Five Hard Lessons From the Wreckage of Blockchain Gaming's Boom

In 2021, a cartoon creature-battler made by a small studio in Vietnam took in about $1.3 billion in revenue. Venture capital poured into anything with "play-to-earn" in its pitch deck. Some of the biggest publishers in the world said, on the record, that blockchain was the future of games.

Five years on, the record looks very different. The biggest game of the boom lost around nine in ten of its daily players. Its bridge was robbed of about $625 million. Token after token fell by more than 90%. Celebrities and anonymous developers alike sold games that never shipped. And, more quietly, one blockchain title after another simply switched off its servers.

This is WGC's attempt to draw the lessons together. It is an editorial, so the opinions are ours. The facts behind them are laid out in our series of history pieces, linked throughout.

By the numbers

• $1.3 billion: Axie Infinity's treasury revenue in 2021.
• Over 99%: how far Axie's SLP token fell from its July 2021 peak by February 2022.
• About $625 million: stolen from the Ronin bridge in March 2022.
• 96%: the fall in monthly metaverse land trading from January to November 2022.
• 90%: STEPN's drop in daily active users between June and September 2022.
• 70%: game developers with no interest in NFTs, according to a 2022 GDC survey. Only 1% said they were integrating them.

Lesson 1: If the game is a job, the job market decides

The defining idea of the boom was that players should earn money for playing. It sounded generous. In practice it turned games into labour markets, and labour markets follow their own rules.

Axie Infinity is the clearest case. Players earned a token, SLP, that was mainly spent on breeding new creatures, which were mainly sold to new players. While newcomers kept arriving, it worked. When they slowed, the economy collapsed: SLP lost over 99% of its value, the average creature fell from $494 to about $37, and daily revenue went from a peak of $17.5 million to around $40,000. STEPN ran the same experiment with jogging and got the same result, faster.

The lesson is not that player-owned economies are impossible. It is that rewards cannot be the main reason to play. If a game is only worth playing because it pays, it will stop being played when it stops paying, and someone else has to fund the payments in the meantime.

Lesson 2: A wallet is a vault, so guard it like one

The second lesson is about security, and it is the most expensive one on this list.

The Ronin hack did not involve breaking any cryptography. Attackers used a fake job offer to plant spyware, took four of nine validator keys, and used a forgotten permission to get the fifth. Nobody noticed for six days. Two years later, Ronin was hit again, this time through a botched upgrade, though white-hat actors returned the money.

The pattern repeated across 2024. At Munchables, a hired developer wrote himself a balance of a million ETH and walked off with about $62.5 million, before handing it back. At Gala Games, a minter account dormant for six months was used to create 5 billion tokens.

In every case, the failure was human: keys, permissions and people. Any studio that holds player assets is effectively running a bank and needs bank-grade habits: minimum permissions, prompt revocation, real-time monitoring and someone whose whole job is to worry.

Lesson 3: Money before the work is a warning, not a feature

NFT sales let studios raise enormous sums before writing a line of game code. That turned out to be a curse as often as a blessing.

At the honest end, Pixelmon raised about $70 million in a day and then had to show buyers its art. The reveal was so poor that one creature, Kevin, became a meme. At the other end were outright scams. The Squid Game token let buyers in but not out, then its creators vanished with an estimated $3.38 million. The developer behind Evolved Apes disappeared with 798 ETH raised for a fighting game that never existed, and US prosecutors charged three men in 2024. And CryptoZoo, promoted by one of YouTube's biggest stars, ended in lawsuits and a buyback that required buyers to give up their right to sue.

The common thread is simple. When you pay for a game you cannot yet play, you are buying a promise. Promises from anonymous creators, or from famous ones with no track record of shipping games, are worth exactly as much as the person making them.

Lesson 4: Players and developers voted with their feet

The boom was driven largely by investors and token holders. Players and developers were far less enthusiastic, and they said so loudly.

In October 2021, Valve banned blockchain and NFT games from Steam. In December 2021, Ubisoft's Quartz NFT programme met such a backlash that its announcement video was pulled from YouTube, and trading in its NFTs was virtually zero in the following weeks. GSC Game World dropped NFT plans for S.T.A.L.K.E.R. 2 a day after announcing them. In February 2022, itch.io called NFTs a scam. In July 2022, Mojang barred NFTs from Minecraft. That year's GDC survey found 70% of developers had no interest in NFTs at all.

The message was hard to miss. Players did not want their hobbies turned into investments, and many developers did not want to build games around economies they could not control. Studios that ignored that lost goodwill they found hard to win back.

Lesson 5: Most endings are quiet

Not every failure is a hack or a scam. Most blockchain games end the way most online games do: the audience thins out and the servers go dark.

On 31 July 2025, Square Enix shut down Symbiogenesis, its NFT project launched in December 2023, and Gala Games closed The Walking Dead: Empires the same day. KLab ended support for Captain Tsubasa Rivals on 28 November 2025. Ubisoft re-released its tactics game Champions Tactics without the blockchain features it launched with. The metaverse worlds that sold plots for millions in 2021 found their land prices down about 90% a year later, and The Sandbox cut more than half its staff in 2025.

What matters for players is what happens to their items when a game ends. Owning an NFT does not keep a game running. If the servers close, a token in your wallet may still exist, but there may be nowhere left to use it.

WGC's take: what changed, and what did not

To be fair to the industry, some lessons have landed. Sky Mavis, the company most associated with play-to-earn, dropped the phrase and is now pitching an MMO on its gameplay. Security is taken more seriously, with bigger validator sets, withdrawal limits and emergency freezes like the one that saved Gala. Players are also far more sceptical, which is healthy.

But the deeper tension has not gone away. A game that lets players trade items for real money will always attract people who care more about the money than the game, and its designers will always be tempted to please them. The best blockchain games of the next few years will be the ones that are worth playing even if every token in them is worth nothing.

What to watch

Watch the games, not the tokens. When a new blockchain game launches, the useful questions are the same ones you would ask of any game: is it fun, who made it, and have they shipped before? Then add three more: who holds the keys to player assets, can you get your money out, and what happens to your items if the servers close? If the answers are clear, give it a try. If they are not, the history in this series suggests you should wait.