Author: Whiteboard Blockchain
One person, one playbook, two exits.
On October 2, the Ethereum Layer 2 network Blast issued an announcement of just a few sentences: operating costs exceeded revenue, and it decided to shut down. Ten days later, the NFT trading platform Blur announced it would shut down its site on November 12.
Both projects came from the same person—Pacman, whose real name is Tieshun Roquerre.
Blur's monthly trading volume peaked at $1.43 billion. DefiLlama recorded cumulative protocol revenue of $2.26.
Not $22,600—$2.26. All fees went to sellers and creators, and the protocol retained $2.26 for itself.
This is the story of a person who took "points plus airdrops" to the extreme, and also a sample of this playbook reaching its end.
In February 2023, Pacman publicly revealed his real identity on X. He wrote one sentence to introduce himself: Dropped out of high school to go through Y Combinator at 17.
After dropping out, he entered MIT and then dropped out again, received $100,000 in startup funding from Peter Thiel, built the decentralized domain trading platform Namebase, which raised $5 million led by Paradigm and was fully acquired by Namecheap at the end of 2021. After selling Namebase, he turned to the NFT market, saw OpenSea's shortcomings in matching speed and listing tools, and began assembling a team to build Blur in early 2022.
Blur's playbook was simple: zero trading fees, with points as the real hook. Users placed bids; the closer the bid was to the floor price and the longer it stayed listed, the more points they earned, which were eventually converted into BLUR token airdrops.
BLUR launched on February 14, 2023, with 360 million tokens airdropped in the first season, worth about $270 million at the post-launch price of about $0.76. The next day, Nansen data showed Blur's daily trading volume surpassed OpenSea for the first time. Galaxy Research estimated that Blur's share reached as high as 80%, leaving OpenSea with only 15.5%.
In March, Blur's single-month trading volume surged to about $1.43 billion. How much of that was wash trading varies widely by algorithm: Dune analyst hildobby calculated about 14%, while CryptoSlam calculated more than 80% in the two weeks after the airdrop.
To attract sellers, Blur also pushed creator royalties down to a minimum of 0.5%. The Block analyst Brad Kay noted that Yuga Labs' quarterly royalty revenue fell from $8.7 million to $2.5 million, mainly due to Blur's low royalties and mining incentives. Three years later, when Blur announced its shutdown, collector Cryptobscure replied on X: "Good riddance. You ruined NFTs."
Zero fees bought trading volume, and points bought liquidity, but the protocol itself never intended to make money from trading from beginning to end. This was the foreshadowing of the first exit.
On November 20, 2023, Pacman launched Blast. The selling point was direct: money deposited earns yield automatically. ETH deposited into Lido staking earned about 4% annualized, and stablecoins were converted into USDB, backed by MakerDAO's U.S. Treasury yield of about 5%.
The points model was copied, with invite codes added. New users needed an invite code to deposit; according to the official documentation, inviters received an additional 16% of the invitees' points, and second-level invites earned another 8%.
At that time, the Layer 2 network had not yet launched, and deposited funds were frozen for about three months. Even so, DefiLlama showed deposits exceeding $510 million by November 26.
Lead investor Paradigm was the first to speak out. Research head Dan Robinson posted on X that day: he disagreed with launching a deposit bridge before the Layer 2 was ready, and disagreed with preventing users from withdrawing for months; many marketing tactics "cheapened the work of this serious team."
Pacman posted several responses in a row on X: Blast's yield came from Lido staking and MakerDAO U.S. Treasuries and had a real economic basis; Paradigm did not participate in Blast's distribution design, and the lockup and invite-code mechanisms were independently decided by the team.
In other words: the money is real, the playbook is mine, and it's none of your business.
The controversy did not stop deposits. The mainnet launched on February 29, 2024, and BLAST launched on June 26, with 17 billion tokens airdropped in the first season. At the launch price of about $0.025, Decrypt estimated the total value at about $430 million, of which 14 billion tokens went to users.
Locked value began declining before the token launch. DefiLlama data showed a peak of $2.26 billion on June 6, which had fallen to $1.63 billion on the day of the token launch. Users received the airdrop and then left.
Blur's first two seasons plus Blast's first season totaled about $850 million in three airdrops. After these tokens were distributed, the vast majority were claimed and sold immediately. The moment the airdrop ended was the starting point of the project's decline.
Blur's cumulative protocol revenue was $2.26. The roughly $70.82 million in fees users generated on the platform over three years all flowed to sellers and creators because of the zero-fee design. Blast's cumulative on-chain fees were about $13.85 million, of which about $7.83 million truly counted as protocol revenue. By the time the shutdown was announced, the entire chain's 24-hour revenue was only $110.
The original wording of Blast's shutdown announcement was: "The costs of maintaining operations continue to exceed the revenue the Layer 2 network can capture."
The same was true for Blur. Monthly trading volume fell from $1.43 billion to less than $7 million by March 2026, a drop of 99.5%. The BLUR token fell from a high of $5.02 on its launch day to $0.0095, a drop of 99.8%. According to DefiLlama data, Blast's locked value fell from $2.26 billion to about $11.5 million at the time of shutdown. L2BEAT, using a broader statistical method, calculated that about $50 million in assets remained on-chain, with the difference mainly being ETH staked in Lido in the bridge contract.
Machi Big Brother Huang Licheng was also a major Blur player, and Decrypt described him as one of the largest lenders in Blend lending. In June 2024, he called on Pacman on X to adjust the mining rules, and half a month later transferred about 18.4 million BLUR entirely into Binance; Spot On Chain estimated a loss of about $3.16 million.
The broader environment did not help either. NFT trading as a whole shrank, and Ethereum Layer 2s kept multiplying. In October alone, Abstract, StandX, and ZetaChain successively announced shutdowns or migrations, and Farcaster changed hands for the second time this year. Users and funds attracted by token expectations could not cover costs once those expectations were met or disappointed.
The $850 million in airdrops ultimately flowed into sell orders in the secondary market, leaving the projects with a bill they could not pay.
Before October 26, Blast users could still withdraw normally on the official page. After that, assets would not disappear, but could only be recovered by interacting directly with the bridge contract on Ethereum; the official team said operating instructions would be published before the deadline.
On November 12, blur.io goes offline, and Blend stops matching new loans. Anyone with a Blend loan must repay it before then, otherwise it will be treated as a default and the pledged NFT will go to the lender. Blur said it would later provide tools to help users recover assets from Blend. The only official account is @blur_io; do not trust any other notices claiming to be official.
BLUR holders still have governance rights over the protocol fee switch, but with the website closed and trading almost stopped, even turning the switch on would collect no money.
After Blur announced its shutdown, Cointelegraph contacted Pacman for comment, but the report contained no response from him. There is currently no public information about what his next step will be.
Three years ago, Pacman defeated OpenSea with points; two years ago, he attracted $500 million in deposits with yield and invite codes. Now, in 10 days, he has shut down the two largest projects he ever built.Points can buy users, but not revenue. The next person to try this playbook will probably issue the same announcement one day: operating costs have exceeded revenue.