GameFi after the shakeout: what the surviving web3 games are doing differently
Most blockchain games launched in the last cycle are gone and funding has collapsed. The projects still standing share a few traits worth understanding.
GAMEFINFTANALYSIS
Sofia Marchetti
10/28/20263 min read
Few corners of crypto have been tested as harshly as blockchain gaming. Research firm Caladan estimated this spring that roughly 93% of GameFi projects are effectively dead, that gaming tokens are down about 95% from their 2022 peaks and that the sector absorbed between $12 billion and $15 billion of capital since 2020. In October, the closures reached the infrastructure layer: Igloo announced it would shut down Abstract, the consumer-focused Ethereum layer 2 behind Pudgy Penguins, after losing tens of millions of dollars running it. Yet a small group of games and networks is still operating, and what they have in common says a lot about where the sector goes next.
How the first model failed
The play-to-earn design of 2021 linked gameplay to a financial loop. Players earned tokens, sold them, and new players bought in to join. As long as inflows grew, rewards looked generous. When inflows slowed, token prices fell, rewards shrank and players left. According to DappRadar data cited by Caladan, Axie Infinity's daily active users fell from about 2.7 million at peak to around 5,500.
The funding cycle made the problem worse. Gaming took about 62.5% of web3 venture investment in 2022, according to Caladan, and annual funding peaked near $4 billion before falling to about $360 million in 2025. Many studios raised large sums before shipping a playable product, while tokens traded freely for years before launch. By the time several games were ready, their tokens had already collapsed. Pixelmon raised $70 million in a 2022 mint and still had no public game four years later. Ember Sword spent around $18 million over seven years before closing without refunds.
Telegram's tap-to-earn wave showed the same pattern at larger scale. Mini-apps of this kind attracted hundreds of millions of users, but Hamster Kombat lost about 96% of its audience within six months of launch, according to the same research. Reach without retention proved no more durable than the earlier model.
What the survivors share
Caladan's analysis points to three traits among the projects that remain: they built a game first and added blockchain features later, they treated tokens as a supporting mechanism rather than the product, and they aimed at mainstream players rather than crypto traders.
Off the Grid, the shooter from Gunzilla Games, is the clearest example. It raised more than $100 million, launched on mainstream platforms and keeps its on-chain economy in the background, so players can ignore it entirely. Cambria, a competitive multiplayer game, built its economy around player-versus-player wagers, which Caladan says exceeded $150 million, giving its token a use tied to skill rather than emissions.
Market data from October points in the same direction. Immutable recorded about $2.55 million in weekly NFT sales in the week to October 17, according to CryptoSlam, led by Guild of Guardians heroes, a sign that in-game items still trade when the game behind them has players. These figures are small next to traditional gaming, but they rest on actual play rather than token rewards.
The infrastructure question
Abstract's closure raises a separate issue. Many gaming and consumer projects launched their own chains on the assumption that dedicated infrastructure would attract users. CoinDesk reported that Abstract processed more than 325 million transactions and attracted 4 million wallets, yet generated only a few thousand dollars a day in chain fees. Igloo is now refocusing on the Pudgy Penguins brand and its token rather than on running a network.
The lesson for studios is that infrastructure is a cost centre, not a business, unless it carries far more activity than most games produce. Several large investors have reached similar conclusions. Animoca Brands, long the sector's biggest backer, has reduced gaming to about a quarter of its portfolio and shifted attention to stablecoins, tokenized assets and AI as it pursues a US listing.
Three scenarios for 2027
In the first scenario, invisible blockchain wins. Games use wallets, stablecoin payments and tradable items in the background, players rarely notice, and success is measured in retention and revenue like any other game. This looks most consistent with what is working today.
In the second, distribution through messaging apps and retail trading platforms brings a new wave of casual players. Telegram's built-in wallet rollout and retail-focused chains such as Robinhood Chain could lower the barrier to owning in-game assets, though tap-to-earn has already shown how quickly such audiences can leave.
In the third, consolidation continues. More studios close, a handful of networks such as Ronin and Immutable remain the main homes for blockchain games, and capital stays focused on infrastructure that serves several industries at once.
The most likely path combines the first and third. GameFi is not finished, but the version of it that promised income for playing is. What remains is a smaller industry that has to compete on the same terms as every other game studio: make something people want to play.