Blockchain in Gaming: Play-to-Earn Models and NFTs

Blockchain gaming stopped being interesting as soon as people treated token emissions as player income. The sector is still large enough to matter. DappRadar recorded 7.4 million daily unique active wallets by December 2024, then 5.8 million daily wallets in Q1 2025 as activity cooled. That is not a dead category. It is a category that has already stress-tested its first economic model.

Play-to-earn worked best as onboarding theater. It showed that players would tolerate wallets, token rewards, and secondary markets if the upside looked immediate. It worked worst as a durable labor model. A February 14, 2026 preprint that measured 12 NFT games found that a small group of top wallets controlled disproportionate NFT ownership, while players in 9 of the 12 games who traded NFTs had negative average profit. That is the core correction. Many players were never earning wages. They were holding volatile game inventory and hoping liquidity stayed available.

Blockchain gaming stayed large after the hype cycle

Blockchain gaming remains one of the few crypto sectors with recurring user activity. DappRadar’s January 2025 report put on-chain gaming at 7.3 million daily unique active wallets, up roughly 3x year over year, before the sector cooled into Q1 2025. That combination matters. Usage is real, but retention and monetization are still cyclical.

Capital already reflects that reset. DappRadar says blockchain gaming and metaverse investment fell to $1.8 billion in 2024, down 38% year over year, and Q1 2025 funding was only $91 million, down 71% quarter over quarter even as deal count rose 35%. Investors are still funding infrastructure, but they are less willing to subsidize reward-heavy user acquisition.

The market implication is simple. “GameFi” is not one model anymore. It now spans competitive games, creator economies, NFT marketplaces, metaverse land systems, and chain-native infrastructure. The old play-to-earn label hides that variation. A better lens is to ask what exactly the player receives, what can be sold, how often new supply hits the market, and who provides the bid on the other side.

NFTs changed distribution of in-game value

NFT integration changed gaming by moving certain assets from closed database entries to transferable wallet objects. Axie’s whitepaper states that game resources and items are tokenized and can be sold on open peer-to-peer markets. Decentraland’s original whitepaper describes LAND as a scarce, transferable NFT, while current marketplace documentation shows active trading for LAND, wearables, emotes, and names.

That ownership shift is real. Players can custody assets, transfer them, price them, and in some cases collateralize them outside the original game loop. But the value of those assets is still game-scoped. A land parcel, avatar item, or breeding token only keeps economic meaning if the game client, community, and marketplaces stay active. Wallet ownership improves exit rights. It does not guarantee demand.

That distinction is why NFTs are more powerful in gaming than pure reward tokens. NFTs can express scarcity, progression, cosmetics, access, or land rights in ways that do not require constant monetary emission. Reward tokens usually need a buyer every day. NFTs need one when the owner exits. The liquidity burden is lower, even if price discovery can still be brutal.

Liquidity structure, not FDV, decides if play-to-earn is sustainable

Play-to-earn economies fail when daily sell pressure scales faster than sinks and market depth. This is the central liquidity problem. A capped governance token can look disciplined on a slide deck, while an uncapped or fast-emitting reward token does the real economic damage. Theoretical fully diluted supply rarely tells you what players are actually selling this week. Emission cadence, burn mechanics, and exchange liquidity do.

Circulating supply is only a starting point. As of March 8, 2026, Etherscan listed 169,340,673 AXS in circulating supply out of a 270,000,000 AXS maximum, or about 62.7% of max supply, with 76,119 holders. On the same date, Etherscan listed 1,935,233,656 MANA in circulating supply out of 2,193,179,327.32 MANA total supply, or about 88.2%, with 280,075 holders. That makes MANA structurally closer to a mature float and AXS structurally more exposed to future issuance. Neither figure, by itself, tells you how concentrated the liquid float is or how much can clear without slippage.

That is the trade-off most token models still miss. Supply optics can look better than real liquidity. A low circulating percentage may support a narrative of future scarcity, but it also means future unlock risk. A high circulating percentage reduces unlock overhang, but it does not solve weak demand. The market pays for exit capacity, not for tokenomics aesthetics.

Axie Infinity shows how emissions outrun demand

Axie Infinity remains the clearest case study because the team described the economic problem in unusually direct terms. The Axie whitepaper says that, early on, the economy was designed to depend on new entrants and that Axie population growth required managing an “ideal inflation rate.” That was honest. It also reveals the structural fragility of early play-to-earn. When onboarding slows, reward-token sell pressure no longer gets absorbed by new user demand.

Sky Mavis later documented the failure mode in public. On January 4, 2022, the team wrote that SLP burn had grown more than 500x through 2021, but SLP creation had still outpaced use and grown more than 160x, calling the inflation unsustainable. On February 3, 2022, the team removed the 50 SLP daily Adventure reward, saying it would cut about 130 million SLP per day, and it removed the daily quest reward for another roughly 45 million SLP per day. The post explicitly warned of “total and permanent economic collapse” if decisive action was not taken.

That is the real lesson from Axie. Reward tokens are not sustainable because they are utility tokens. They are sustainable when the game creates enough reasons to spend, burn, or lock them before players rush to sell. Sky Mavis later capped SLP supply at 44 billion and introduced a buyback and stability mechanism in September 2023, but that policy itself was an admission that the token needed active monetary management rather than passive faith in growth.

Infrastructure risk compounded the economic risk. On March 29, 2022, Ronin disclosed that 173,600 ETH and 25.5 million USDC had been drained from the bridge after the compromise of validator keys. The official post said the root cause was the small validator set, and detailed how an allowlist path linked to the Axie DAO validator was abused. When a game economy depends on bridged liquidity, wallet trust, and a small validation perimeter, security becomes part of token design whether teams like it or not.

Decentraland shows a slower, property-style economy

Decentraland represents a different model. Its original whitepaper framed LAND as scarce digital real estate and MANA as the token used to claim parcels and make in-world purchases. The protocol describes LAND as acquired by burning MANA, and the whitepaper set a baseline exchange rate of 1000 MANA for 1 LAND for new parcel claims. That structure ties token utility to land scarcity and creator activity rather than to constant gameplay emissions.

Current product design still follows that logic. Decentraland’s documentation shows a marketplace for LAND, estates, wearables, emotes, and names, while in-world docs describe Genesis City as thousands of community-owned parcels. This is closer to a property and creator economy than to a wage economy. Players and creators monetize by building, trading, renting, or selling scarce digital assets, not by claiming a faucet every match.

That model is slower, but it is usually healthier. It reduces the daily necessity to dump a reward token into the market. It also means demand is more cyclical and more narrative-driven. DappRadar reported that metaverse NFT trading volume fell 80% in 2024 versus 2023, and that metaverse NFT collections generated only nearly $15 million across about 32,824 sales in Q1 2025, down 28% in volume from the prior quarter. A property-style game economy avoids reward-token inflation, but it is still exposed to low turnover and weak speculative demand.

What token economy design should optimize now

Blockchain games should optimize for tradable float discipline before they optimize for headline supply stories. That means smaller and slower emissions, clearer sinks, deeper venue support, and less dependence on a single reward token to carry onboarding, retention, and speculation at once. A governance token can coordinate upside. It should not be the hidden subsidy for a weak gameplay loop. A reward token can be useful. It should not be treated like a salary rail unless the game can consistently create spend demand on the same cadence that rewards are minted.

The strongest Web3 gaming models now look less like pure play-to-earn and more like play-and-own, play-and-trade, or play-and-create. The revenue path can still be meaningful for players. It just cannot rely on permanent token inflation and thin exit liquidity. For teams doing token economy design, this is where serious tokenomics consulting starts. At FinDaS Tokenomics, the right first question is not “what should the FDV be.” It is “who can sell, how often, into which liquidity, and what organic reason exists for someone else to buy.” That question usually explains the outcome before the narrative does.