Every launch is a self-resolving prediction market
Each new coin on wdyt opens a 30-minute market on one question: will it launch? This page walks through the whole mechanism, from the first buy to the payout. The numbers in every example come from the same math the program runs.
One question, two sides, one clock
Yes buys the coin. No buys a share of the pot.
Buys the coin before it exists: every Yes token becomes one coin if it launches. The price follows a curve that rises as SOL comes in on Yes, so the first buyers get about 4x as many coins per SOL as the last. The SOL on Yes is the launch's funding, and when it reaches 75 SOL the coin launches.
Buys a share of the pot, at exactly what that share would pay if the market failed right now. Every Yes buy that comes in afterwards raises the share price, so an early No can multiply its SOL while a No bought at the last second can only get its own SOL back. SOL on No never counts toward the launch. No opens once Yes reaches 22.5 SOL, 30% of the target, so No is a bet against launches that are already under way, not a free option on every new coin.
The No side opens once Yes reaches 22.5 SOL. An early No bought at 9.43 SOL a share. A No bought at the last second pays 14.54, the full value of the pot per share, so it has nothing left to win.
It resolves itself
Whichever happens first decides the market. There is no third outcome.
The outcome is the market's own state: how much SOL is on Yes, and what time the chain says it is. Nobody reports a result.
It settles the moment either condition is met, with nothing to wait for or appeal.
The program, its rules and every trade are on-chain. Anyone can verify a result, or send the resolving call themselves.
What each side gets
The program settles it on-chain the moment either happens: no oracle, no dispute window. Amounts are after the 1.25% fee.
2.09x of the 1.000 SOL you paid.
0.25x: your share of the 25% of No-side SOL that comes back. The rest goes into the coin's locked pool.
Your shares are 4% of the No side, so you win about that share of every Yes SOL that arrives after you. 30 SOL of Yes came in after you and it still fell short: that is where your profit comes from.
Why a last-second No can't snipe a failing launch
The pot is empty and only Yes can be bought until Yes reaches 22.50 SOL. The Yes side owns the pot through a starting share, so until someone buys No, any SOL that comes in on Yes would go back to Yes.
Launch day
- Both pots become liquidity. 77.87 SOL here: both pots, less the No refund (1.234 SOL), a 0.05 SOL fee that pays for creating the pool and the 1% launch fee (0.787 SOL). The pool gets proportionally fewer coins for the launch fee, so it doesn't move the opening price.
- No-side SOL makes the launch stronger. Half of it lifts the opening price (2.4% here); the other half is paired with 4.82M newly minted coins, so the pool is deeper.
- Locked forever. The liquidity position is locked permanently. Nobody, including the creator, can pull it out.
- Yes holders are paid first. Every Yes token becomes one coin, sent to its holder's wallet. Trading opens 60 seconds after the pool exists, and the coins go out in that window.
- Fees keep flowing. Every trade in the pool pays 1.25% in SOL: 0.64% to the creator, 0.2% to the coin's holders, 0.16% to the protocol and 0.25% to Meteora.
- No holders get a quarter back. 25% of the No-side SOL (1.234 SOL here) goes back to No holders, split by share. The rest is in the pool, deepening the market for the coin they bet against.
Leaving early
- No selling back to the market. SOL in the pots can't leave before the market resolves. That keeps the Yes progress honest and the No prize real.
- Sell to another trader instead. Every market has its own on-chain order book for Yes and for No. Place an order, or fill someone else's.
- Book trades never move the market. Only Yes and No buys add to the pots and to the progress toward the line.
- Taker pays the fee. The trader who fills an order pays 1.25%; the one whose order was resting pays nothing. A resting order can be filled right before an outcome is decided, so price it with that in mind.
- Ordinary tokens. Yes tokens and No shares are standard Solana tokens, so any other venue can list them too.
Fees and the creator's bond
Goes to the wdyt treasury.
Paid instantly to whichever wallet, bot or site routed the trade. With none, it goes to the protocol.
Held until the market resolves. Paid to the creator only if the coin launches; otherwise it goes to the protocol.
The launch fee, paid to the protocol once. The pool gets proportionally fewer coins, so the opening price is the same as without it. A 0.05 SOL migration fee pays for creating the pool.
After a launch, every trade in the locked pool pays 1.25% in SOL: 0.64% to the creator, 0.2% to the coin's holders, 0.16% to the protocol and 0.25% to Meteora. Payouts are free.
Posted when the market is created, so spam launches cost something.
The creator gets the whole bond back.
The bond goes to the No side, building up evenly from the open to the deadline. Each No holder earns the part that builds up after they buy, so a last-second No can't grab it. Whatever the No holders didn't earn goes back to the creator.
Why not a bonding curve?
Doubt is invisible: skeptics simply don't buy. When a coin fizzles on the curve, its holders can only sell back into it, at whatever price the curve has fallen to.
Doubt is a position with a price. The launch has to earn its liquidity, and if it doesn't, the people who called it get paid in SOL.
On a curve, skeptics are invisible. Here, the SOL on No shows how many people think a coin won't make it, and they get paid if they're right.
A coin launches only if Yes reaches its target before the clock runs out, and the SOL on No deepens its pool when it does.
A failed launch doesn't leave a dead coin behind. The market resolves, and everyone is paid in SOL, automatically.
No oracle, no committee, no dispute window. The program reads the market's own SOL and settles it.
See it live
Every open market shows its rules, its clock and what one SOL on each side pays right now.
Building an app or an AI agent? Developers