Papertrade Launch, 1000x Leverage & Onchain War Stories - blurr
Summary
Paper Trade replaces the traditional casino bankroll with a tokenized queue intended to return the house’s structural edge to participants. When the bankroll is insolvent, winning P&L becomes debt that later trader losses repay; when it exceeds its adjustable limit, the surplus can be distributed pro rata to stakers. blurr’s objective is to “externalize the profits that market makers and exchanges receive” back to users.
Transparent 1000x exposure is possible because Paper Trade reads Hyperliquid’s BBO directly on-chain through HyperEVM precompiles. At 1000x, a move beyond 0.1% can liquidate one side before the winning side offsets it, so volatile conditions initially restrict positions to BTC and ETH; blurr suggested SOL might work closer to 300x. The selling point is inspectability: “I specifically designed it so that you can see how it works.”
The hardest product problem is ordering transactions fairly when tens of thousands of users arrive together. Paper Trade routes activity through relayers and an internal mempool-like scheduler, prioritizing liquidations over closes and closes over opens; direct contract access would otherwise become a gas war in which interface users are “lambs to the slaughter.” If HyperEVM pauses while prices move, Paper Trade may freeze the market at its last price rather than let the first returning block become a race for free money.
The trading economics combine zero slippage and no funding with hard limits on scalpability and system exposure. Positions execute around the Hyperliquid BBO, but size restrictions prevent infinite positions; Paper Trade charges 1% on profits and losses, while its Rollbit-inspired payoff curve imposes a heavier charge on tiny moves. A $6 million winner against a $5 million bankroll would receive $5 million and enter a queue for the remaining $1 million.
Token emissions are strongest when the bankroll most needs recapitalization, creating a potentially unbounded but economically decaying supply. blurr could not recall the initial figure precisely, saying it might be $100 or $1,000 per paper before tentatively describing $1,000 for every $1,000 of paper losses; above roughly $2 million of bankroll, emissions ratchet downward toward zero. The token owns house commissions after a gas allocation, must be staked for emissions, and will not be tradable on day one: timing depends on how the live system behaves.
Paper Trade evolved from a playful mechanism into infrastructure designed around the possibility of 20,000 users on day one. blurr repeatedly attacked his own design—“How would I break it?”—including removing administrator-level attack paths, and sees failure as either a technical breakdown or nobody finding the product useful. His broader ambition is for others to inspect and iterate on the queue, because “it is a very new and unique mechanism.”
blurr’s larger crypto thesis is that “the purpose of a system is what it does,” and what crypto demonstrably does is move value safely online. Thread Guy’s examples included Bangladeshi click-farm workers using cheap USDT on Tron amid local-currency erosion, and the difficulty of splitting a restaurant bill in the United States before Venmo and Cash App. blurr’s conclusion is that financial adoption should precede attempts to use crypto for broader political or social goals.
His career—and his warning to younger builders—centers on taking asymmetric risks before the edge becomes institutionalized. RuneScape led to CryptoKitties genetics, Compound’s ticking 18-decimal balance, Uniswap sniping, three hours as effectively Curve’s only farmer, and a $13 million on-chain incident; today, venture capital absorbs many would-be “pirate developers.” With conventional labor buying less freedom, he thinks young people may still need to “flip the coin,” even as interconnected technology makes civilization feel like repeated heads until one terminal tail.
Deep dive
1. Paper Trade began as a mechanism that finally became buildable
blurr and Jez had worked together for roughly five years, but neither had independently built a complete product from scratch. New tools collapsed work that once required “a team of seven or eight people,” particularly interface and infrastructure tasks outside a founder’s core expertise.
The seed was an article, apparently written in summer 2024, about a queue-based casino bankroll. They first modeled deterministic games such as blackjack and coin flips, where outcomes were easier to express, but concluded those products were not interesting enough to launch.
Asked why accept the reputational downside when success might barely change their lives, blurr offered no convincing grand mission. He emphasized that the risks could outweigh the benefits: if everything went well, they might be no better off, while failure would hurt both of them. He then used the maxim, “A ship in the harbor is safe, but ships are not made for that.”
Jez’s tweet announcing the project drew more interest than expected. blurr realized that 20,000 users on the first day would require a fundamentally different product, so they redesigned around that possibility and around every way traders, administrators and infrastructure could break it.
2. The queue removes the house bankroll without removing the house edge
A conventional casino takes the other side of every bet and therefore needs enough capital to survive winners; its bankroll determines maximum bet size and creates a severe cold-start problem. Paper Trade instead pairs the system with an emission curve that rewards users more aggressively when the house is least solvent.
The incentive is intentionally counterintuitive: users who believe the house will be profitable over time may accept losses because those losses effectively buy into the system. “You encourage losing money to effectively buy into the system.”
blurr’s industry map is blunt: market makers and exchanges are the two crypto businesses that earn substantial money, and most of that money ultimately comes from users. HLP- or LLP-style liquidity pools gain specialized advantages, particularly liquidation flow and related fees, that ordinary market makers cannot easily reproduce.
Paper Trade keeps an exchange-like edge but tries to return the advantage it creates to users, including users who lost in the system, rather than leaving hidden economics with outside market makers. blurr’s framing is to synthetically put everything on the exchange while making the resulting house advantage flow back to users.
3. Hyperliquid’s on-chain BBO makes inspectable 1000x trading possible
Paper Trade reads the Hyperliquid BBO as its oracle through HyperEVM precompiles, rather than importing an off-chain price feed. blurr described one HyperEVM block and three HyperCore blocks per HyperEVM block and emphasized that the oracle can be read directly from the chain.
That access was decisive: other oracle systems become a product’s “stumbling block,” introduce authorized operators and theoretically expose keys or update processes. Hyperliquid can itself be manipulated, but blurr said its risk parameters are intended to ensure that manipulation is not profitable for Paper Trade.
Thread Guy pressed on why this architecture is not standard. blurr’s answer was technical difficulty: Hyperliquid’s precompiles made the approach possible, whereas many existing 1000x venues are effectively internal systems whose mechanics are not publicly visible.
The leverage remains dangerous despite the oracle. At 1000x, a move beyond 0.1% can liquidate one leg before the winning leg offsets it, leaving the house exposed. In volatile conditions, Paper Trade offers only Ether and Bitcoin; blurr said SOL might eventually support something nearer 300x, subject to much more complicated risk controls.
4. Fair execution is primarily a scheduling problem
blurr deliberately preserved the ability to close positions even if markets must be frozen. He described oracle restrictions that could set the value to zero and stop new positions while still allowing existing positions to close.
A HyperEVM maintenance pause creates a specific exploit: open a large long-short straddle, close the losing leg at the stale price, then close the winning leg after the chain resumes. If the deviation is extreme, Paper Trade may freeze the market at the last price instead of turning the first post-pause block into a race.
Direct contract access during a large launch would produce a gas war. Everyone using the interface would be “sent like lambs to the slaughter,” so early activity passes through relayers and an internal mempool-like ordering system.
Priorities reflect trader harm: liquidations outrank closes, and closes outrank opens. Deposits and other operations are expensive because Paper Trade must deploy proxy contracts, and the system also considers size and other factors. The goal is to make 50,000 simultaneous users workable rather than merely first-come, first-served.
5. No funding and zero slippage come with hard exposure limits
Paper Trade charges no funding. blurr argued that Bitcoin and Ether perps do not often enter significant backwardation or contango because spot is globally liquid and transferable; funding would matter far more for an RWA whose market closes overnight or over a weekend.
A low-leverage long could consequently be cheaper to hold on Paper Trade than elsewhere. That design would not safely generalize to oil or similar assets, where a known 10% opening move could make financing “a significant part of the deal.”
Execution uses the average BBO with zero slippage, but size is not infinite. Size restrictions are meant to stay invisible to ordinary users while blocking an attacker from opening a position large enough to overwhelm the bankroll.
Paper Trade takes a 1% commission on profits and losses. Its Rollbit-derived payoff curve charges proportionally more when a position barely moves and improves the payout farther from the mark, making easy scalping difficult.
6. Insolvency creates queued winnings, not an immediate claim on the bankroll
The bankroll is initially limited to about $5 million, restricting both the house’s exposure and the economic reward for manipulating Hyperliquid’s BBO. blurr estimated that moving the relevant Bitcoin order book by 0.01% could require at least low- to medium-eight-figure amounts in some cases.
If a closed position has $6 million of unrealized profit against a $5 million bankroll, the trader receives $5 million and holds $1 million of debt. That balance sits in the queue and is paid as later traders lose.
Thread Guy tested the catastrophic case: Bitcoin rises about 7% in an hour, everyone is highly leveraged long, and debt reaches perhaps $150 million. blurr conceded that this is theoretically possible, though he does not expect it; such a situation could cause participants to stop or leave.
Each position is isolated and temporary. A trader’s opening balance remains theirs rather than disappearing into a pooled margin account. blurr acknowledged that a scenario in which everyone wins could technically leave the system with its original balances but no ability to pay all paper gains immediately.
7. Token emissions act as an automatic recapitalization bid
The token owns the commissions generated by the house, with Paper Trade retaining a small amount for gas. Gas deserves its own carve-out because blurr expects Paper Trade could become one of the few very large users of HyperEVM and could otherwise spend more on gas than commissions generate.
blurr could not recall the maximum starting emission precisely. He said it might be $100 or $1,000 per paper, then tentatively described the mechanism as beginning at $1,000 for every $1,000 of paper losses. The transcript does not establish a clean token-per-dollar figure. Emissions are paid immediately after a loss and can be staked without a waiting period.
Once the bankroll reaches roughly $2 million, token output per dollar of loss begins ratcheting downward toward zero. Supply could technically expand indefinitely if the bankroll repeatedly falls below that threshold, but blurr expects the ratchet to create a practical ceiling.
Above the adjustable $5 million bankroll limit, anyone can press a button to distribute the surplus pro rata to stakers: at $6 million, roughly $1 million becomes distributable. The token will not trade on day one; release depends on live results that blurr compared to an uncertain Monte Carlo simulation.
8. The mechanism matters even if the first product does not
blurr’s most immediate failure case was “many people getting angry and tweeting at Jesse,” followed by the more substantive possibility that Paper Trade simply receives no useful adoption. He does not describe the project as a child, but the pair have invested enough effort that irrelevance would still matter.
He recalled Lighter publishing part of its prover stack and offering rewards for improvements. The expected gain was perhaps 20–30%, but the result was about 10 times faster. Thread Guy referred to roughly 3,400 people; blurr responded that this was about how many tokens were burned trying to do it, leaving the exact participant count unclear.
blurr said the Paper Trade design had been worked on by only two or three people in addition to the model. He hopes public release produces successive iterations, just as CryptoKitties and later NFT systems built on earlier experiments: “It’s worth seeing the light of day because it’s a very new and unique mechanism.”
9. Jez supplied the rare collaboration blurr could not replace alone
blurr had historically found collaboration costly because explaining and delegating work took longer than completing it himself. Jez is the exception: he can hand Jesse a small logical problem, receive a solution and explanation, and finish faster than he could have alone.
That complementarity is personal as well as technical. They met in New York in summer 2021, apparently through NFTs, quickly understood one another, and Jez became “probably one of my best friends.”
blurr is not especially worried that Hyperliquid will simply absorb Paper Trade. Its core goals differ, and Hyperliquid has little reason to optimize gas capacity for this exact workload, though sustained Paper Trade demand could eventually force a discussion.
His pseudonymity was accidental rather than ideological. He entered crypto around 2017, before crypto Twitter personalities were an obvious career strategy, and simply never developed an interest in personal social media. He worked alone behind a laptop and saw little upside in visibility; Thread Guy described Paper Trade as more “cult classic” than mass-market acquisition funnel.
10. Crypto’s real purpose appears where legacy money fails
blurr’s governing heuristic is that “the purpose of a system is what it does,” not what its founders said it should do. Crypto’s demonstrated function is secure value transfer online; financial applications are therefore the clearest evidence of what the system is for.
Thread Guy supplied the example of paying a Bangladeshi click farm that accepted only USDT on Tron. Cheap transfers mattered while local cash allegedly lost around 40% of its value annually; blurr agreed that crypto solves problems that people in wealthy countries often do not encounter.
Thread Guy also described his personal conversion. Britain already offered instant, free bank-to-bank transfers, but splitting a restaurant bill in the United States around 2018 resembled an international wire—roughly $30 and potentially weeks—making crypto payments suddenly legible before Venmo and Cash App filled part of the gap.
blurr remains skeptical of forcing blockchains into voting, charity or broad social coordination merely because early Ethereum culture imagined those uses. “This thing is for finance. It trades. It’s value on the internet.” Skilled financial actors will otherwise overwhelm idealists using the wrong tool.
11. RuneScape trained the instincts that Compound and CryptoKitties monetized
Before crypto, blurr played RuneScape for roughly 12 hours a day and had never held a conventional job. His father joked that he would end up “super dead, super rich, or in supermax”; a privileged upbringing gave him enough time to find an internet-native economic outlet.
The decisive moment came toward the end of 2018 after depositing a couple hundred dollars into Compound. Watching interest accumulate through all 18 decimal places—even fractions of a cent—felt like a collectible game with real money. He thought that if it affected him so strongly, it could become important.
At 19, influenced partly by the Ethereum white paper’s tokenization thesis, he left university shortly afterward. The downside was recoverable; if the thesis was right, working on it had “extremely high EV.” His description: “I threw myself off a cliff.”
CryptoKitties provided the first edge. Each cat encoded visible and recessive genes in a byte string; blurr bought cheap cats carrying valuable recessive traits, bred them until the traits surfaced, and sold the offspring at multiples. Three weeks of that replaced RuneScape permanently.
12. Early on-chain fortunes came from reading mechanics before the crowd
Uniswap sniping rewarded being first because an initial buyer had little immediate downside beyond liquidity removal or being frozen. Contracts could enforce “first or do not buy” with strict slippage checks, turning transaction ordering into a game before developers learned to counter it.
Ethereum’s roughly 15-second blocks meant implementation quality barely mattered. Even “the worst piece of JavaScript” could finish within the block; the competitive advantage was game logic and knowing how to make the code act on-chain, not shaving hundredths of a second in low-level code.
Curve’s launch is blurr’s cleanest specimen. An account called 0xChad deployed public contracts from GitHub, written in then-obscure Vyper; blurr figured out the farming mechanics and was effectively the only farmer for about three hours, created the initial Uniswap pool and sold individual CRV tokens at “insane” prices.
At Fei’s launch, depositing ETH returned stablecoins plus roughly a 20% token bonus, while a penalty escalated as the stablecoin deviated from its peg. blurr sold early and pushed it far enough off peg that later holders could barely exit, capturing what he regarded as literal free money.
13. Crypto must absorb finance before it can redirect political power
After roughly four years of feeling that the government had “a knee on our neck,” Trump’s November 2024 win briefly felt like permission to build and a declaration that crypto had won. blurr then watched “old demons” reappear, tempering that conclusion without changing his belief in the rails.
Political transformation requires capital first. He cited BitMEX co-founder Ben Delo and Christopher Harborne, whom he described as a Tether founder and effective inventor of the stablecoin, giving roughly $100 million to UK political parties—by his account, more than all parties together had received in any year—as the model: build useful financial products, accumulate wealth, then demand change.
Meme coins are loud but, in his view, do not contain especially serious money relative to crypto’s larger financial system. Their real cost is reputational: many outsiders’ only exposure is fraud-like behavior, so the entire technology becomes “tarred” by the most visible casino.
Adoption is generational and often invisible. Borrowing the line that science advances “one funeral at a time,” blurr expects younger users to treat digital value rails as normal; the strongest victory would be people using crypto under the hood without knowing or caring that they are doing so.
14. Risk avoidance is shrinking opportunity while technology raises the stakes
blurr attributes part of Europe’s technology deficit to education that systematically suppresses risk. He contrasted a 16-year-old who wants to start a bank and receives encouragement with the response he associates with Great Britain: “What are you doing? Stop this.” The resulting company map is the United States’ mass of bubbles versus Europe’s rare Spotify.
Conventional labor no longer buys the freedom it bought prior generations: its value is “tunneling to the center of the Earth.” For young people, school and college followed by compliant employment work only through narrow channels; creating something uncertain may be mandatory even when some coin flips end badly.
Thread Guy framed civilization as repeated coin tosses whose frequency rises with technological interconnection: food stops reaching New York, one dependency fails, or an AI risk becomes real. blurr agreed that greater interconnectedness creates more risks. He separately described himself as a “doomer” after years of travel and exposure to the world’s worst conditions, while cautioning that most people should not adopt that burden.
Thread Guy’s counterargument was that catastrophe has not happened and “the game still lives.” blurr said he did not see an obvious holistic rebuttal. He argued that markets will build data centers despite local political resistance because “the market does what the market wants,” leaving few effective levers to halt the process.
15. A $13 million on-chain incident closed one pirate era without ending the game
blurr described a $13 million incident involving an Aave position in aTokens, with MKR against dollars. He had borrowed against the deposit, and the position sat almost exactly at the solvency boundary, so any adverse oracle move could trigger liquidation while he investigated an unfamiliar transaction and signature.
He could inspect MetaMask but did not know whether canceling an approval would prevent the action from being approved again. His instinct was to stabilize the position, take what he needed from the address and investigate afterward. “In the end, they got me.”
Thread Guy called blurr an “on-chain pirate,” but blurr thinks the lone pirate-developer archetype has faded. Young developers who once shipped experiments directly are now intercepted by a large venture-capital class—“Welcome to YC, buddy”—reducing the supply of strange public infrastructure that pirates once exploited.
Thread Guy described the remaining opportunity as a dark forest in which participants learn obscure contracts and hide their edge before others notice it. blurr agreed that a whole game remains, though the particular style he practiced is smaller than it once was. The game evolves, even if the old players eventually accept that “the Lost Boys in Peter Pan should return home.”
Verification Notes
- The transcript is unclear on whether the Lighter experiment involved approximately 3,400 participants or approximately 3,400 tokens burned; the digest preserves both attributions without resolving them.
- The transcript does not establish a precise initial token-emission rate or a token-per-dollar formula.