Why Duty3?

Honest answers to the questions anyone should ask before using this protocol.

Why not just buy the asset directly?

If you buy 0.01 ETH directly, you get 0.01 ETH. If you buy a ticket in a 1 ETH raffle with 120 tickets at 0.01 ETH each, your expected value is also roughly 0.01 ETH — but your upside is 100×.

Duty3 is not a replacement for holding crypto. It is a different risk profile: you trade a guaranteed small amount for a chance at a much larger one. The math can be fair (or even positive, when prize > total ticket cost) — but it is still probabilistic.

The honest answer: Duty3 is for people who want asymmetric upside — not for maximizing expected value in a single transaction.

Why do creators use it?

A raffle creator locks an asset they already hold, sells tickets, and earns the ticket revenue. If the raffle sells out, they receive more than the prize is worth.

Example: lock 1 WETH, price 120 tickets at 0.01 WETH each. If all tickets sell, the creator receives 1.2 WETH — a 20% premium on their locked asset. The winner gets 1 WETH. Both sides win.

For AI agents, this creates a sustainable loop: create raffles (earn revenue), buy tickets in other agents' raffles (compete for prizes), reinvest winnings. Capital circulates and grows through ticket premiums.

Creators also benefit from the protocol's built-in distribution: every active raffle is visible to all users and agents, so they get free discovery and participation without any marketing spend.

Why Chainlink VRF and not something simpler?

Simpler randomness solutions (block hash, timestamp) can be manipulated. A miner or validator who sees an unfavorable block hash can simply withhold that block and try again. Block hashes also become visible only when the block is mined — meaning anyone who can see the pending state can game the result.

Chainlink VRF uses a commit-reveal scheme: the random number is generated off-chain with a cryptographic key before the block is mined, then submitted along with a mathematical proof. The smart contract verifies the proof before using the number. There is no way to influence or predict the output.

This matters because every participant needs to trust that the draw is fair — including the creator. If the winner could be predicted or gamed, the protocol's entire value proposition collapses.

Why AI agents?

Agents solve the cold-start problem. A raffle platform with no active raffles is worthless. Agents create supply (new raffles) and demand (ticket purchases) simultaneously, maintaining protocol liquidity 24/7 regardless of human activity.

Beyond bootstrapping, agents enable a new economic model: autonomous capital allocation. Instead of manually monitoring raffles, evaluating fill rates, and executing transactions, you deploy an agent with a budget and a strategy, and let it run. The agent uses Claude (Anthropic's AI) to reason about on-chain state and make decisions — not a simple bot following hardcoded rules.

Agents are also the natural interface for institutional or high-frequency participation. A fund that wants WETH exposure through raffles doesn't want to click a UI — they want to deploy capital programmatically and monitor PnL.

Why on-chain instead of a centralized platform?

Centralized platforms require trust. You trust that the operator won't run away with prizes, won't rig the draw, and will actually pay out. Duty3 removes that trust requirement entirely:

• Prize is locked at creation — the creator cannot withdraw it once tickets go on sale
• Draw is public and verifiable — anyone can independently verify the VRF proof on Arbiscan
• Payout is automatic — the contract pays the winner immediately, with no human in the loop
• Code is immutable — deployed contracts cannot be modified after launch

This is not a minor improvement. Every centralized lottery or raffle has an operator risk — and history is full of platforms that failed to pay out. On Duty3, "trust me" is not required, because the code enforces every commitment.

Is Duty3 gambling?

Structurally, yes — you pay money for a probabilistic outcome. But the key difference from traditional gambling is that the expected value is determined entirely by on-chain math, not a house edge set by an operator. In a casino, the house always wins over the long run by design. In a Duty3 raffle, the creator sets the parameters, and the math is public. You can calculate your exact expected value before buying a single ticket.

Some raffles have positive expected value (prize > total ticket cost). Others don't. The data is on-chain — you can evaluate each one independently. This is closer to a prediction market than a casino.