Most blockchain games have now shut down. The ones that remain are moving away from play-to-earn and towards cheaper infrastructure and fewer token rewards. In a report covered by CoinDesk on 23 April 2026, the trading firm Caladan estimated that about 93% of GameFi projects are effectively dead and that the sector burned through up to $15 billion. In August 2026, one studio built on the "fully on-chain game" idea announced it was closing. Three months earlier, in May 2026, Ronin, the chain behind Axie Infinity, moved onto Ethereum and cut its token issuance by roughly 9x. Smaller and less inflationary is where blockchain gaming is heading. Mass adoption is not.
What blockchain gaming is
A blockchain game records some of its items, currency or characters as tokens on a public blockchain instead of only on the publisher's servers. Unique items are usually NFTs. Ethereum.org describes these as "tokens that are individually unique", whose ownership "is stored on the blockchain for anyone to verify publicly". It lists "ownable items for games" among their uses.
The labels you will see:
- Play-to-earn: players receive tokens or NFTs for playing, which they can sell.
- Pay-to-earn: players first buy an NFT, such as a character, land or a card, before they can earn.
- GameFi: the umbrella term for games built around these earning models.
What happened: the 2021–2025 boom and bust
All figures in this section come from Caladan's report as summarised by CoinDesk (23 April 2026). Caladan published no methodology with it.
- Gaming took about 63% of Web3 venture funding in 2022. By 2025 its share had fallen to single digits, as capital moved to AI, real-world asset tokenisation and layer-2 infrastructure.
- Funding to game studios fell 93% by 2025, and gaming tokens are down about 95% from their 2022 peaks.
- According to DappRadar data cited in the report, more than 300 blockchain games have shut down. Axie Infinity's daily active users fell from about 2.7 million at the peak to about 5,500.
- A Coda Labs survey cited by Caladan found that only 12% of gamers had tried a crypto game, even at the height of the boom.
Caladan's explanation: studios raised large sums before shipping a playable product, development took three to five years, and tokens traded live the whole time, so many collapsed before the game launched.
Closures with dates
- Ember Sword (Bright Star Studios): announced the end of development and server support on 21 May 2025, citing insufficient funding, and took the servers offline.
- Symbiogenesis (Square Enix): the NFT story project ran a final season that ended in July 2025.
- Pirate Nation / Proof of Play: the game closed in August 2025 because "the audience no longer justified continued investment". On 4 August 2026 the studio announced it was ceasing operations. It open-sourced the Unity client, smart contracts and internal tools, and released its Founder Pirates artwork under CC0. An independent foundation says it will keep supporting the PIRATE token, which "has lost its principal game utility".
- Axie Classic (Sky Mavis): Sky Mavis said Season 16 (12 August to 2 September 2026) was the last season under the current structure, and that after it the company would pause competitive seasons and active development. It plans to hand the game's direction to the community, but said "the details aren't settled yet".
The newest development: Ronin's move to Ethereum
On 12 May 2026, Ronin ran a hard fork at block 55,577,490 with about 10 hours of planned downtime. On 13 May its official blog announced: "Ronin's migration to Ethereum is complete." According to Ronin:
- The chain went from a sovereign layer-1 to an OP Stack chain (Optimism's software). Conduit led the technical work.
- RON is still the gas token.
- RON emissions fell from 45 million to 5 million a year, a 9x cut. Ronin puts expected inflation from emissions at about 1.2% a year, against more than 20% before the upgrade.
- Future emissions go to developers through a monthly "Proof of Distribution" system. It scores builders on gas spend, new and active users, and NFT, DEX and contract volume.
- Ronin's share of marketplace fees rose from 0.5% to 1.25%.
- ZK fraud proofs are planned but not yet live.
These are Ronin's own statements. The reasoning is clear, though: fewer new tokens, rewards paid to developers instead of directly to players, and security shared with Ethereum instead of a small set of the chain's own validators.
Four checks before you spend on a blockchain game
Answer four questions about a blockchain game to see which of the article's warning signs apply to it.
Is there a live, playable game today, not just a mint, a token or a roadmap?
- Yes, I can play it now → next question
- No, or not yet → No playable game yet
Can you find how many new tokens are issued each year, and who receives them?
- Yes, issuance and recipients are published → next question
- No, or the figures are unclear → Token issuance is unknown
Does the studio say what happens if it closes: open-source commitments, a published shutdown policy, and who controls the smart contracts?
- Yes, at least some of this is published → next question
- No, nothing is published → No plan for a shutdown
Are people playing who are not there to earn?
- Yes, there are regular players → No warning signs on these four checks
- No, activity is mostly earning → Players are mostly there to earn
Caladan describes studios that raised money before shipping, spent three to five years in development and collapsed before launch. Its example, Pixelmon, raised $70 million in a 2022 NFT mint and has no public game four years later.
Issuance can weigh heavily on a game token. Ronin's expected inflation was more than 20% a year before its upgrade and is about 1.2% after it, by Ronin's own estimate. Without that figure, you cannot judge this risk.
An NFT stays in your wallet after a closure, but what the item does may not. When Ember Sword's servers went offline, its items had no game left to be used in.
Low engagement from regular gamers is the failure pattern Caladan describes. A Coda Labs survey it cites found only 12% of gamers had tried a crypto game, even at the height of the boom.
The game passes the article's four checks. That does not make it safe: even in Proof of Play's best-case closure, the code was open-sourced but nobody is guaranteed to rebuild the game.
This walks through the article's four checks only. It does not assess a token's price, a studio's finances or the security of its contracts, and it is not a recommendation to buy or avoid any game.
What owning an in-game NFT does and does not protect
The original promise was that players own their items even if the publisher disappears. The closures above show where that promise stops:
- What survives: the token record. An NFT stays in your wallet and can still be transferred.
- What does not: what the item does. When Ember Sword's servers went offline, there was no game left to use its items in. Pirate Nation's token kept trading after the closure but "lost its principal game utility".
- Best case so far: when Proof of Play closed, it open-sourced its code and released its art. That lets other people rebuild the game, but it does not guarantee anyone will.
Practical checks before spending money on a blockchain game
- Is there a live, playable game? Caladan's report gives Pixelmon as an example: it raised $70 million in a 2022 NFT mint and has no public game four years later.
- How many new tokens are issued, and to whom? Ronin cutting issuance from over 20% to about 1% shows how much inflation can weigh on a game token.
- What happens if the studio closes? Look for open-source commitments, a published shutdown policy and who controls the smart contracts.
- Is anyone playing who is not there to earn? Low engagement from regular gamers is the failure pattern Caladan describes.
What to watch next
- Whether Ronin's builder rewards bring in new games with real user numbers. Its scoring weights are reviewed monthly.
- Whether the community takes over Axie Classic after Sky Mavis steps back.
- Whether investment keeps moving from individual games to infrastructure, which is the shift Caladan describes.
Limitations: the sector-wide figures come mainly from one market maker's report, published without a methodology, and the Ronin details come from Ronin's own blog.
