passport-style).
Sources: Galxe + NodeReal SBT
Verb: Buy a sealed crate NFT; open it to reveal probabilistically distributed contents (other NFTs and/or tokens).
Effect: Crate burned; one or more output NFTs minted using verifiable randomness.
Token role: Big primary-sale revenue spike; pure sink if crates cost tokens to open.
Best example: Star Atlas Meta-Poster loot boxes dropped tiered NFT badges; many Solana/Magic Eden games use similar reveal mechanics. Sorai-style series across Asian markets.
Pros: Strongest monetization spike in gaming history (loot-box psychology); fits short attention spans.
Cons: Regulatory risk — several jurisdictions (Belgium, Netherlands) treat loot boxes as gambling. Requires verifiable randomness (Chainlink VRF) to avoid manipulation.
Standard: ERC-1155 for crates + Chainlink VRF for randomness.
Sources: Star Atlas Loot Box
Verb: List, auction, and bid on NFTs without leaving the game client.
Effect: Game embeds a marketplace contract or aggregates an existing one; royalties and fees flow back to the studio.
Token role: Transaction-fee tax on every trade is one of the most reliable long-term revenue streams (Sky Mavis charges 2% + 0.5% Ronin treasury + 0%–10% creator royalty on Mavis Market).
Best example: Mavis Market on Ronin — fully native to Sky Mavis games, multi-currency (RON/AXS/SLP/ETH auto-converted), with treasury/validator share.
Pros: Captures secondary-market value (which is most of NFT trading volume); deepens engagement; analytics goldmine.
Cons: Royalty enforcement is collapsing on many chains; users sometimes prefer OpenSea/Magic Eden liquidity.
Standard: Seaport (OpenSea), Reservoir aggregation, or custom listings contract.
Sources: Mavis Market Fees
Verb: Use the same NFT in multiple games or platforms.
Effect: Game B reads the contract address from Game A and grants in-engine usage (visual, stat bonus, access).
Token role: Indirect — shared NFTs become marketing surface for partner games.
Best example: Loot (for Adventurers) — a JSON-only Ethereum NFT that spawned hundreds of derivative games and renderers. Ready Player Me avatars work in 3,000+ apps. Yuga's M3taloot initiative ties BAYC/MAYC into wider metaverse experiences; ApeChain consolidates them on a Yuga-controlled L2.
Pros: Massive moat for the originating collection; cross-promotion.
Cons: Coordination problem — no studio wants to do free work integrating someone else's NFT; usually only happens when one collection is too big to ignore (BAYC, Pudgy).
Standard: No standard required for read-only reference; ERC-6551 enables an NFT to carry assets across games.
Sources: Loot Project, Yuga + M3taloot
Verb: Burn an NFT to claim a physical good, IRL benefit, or different on-chain asset.
Effect: NFT permanently destroyed on-chain; off-chain fulfillment system ships the physical item.
Token role: Removes supply at the moment of redemption — supports floor; turns digital asset into single-use coupon.
Best example: RTFKT × Nike Cryptokicks iRL — owners "forged" digital sneakers into physical pairs (19,000-pair series, NFC-chipped, NFT-linked). Nike .Swoosh, Adidas Originals, and luxury houses (Tiffany, LVMH) followed similar patterns.
Pros: Bridge between digital and physical brand value; very strong PR moment; clean sink.
Cons: Logistics is the hard part — RTFKT failed to ship to non-US holders, generating significant blowback. Customs, sizing, fraud all become problems the studio must solve.
Standard: ERC-721 with burn; off-chain KYC/shipping system; sometimes NFC for proof-of-redemption.
Sources: Nike + RTFKT Phygital
Verb: NFT's visual output is regenerated by an AI model based on user actions, community votes, or external data.
Effect: Metadata URI points to an endpoint that re-renders via an AI pipeline; on-chain state seeds the prompt.
Token role: Often tied to a curation token (stake to vote on outputs).
Best example: Botto (decentralized autonomous artist) — generates ~8,000 candidate images weekly, a "taste model" narrows to 350, BottoDAO token holders vote, winner is auctioned on SuperRare. Botto has sold $5M+ in works since 2021. Art Blocks remains the gold standard for generative on-demand minting with deterministic seeds.
Pros: Endless content for trivial cost; strongest narrative for an AI-native game economy; community curation is engaging.
Cons: Provenance/copyright legally unsettled; model deprecation breaks future renders; off-chain compute centralization.
Standard: ERC-721 with dynamic tokenURI; on-chain seed + off-chain renderer (Art Blocks style); IPFS pinning for stability.
Sources: Botto Decentralized Artist
- Wrapping / bridging — wrap a TON NFT to an EVM chain to access EVM tooling, or vice versa. Pattern: lock-and-mint on origin, burn-and-unlock on destination.
- Royalty splitter NFTs — NFT represents a share of future revenue (0xSplits, Royal music).
- Time-locked / vested NFTs — NFT becomes transferable after a date or condition.
- Insurance / hedging NFTs — Nexus Mutual-style coverage as a tradable NFT.
- Subscription NFTs (ERC-5643) — NFT with expiry-based access (recurring billing on-chain).
- Lottery / raffle NFTs — drawing rights as NFTs; verifiable randomness via VRF.
A look at how marquee Web3 game franchises across all chains combine these primitives.
Verbs: mint, draft, play, fuse (forge), trade, stake.
The signature interaction is Fusing in The Forge: two identical Plain cards burn together to mint a Meteorite, then Shadow, Gold, Diamond — four upgrade tiers, each fusion costing $GODS plus Flux earned by playing weekend ranked games. Costs scale by rarity (Common 0.1 GODS + 20 Flux → Legendary 1.5 GODS + 200 Flux). Shine Fusing later allowed mid-tier shines without leaving the chain. The 2025 migration to Immutable zkEVM introduced an upgraded Forge and refreshed staking rewards.
Why it works: Forge is a constant token sink tied to skill (Flux only drops to ranked winners), and the upgrade path is visibly cosmetic and signals dedication. Trading the resulting NFT is optional — most stay in collection. The flux gate prevents whales from one-shotting full Diamond decks.
Failure mode: Until zkEVM migration, on-chain frictions and high $GODS cost gated casuals; new-card-set inflation periodically deflated older shines.
Verbs: mint (drop), draft lineups, score multiplier accrual, trade, rent (lend), craft via Essence. Cards are tiered Limited (≤1000), Rare (≤100), Super Rare (≤10), Unique (1). Each card carries season-bound XP and "multipliers" that grow with use, making the card itself a progressing asset. Sorare's 2025 introduction of Essence and the Card Factory added a crafting verb: combine cards to manufacture or improve assets, deepening sinks beyond just primary drops. Why it works: Rarity ties to competitive tournament access, not just resale. Rentals let collectors monetise idle inventory and new players field competitive lineups without buying-in. Failure mode: Seasonal multiplier reset still bleeds value off older cards; reliance on real-world player performance means injuries/transfers can crash a card overnight.
Verbs: mint via packs, combine (BCX merge), level-up, rent (daily/seasonal), delegate, burn for DEC. Cards track Base Card Experience: combining two identical cards burns one and consolidates BCX into the survivor, raising level and unlocking new abilities. The rental market — paid in DEC, daily or seasonal — is one of the most mature in Web3 gaming, with secondary tools like PeakMonsters surfacing thin-market level breakpoints. Why it works: BCX combining creates meaningful supply destruction (one NFT literally disappears per merge) and the rental layer turns idle cards into yield without selling. Burning cards back into DEC is the safety-valve sink. Failure mode: DEC inflation in 2022-23 mirrored Axie's SLP issue; rental yields collapsed when reward pools shrank.
Verbs: mint, deck-build, play matches, energy-bank between turns, use Keys (perks), earn $PRIME.
Parallel's mechanical novelty is the Energy Banking System — unspent energy carries between turns, so the NFT deck composition rewards forward planning. Keys (Overclock, Multifold) are NFT perks that boost PRIME earnings or amplify card effects. Starter progression in 2025 was reworked so new players unlock Paragons and faction decks without holding NFTs.
Why it works: Free-to-start with NFTs as power/yield modifiers (not paywall). PRIME accrues via competitive play, not idle.
Failure mode: Like all TCGs, deck rotations destroy old-card value if not handled carefully.
Verbs: mint, breed (SLP + AXS burn), battle, scholarship/rent, stake, body-part swap, land farm. The canonical case study. Axie's breeding burns both SLP (earned by winning) and AXS (governance). Scholarships pioneered "play-to-earn" delegation. The collapse, well-documented, came from faucets without sinks — scholars farmed SLP, breeders didn't burn enough, and population growth meant new players bought existing Axies rather than commissioning births (Yu-kai Chou analysis). SLP fell 95%+. 2026 reform: Sky Mavis introduced bAXS (Bonded AXS) — a non-transferable in-game wrapper of AXS used for breeding and staking — and is replacing Homeland with Terrariums, a new land-yield model. This is an explicit attempt to de-link in-game reward emission from the open market, a major lesson the whole industry is now applying. Why breeding worked at first: It's the cleanest NFT verb — you literally make new NFTs from old ones, with genetic inheritance creating long-tail desirability. Why it failed: Unlimited supply with weak demand-side sinks. Breeding cost was denominated in inflating tokens. No friction.
Verbs: mint, breed (sire), trade. The original NFT-breeding template. Each kitty has 256-bit "genes"; pairing produces offspring with inherited traits. Two failure modes hardened into industry lore: (1) market saturation — every kitty is a breeding factory, so supply explodes; (2) Ethereum L1 gas costs routinely exceeded kitty floor price, killing casual play. Lesson the genre kept: breeding requires either a consumable rare ingredient (Axie's later "AxieEgg" attempts), generational caps, or expiry — otherwise it's a perpetual supply faucet.
Verbs: farm, mint pets via vending machines, hatch (potion + lab), level skills, build on owned plots, spend energy.
Land NFTs are scarce (5,000 plots) and gate access to higher resource yields. Pets are companion NFTs minted in-game, traded on Ronin marketplace. The token model evolved: in early 2025 Pixels phased out $BERRY entirely, replacing it with off-chain "Coins" and reserving $PIXEL for high-tier verbs (NFT mint, guild join, premium boosts).
Why the reform matters: Pixels deliberately moved low-value loops off-chain and kept blockchain for high-value, low-frequency actions — directly addressing the CryptoKitties gas problem and the Axie inflation problem.
Verbs: craft cosmetics, upgrade weapons/armor, recharge Hourglass NFTs, expand personal metaverse with SPACE, refine Cosmetic Shards.
Big Time pioneered a "Cosmetic-Based Economy" — gameplay items themselves are not NFTs, but the cosmetic skins and Hourglasses (yield-bearing NFTs that boost $BIGTIME earnings) are. Workshops (Forge / Armory / Time Warden) require SPACE NFTs to operate, creating a clean B2B layer between cosmetics creators and consumers (Big Time crafting guide).
Why it works: Decouples cosmetics from balance — no pay-to-win — while still giving NFTs functional yield via Hourglasses.
Verbs: capture Illuvials, level them, fuse 3 max-level into 1, mine land for fuel, equip Arena weapons/suits, bet in Leviathan Arena. The fusion verb here is more demanding than Gods Unchained: three fully-leveled identical Illuvials merge into one stronger Illuvial. Land owners generate fuel that both demand-side games (Arena, Beyond) and supply-side games (Zero, Overworld) consume. Patch 1.1.0 added Gauntlet Leviathan Mode which stakes the player's actual NFT roster against others. Why it works: Multi-tier supply control (3:1 burn), and the fuel economy ties land to combat — land isn't passive yield, it powers the rest of the world.
Verbs: mint ship/crew/land NFTs, enlist ships in SCORE faction missions, claim land Stakes, mine ore, refine, craft, faction-align. Ships sit idle if held; staking them into SCORE deploys them to a faction's Council of Peace zone, accruing rewards while burning fuel/ammo/food consumables. Land Stakes anchor planetary build loops. Why it works (in theory): Every NFT has active utility cost — fuel and food drain — so holders can't passive-farm forever. Why it struggles: Long development cycle and limited live gameplay mean most of the economy still revolves around speculation rather than friction-creating play.
Verbs: battle royale play, collect rare drops as NFTs, earn season-pass NFTs/AVAX rewards. Notably hides its blockchain layer — players engage with a normal Epic Games shooter. The Avalanche Battle Pass (Ava Labs + Playfull + Magic Eden) functions as a cross-game soulbound-ish progression layer that mints NFT rewards for playing any participating game (Decrypt coverage). Why it works: Web2 UX + opt-in Web3 inventory. Identity is contextual rather than mandatory.
Verbs: extract resources, craft weapon skins from fragments, trade on SHRAP marketplace, stake SHRAP, mod. Skins are crafted by combining three same-rarity fragments — an Axie-style 3:1 burn but for cosmetics, not creatures. Marketplace pricing sits at $3-13 per skin, deliberately accessible. Why it works: Cosmetic NFTs from in-game drops have playtime backing, not fiat. Modding tools encourage UGC.
Verbs: draft 3-champion team, combine 2 Champions in The Forge → new Champion with merged DNA, use Ethereal (non-NFT, time-limited) champions for try-before-buy.
Forge crafting costs Gold or $OAS. Ethereal Champions are explicitly non-tradable, expiring rentals — a clean F2P on-ramp that doesn't bloat NFT supply.
Why it works: AAA brand legitimacy + smart separation of NFT scarcity (75,000 minted champions) from gameplay accessibility (Ethereals).
Verbs: build base on Land NFT, raid for RUBY, stake RUBY for MAVIA, rent land, scholarship via partnerships. Land rarity directly multiplies earning caps: Common +200%, Rare +257%, Legendary +350% earnings, with daily-cap multipliers 300/450/700% (ChainPlay analysis). Phase 2 added Ruby staking → MAVIA, and PvP battles. Land rental and free-share "partnerships" formalised the scholarship model post-Axie lessons. Why it works: Tiered yield + explicit daily caps prevent runaway inflation.
Verbs: Ascend Guardian → Radiant NFT (spend Radiant Seals), Enlighten with duplicates (Radiance up), sacrifice at the Altar (burn NFTs for rare new ones), guild raids. The Altar of Sacrifice is the explicit deflationary verb — burn multiple NFTs to mint one rarer. Radiance is on-chain progression baked into the asset itself (a dynamic-NFT pattern). Migration from ImmutableX → zkEVM completed in 2024-25. Why it works: Ascension creates a clean F2P → P2O upgrade path (you must own the in-game character first, then convert to NFT).
Verbs: assemble mech from 10 NFT parts (5 body + 5 attachments), set bonuses, swap parts, trade unused parts. The signature mechanic is composability: every mech is 1 of 1.5 quadrillion possible combinations from individually-traded body/arm/leg/head/shoulder/knee/satellite/accessory NFTs. Same-set equips unlock passive bonuses (Viper: +15% crit; Sentinel: +20% shield). Why it works: True modular NFTs — secondary markets form per part, not per finished asset. Players upgrade incrementally. Why it nearly didn't: Initial deployment on Ethereum mainnet meant $30+ gas just to open a parts box. Composability without cheap chain is unworkable.
Verbs: mint baby pet, feed/care/grow, battle in arena for TAMA, climb leaderboard, breed.
Tamadoge added NFT breeding post-launch. Despite early hype it now sits 99.8% below ATH ($0.00034 in May 2026) — a textbook example of speculation-without-retention. Pet Simulator X (Roblox) faced a player revolt when NFTs were grafted onto a Web2 game post-hoc.
Lesson: Pet sims with NFTs work only when the care loop is intrinsically rewarding; if the NFT is just a wrapper for tradable yield, players churn the moment token price dips.
- Burn-to-upgrade / fusion — Gods Unchained Forge, Splinterlands BCX, Illuvium 3:1, Guild of Guardians Altar, Shrapnel fragments, Champions Tactics Forge. Every long-lived game has this. It's the strongest, most legible NFT sink.
- Rental/scholarship — Splinterlands, Axie, Sorare, Mavia, Pixels. Idle inventory becomes yield; new players access end-game without capital.
- Tiered scarcity with utility gating — Sorare's Unique/SR/R/L, Mavia land tiers, Pixels' 5,000 plots. Rarity gates something a player can do, not just bragging rights.
- Token sinks tied to play, not idle holding — Star Atlas fuel, Big Time Hourglass recharge, Mavia daily caps.
- Composable / modular NFTs — Wreck League's per-part economy is the clearest signal. Most games still ship monolithic NFTs.
- Bonded / non-transferable wrappers — Axie's 2026 bAXS model decouples in-game emission from open-market token price. This is likely to spread; it's how you keep a token usable for breeding without it becoming an extraction vehicle.
- Cosmetic-economy separation — Big Time and Shrapnel keep balance off-chain, cosmetics on-chain. Sidesteps pay-to-win and gas-cost-vs-floor-price death spirals.
- Try-before-buy non-NFT analogues — Champions Tactics' Ethereal Champions, Parallel's free Paragons. The NFT becomes an upgrade, not a paywall.
- Soulbound progression — Avalanche Battle Pass, achievement SBTs. Identity and achievements as non-transferable tokens, leaving only intentional assets tradable.
- PFP → functional NFTs. Value is increasingly anchored to utility and in-game relevance, not speculative scarcity. Cosmetic PFP collections are out; assets with verbs are in.
- Dynamic NFTs / on-chain progression. Metadata that mutates with play — Guild of Guardians' Radiance, Sorare's growing multipliers, swords that level up. The NFT is the save file.
- Hidden blockchain UX. Off The Grid, Champions Tactics, Pixels' Coins-for-low-tier model all minimise wallet/gas friction for casuals. Chain is a backend, not a gate.
- Hybrid economies. Surviving games balance earn with crafting/upgrade/cosmetic spend — Axie's bAXS reform is the canonical case.
- Soulbound identity layers. Achievement and reputation as non-transferable tokens are an alternative to leaderboard-as-database — and naturally resistant to bot/scholar farming.
- Faucet without sink (Axie SLP, CryptoKitties supply, Tamadoge).
- In-game token = open-market token, 1:1 — extraction always wins. Bonded wrappers fix this.
- NFT = paywall, not upgrade — kills onboarding (early Wreck League, original Axie).
- Cosmetic-only with no gameplay verb — PFPs in disguise; players churn when speculation cools.
- Gas > floor price — CryptoKitties' fatal mistake; only L2/subnets/app-chains avoid it.
Ranked by implementation simplicity × retention/economic impact. Assumes limited engineering, no AAA art budget, mobile/Telegram-style game.
- Soulbound achievement NFTs (Patterns 7/15). Cheapest possible pattern, immediate retention lift. One contract, one mint function, infinite badge variants. ERC-5192 or TON SBT.
- Access pass NFTs (14). Simplest revenue mechanism. Sell a capped pass that unlocks future drops/airdrops. Catizen pattern works on any chain.
- Burn-to-upgrade fusion (2). One contract function (
burnAndMint), instantly creates a meaningful sink. Doesn't require any new game mechanics. - NFT staking with capped emissions (10). Lock-up improves floor; one contract; simple math. Avoid pure-emissions Ponzi by capping rewards or sourcing them from sinks.
- In-game marketplace with fee (17). Even a 2–5% trade fee compounds; can be a Seaport hook rather than a full custom build.
- Gacha crates with VRF (16). Strong monetization spike; one randomness integration. Watch the regulatory exposure.
- Rentals via app-level user-role (5). ERC-4907 (or its TON equivalent) unlocks scholarship dynamics with a single permission table. Massive new-player funnel.
- Quest NFTs feeding into pass eligibility (15 → 14). Combine patterns: complete quests → mint cheap NFTs → those NFTs gate the lucrative pass. Galxe blueprint, fully self-serve.
- Cosmetics over stats (12). Pay-to-look beats pay-to-win for community goodwill and is mechanically trivial — pure metadata swap. Stackable ERC-1155.
- Dynamic metadata via ERC-4906 events (3/9). Tiny code change (emit one event when state mutates) gives marketplaces "living" NFTs at near-zero cost. Pair with on-chain leveling.
Patterns to defer until later: Breeding (1) — economic balancing is brutal and can sink the whole game; Land (8) — engineering cost is enormous; Fractionalization (6) — regulatory and product-market risk; Cross-game interop (18) — requires partner muscle; AI-generated art (20) — model ops and copyright complexity unless it's the core pitch.
| Cube Worlds CNFT today | Interaction present? |
|---|---|
| Mint (admin-curated AI art, once per user) | ✅ Pattern: minting |
5 active types (Whale / Diamond / Coin / Knight / Common) by vote count + referrals; Dice remains in the enum on legacy rows but is no longer awarded |
|
| Stored as TON cNFTs (compressed), Pinata IPFS metadata | ✅ Infrastructure |
| Tradable on Getgems? | |
| Any of: fusion / rental / staking / evolution / quests / passes / cosmetics? | ❌ None |
The CNFT is currently a proof-of-mint trophy. Every modern long-lived Web3 game has 3–5 active NFT verbs. Cube Worlds has 1 (minting).
Ranked by (implementation simplicity × retention impact) and mapped to existing code.
Already in the previous market-research plan. The fastest unlock. Add getCnftMultiplier(type: CNFTImageType): number next to the existing pickCNFTType() in src/common/models/CNFT.ts:32–43 — pure type→multiplier table, no IO. Then a thin getUserCnftMultiplier(userId) wrapper in the claim flow (and any future clicker-reward / season-pass / tournament flow) looks up the user's CNFT via getCNFTByWallet() and applies the table. addPoints() consumes the multiplier. Zero new on-chain work; pure DB lookup. This is the foundation everything else builds on — without it, no other NFT verb has value. (Suggested table: Common 1.0×, Knight 1.1×, Coin 1.25×, Diamond 1.5×, Whale 2.0×. Dice is a legacy enum value still on past records — treat as baseline or migrate.)
Cheapest possible pattern. Mint a TON soulbound (TEP-62 with transfer disabled) for milestones: first claim, 10-day streak, 100K votes, first wallet-bind, referral threshold. Each badge is a separate cNFT type — extends the existing CNFT model (src/common/models/CNFT.ts) with a soulbound: boolean flag and adds a BadgeKind enum. AI-art prompt can vary by milestone, reusing the existing Stability AI pipeline. Display on the leaderboard for instant social proof.
Critical sink. CNFTs are currently mint-once-per-user — needs to evolve into mint-many. Two design options:
- Conservative: Allow users to burn 2 Common CNFTs → mint 1 Coin; 3 Coins → 1 Diamond; 3 Diamonds → 1 Whale. Burn happens on TON; new mint reuses the existing admin pipeline but auto-runs.
- Aggressive: Allow burning CNFTs for CUBE jetton at a rate set by the type, creating a clean two-way sink (mint costs CUBE → upgrade costs more CUBE/burns → burn returns some CUBE). This pairs with the existing SATOSHI exchange and gives CUBE a real demand floor.
Either way, it's the single biggest sink mechanic in Web3 gaming.
Fanton's pattern, perfectly matched to Cube Worlds. Whale/Diamond/Coin holders can list their CNFT for rent (denominated in CUBE); a renter pays daily fee → during rental, the renter's claim (and future clicker / tournament) gets the multiplier from §1, owner banks the CUBE. Implementation: new Rental model linking User + CNFT + expiresAt; the multiplier lookup in getCnftMultiplier(userId) checks rentals as well as ownership. TON has no ERC-4907 equivalent — this is app-layer, which is actually fine and simpler.
Catizen's Airdrop Pass made $2M as a single feature. A seasonal "Cube Pass" minted as a transferable NFT (Common type, season-themed art) sold in Telegram Stars. Pass holders get 2× claim, exclusive AI-art prompts, weekly bonus drops, and the right to a season-end commemorative SBT. New SeasonPass model linked to User + Season. Stars webhook handler on Fastify side; transfer pass NFT via existing TON pipeline.
- Breeding — economic balancing risks ruining the game; Axie/CryptoKitties cautionary tale; defer until base economy is healthy.
- Land / plots — engineering cost is enormous; doesn't fit the bot UX.
- Composable / equippable (ERC-7401) — meaningful only when there are slots; not yet relevant.
- Fractional ownership — regulatory risk, no clear win.
- Cross-game interop — needs partner muscle; revisit once a community exists.
- Phygital burn-to-redeem — logistics is the hard part; too early.
The market-research plan's Phase 1 already proposed CNFT-as-multiplier. Pair it with achievement SBTs in the same PR — both touch the CNFT model, share the AI-art pipeline, and together unlock the value of every later pattern:
getCnftMultiplier(userId)pure helper + wire intoclaim-handler.ts(and re-use from any future clicker / season-pass / tournament reward path).BadgeKindenum onCNFT.ts; soulbound flag; mint badges from a small set of milestone events (already trackable via theBalancemodel'sBalanceChangeTypelog).- Frontend badge gallery on the user profile area.
- (Existing market-research items: unhide clicker, referral dashboard, mint notifications.)
Phase 2 of the earlier plan should add fusion + season pass; Phase 3+ adds rentals.
- Gods Unchained's Forge (stair-step fusion with token + flux gate). Cleanest sink mechanic in the entire industry.
- Fanton's rentable NFTs (multipliers, app-layer rental, owner-renter CUBE split). Direct match for Cube Worlds' type system.
- Catizen's Airdrop Pass (seasonal Telegram-Stars pass; $2M revenue line). Highest-ROI revenue addition you can make.
Standards & primitives:
- EIP-4907 Rental NFTs
- EIP-5192 Soulbound NFTs
- RMRK ERC-7401 Nestable · ERC-7401 overview
- ERC-998 vs ERC-6551 · ERC-6551 Token Bound Accounts
- TON NFT Standard TEP-62 · TON NFT 2.0 Docs
Games:
- Axie: Breeding whitepaper · Breeding guide · Yu-kai Chou collapse analysis · Outlier Ventures P2E 2.0 · Lunacian Homecoming
- Gods Unchained: Fusing in the Forge · Shine Fusing · Flux/Fusing/Forge · zkEVM migration
- Splinterlands: Combining Cards · Rental System · Burning Cards · Rental-market analysis
- Sorare: 2026 mechanics · Sorare Football
- Parallel: DappRadar guide · Flagship analysis
- Pixels: Decrypt overview · Land guide
- Big Time: Help center · Crafting guide · SPACE wiki · $BIGTIME sinks
- Illuvium: Guide · Patch 1.1.0 Leviathan
- Star Atlas: Faction Fleet · SCORE gameplay · Loot box drop
- Off The Grid: Decrypt overview · Avalanche Battle Pass
- Shrapnel: CoinGecko guide · Marketplace launch
- Heroes of Mavia: Land NFT rarity · Phase 2 staking
- Guild of Guardians: zkEVM migration · NFT migration complete
- Wreck League: Guide · Mech parts · Decrypt preview
- Champions Tactics (Ubisoft): FAQ · 80.lv launch
- CryptoKitties: IEEE Spectrum collapse
- Tamadoge: Bitget price status
Pattern references:
- Decentraland Linked Wearables
- Otherside Otherdeed
- Chainlink Dynamic NFTs · Sports data oracles
- ApeStake.io
- Nouns DAO Governance
- Catizen Airdrop Pass
- X Empire NFT Vouchers on TON
- Mavis Market fee structure
- Loot Project · Yuga + M3taloot
- Nike + RTFKT phygital · Redeem-and-retain essay
- Botto decentralized artist
- Tessera shutdown
- Galxe + NodeReal SBT
- Sequence on Soulbound tokens · CoinGecko SBT
- Phala on dynamic NFTs
- Bitrue 2026 adoption · Data40 P2E 2026 · Mordor Intelligence
- Fanton TON Innovators
- CryptoKitties research: Frontiers in Physics