Subscribe to Bankless or sign in

gm Bankless Nation,
A billion people play the lottery every year, and the house always wins. In today's essay, David Hoffman makes the case for Megapot, an onchain lottery built on
Base that finally lets you be the house.
Thanks for being a Premium subscriber,
luma 🫡
Sponsor: MegaETH — Crypto has new apps, finally.

A billion people play the lottery every year. Half of American adults buy a ticket annually. Eighty percent of UK adults do the same. It's the most universally played game on the planet - and yet, despite the fact that the internet reshaped virtually every other industry on Earth, lotteries have barely changed in 50 years.
That’s the motivating force behind Megapot - a startup building a lottery at internet scale.
I think Megapot is pretty cool, and the vision behind Megapot has a lot of microcosms of so many reasons why
Ethereum is cool: disintermediation, democratization, and coordination, all made possible via smart contracts.
Patrick Lung, Megapot Founder, understands the potential scale of what he’s building, while also simply being an all-around nice dude with a sharp sense of what it takes to find product market fit.
So, naturally, we backed him at Bankless Ventures.
What is Megapot?
Megapot is an on-chain lottery protocol built on Base. Tickets cost $1, draws happen every 24 hours, and there’s one single global prize pool for ticket purchases all over the world.
The prize pool itself is on Base, but you can buy Megapot tickets from any EVM chain. They’ve got some intents-based magic in the backend to make it all seamless. You send assets from whatever chain you're on, and you get a ticket back.
Better Economics
The traditional lottery is a bad deal for the player.
When you buy a $1 Powerball ticket, only ~50 cents actually makes it into the prize fund. Retailer commissions, state government programs, administrative overhead, and operator profit all take a cut before the money gets put into the prize pool.
With a $1 Megapot ticket…
- ~70 cents goes back to the player in expected value
- ~20 cents goes to liquidity providers (LPs), who backstop the prize pool
- ~10 cents goes to whoever referred the ticket purchase
Megapot returns roughly 40% more value per dollar to the ticket buyer, and the other 30 cents that still stays inside the Megapot system.
40% better expected ROI for ticket buyers is a stark improvement, but let's be real - it’s not like lottery ticket purchasers are economically rational anyway. It’s a tailwind, but it's not a 10x tailwind.
Rather, it’s the latter part that’s critical - instead of 30% of the ticket disappearing into other pockets, that value gets distributed to two key players fundamental to Megapot's growth: LPs and Distributors.
The LPs: Be the House
This is where Megapot gets philosophically interesting.
Traditional lotteries have a chicken-and-egg problem: you need a big jackpot to sell tickets, but you need to sell tickets to build a big jackpot. Megapot solves this with liquidity providers. LPs deposit capital into the prize pool, effectively backstopping the jackpot from day one. LP’s supply insurance - if a player hits, LPs pay it out.
In return, LPs earn the 20% fee from every ticket sold. In Megapot's first week of operation, LPs were earning approximately 73% APY on their deposits. That's a real yield derived from real value - not from token emissions or recursive DeFi loops.
There's variance, of course. If someone hits the jackpot, LPs take a loss on that draw. There’s nothing preventing a jackpot being hit multiple times inside of a week - a scenario in which the LPs will be down bad. It’s obviously not zero risk - LPs can lose principal.
But the system is designed such that over time, the math works in the LP's favor. Every player who buys a ticket is voluntarily accepting 70 cents on the dollar. Over a long enough time horizon, the house wins.
As an investor, I find this to be a very attractive high-risk yield account.
With Megapot, it’s permissionless to be the house.
The Distributors: Be the 7-11
Megapot's third economic participant is the referrer. Every ticket purchase includes a 10% fee that goes to whoever distributed that ticket, whether that's a website that integrated Megapot's protocol, a Telegram bot, or just someone who shared their referral link.
Traditional lotteries pay gas stations 6-8% for selling tickets. Megapot pays 10%, and it's baked into the protocol. You can't sell a ticket without paying the referral fee.
This is how Megapot plans to scale without needing to build 160 different versions of its app. If someone in Indonesia wants to build a lottery front-end tailored to their local market, they can build on Megapot's open protocol, earn the same 10% fee that Megapot's own front-end earns, and contribute volume to the same shared prize pool.
The protocol doesn't know or care who's sending the volume. It just pays whoever referred the ticket.
Want to buy a Megapot ticket? USE MY REFERRAL CODE! 😊
The Flywheel: What Happens When LPs Never Leave?
The economics of Megapot's prize pool get interesting when you model what happens if LP capital simply stays put.
If LPs never withdraw, those earnings stay in the pool, compounding its size over time.
Megapot launched with a $1.1 million prize pool and a jackpot of roughly $220,000 (the jackpot is approximately 20% of the total prize pool, since the pool must be capitalized to cover every possible winning combination, not just the top prize).
Here's what happens to the pool and jackpot at different daily ticket volumes, assuming LPs never withdraw their earnings:
(Reminder, 1 ticket costs $1)
At 10,000 tickets/day (roughly the launch pace implied by the 73% APY):
At 100,000 tickets/day (moderate global traction):
At 1,000,000 tickets/day (approaching real lottery scale):
A few notes here:
Yields compress as the pool grows. If LPs earn a fixed 20 cents per ticket against an ever-larger base, APY naturally shrinks - from 66% at launch to 28% by year two at 10K daily tickets. Early LPs capture the best returns, just like early liquidity providers on any DeFi protocol.
This model is conservative. It holds ticket volume constant, ignoring the central flywheel: bigger jackpot -> more ticket demand -> faster pool growth -> bigger jackpot. Powerball drawings above $500 million trigger exponential surges in ticket sales. If Megapot exhibits even a fraction of that behavior, the growth trajectory looks more like the 100K or 1M rows than the 10K row.
Obviously some LPs will leave. I just find it useful to hold that variable constant for demonstrative purposes. While some LPs exit the system, others will also join too, especially as ticket sales increase.
This is the reflexive loop at the heart of Megapot’s design - capturing this phenomenon will be the difference between Megapot becoming a global institution or a failed startup:
- LPs deposit capital.
- That capital creates a big enough jackpot to attract players.
- Players buy tickets.
- LP earnings grow the pool.
- The jackpot gets bigger.
- More players show up.
- More LPs deposit to capture the yield.
- The pool grows further.
Each participant in the system, the player, the LP, and the referrer, reinforces the incentives of the other two.
So the fundamental question is: can the Megapot team get this flywheel to spin before momentum dies?
The Moral Case
There's been a growing societal conversation about the "hyper-gamblification" of everything - sports betting apps flooding your phone with notifications, prediction market platforms marketing escape from poverty through clever trades. Much of this criticism is warranted.
Megapot offers something structurally different. Yes, it's a lottery, and yes, buying a ticket is negative expected value for the player.
But Megapot offers a strict improvement on the status quo:
- 40% better expected value for ticket purchasers.
- Retail has the option to be the house.
- The elimination of non-productive rent seekers.
The question isn't whether people will play lotteries. A billion people already do. The question is whether the system that serves them should be opaque, extractive, and geographically siloed, or transparent, economically efficient, and globally accessible.
Megapot is Ethereum's answer.
📍 Disclosure: Bankless Ventures is an investor in Megapot.

We're past "in it for the tech" or "in it for the money."
MegaETH is bringing you products worth using, powered by USDM.
📈 The Asset
- Bitmine purchased another +70k ETH last week
- The Ethereum Foundation staked +20k more ETH
- Forgotten ETH opened for old ETH recoveries
🏛️ The Protocol
- cart00n.eth shipped a map of Ethereum’s earliest smart contracts
- Justin Drake co-authored a paper with Google Quantum AI on cracking Bitcoin + Ethereum signatures via quantum computing
📱 The Apps
- Aave unveiled its V4 protocol on mainnet
- Asphodel is coming to Ethereum via the Kamigotchi creators
- Circle is rolling out Circle Wrapped Bitcoin
- Clanker is launching the Clanker Ecosystem Fund
- Cloudfare intro’d EmDash, a CMS ft. support for Base x402 payments
- Coinbase secured conditional OCC charter approval
- Decentraland launched on Epic Games and Google Play
- Ethervista is looking to be acquired
- Evm.now, a slick new block explorer, has arrived
- Infinex added support for Lighter perps
- Paragraph debuted Publish, an onchain Gumroad friendly to agents
- Safe unveiled Safenet, a self-custodial defense protocol
- ScopeList will continue developing Tally’s gov platform
- Uniswap Foundation projects runway through Jan. 2027
🤫 The Privacy Stack
- Ethereum researchers proposed Private Information Retrieval for private onchain reads
- Alpha Network, Aztec’s L2 centered around private smart contracts, debuted
- Privacy Pools revealed its latest roadmap plans
- Safe previewed Private Safes at EthCC
🐸 The Culture
- EthCC kicked off this week in Cannes, France
- Alabama granted legal status to DUNA DAOs
- Base announced its 2026 strategy roadmap
- Shape took Protocol Guild’s 1% Pledge
💽 The Tech
- Gnosis and ZisK announced the Ethereum Economic Zone, a new L1-L2 framework focused on synchronous composability
- Linea rolled out its Yield Boost program
- StarkWare published a primer on its new circuit-based recursive proving method
- x402 Foundation will operate under the Linux Foundation umbrella

Google dropped a quantum warning aimed squarely at crypto, so Ryan and David dug into what it actually means for Bitcoin and Ethereum.
They also cover Trump's Iran speech, what 3 more weeks of conflict could mean for markets and crypto, the lessons of the $285M Drift hack on Solana, Aave V4 going live, the Ethereum Economic Zone, and more.
Tune into this week’s Rollup to catch it all! 👇