Who Sets the Price? How Standom's Market Maker Works

When you back Yes on a take, someone has to be on the other side. On Standom that someone is not a bookmaker setting a line, and it is not another player who happened to want the opposite at exactly the same moment. It is an automated market maker — a formula that will always quote you a price, and whose behaviour is fully determined by how many Stars have gone each way.

This article explains what that formula does, why we chose the one we did, and what it means for you when you are deciding how much to commit.

The problem a market maker solves

The obvious way to run a prediction market is an order book: buyers post what they will pay, sellers post what they will accept, trades happen where the two meet. That works beautifully on a stock exchange with thousands of participants per second.

It works terribly for a market about whether a specific film crosses a specific number in its opening weekend. There might be eleven people in the world interested in that question at 2 AM on a Wednesday. Nine of them want the same side. An order book leaves them all staring at an empty screen.

An automated market maker fixes that by always being willing to take the other side. You never wait for a counterparty. The price you are quoted is a function of the current state of the market, and your own participation changes that state.

What the formula actually does

Standom uses a liquidity-sensitive logarithmic market scoring rule — LS-LMSR, if you want to look up the literature. Three properties matter, and you can reason about all three without touching the mathematics.

The two sides always sum to 100. Yes at 71 means No at 29, by construction. There is no gap between them where a house margin lives. This is worth saying plainly, because it is the opposite of how a traditional betting line works: there, the two sides deliberately sum to more than 100 and the surplus is the operator's cut.

Backing a side raises its price. Commit Stars to Yes and Yes gets more expensive — for you, as you go, and for everyone after you. That is not a penalty. It is the mechanism by which the market aggregates belief: the price is high precisely because people committed to making it high, and it carries information for the same reason.

Liquidity grows with the market. This is the "liquidity-sensitive" part, and it is the reason we use this variant rather than the simpler original. In the plain version, you have to pick a liquidity parameter up front — how much conviction it takes to move the price. Pick it too low and a new market swings wildly on tiny volume. Pick it too high and a market that genuinely should move stays frozen. Neither is knowable in advance for a market that might attract ten players or ten thousand.

The liquidity-sensitive version derives that depth from the market itself. A fresh take is thin: modest positions move it a lot, which is correct, because a fresh take genuinely does not know much. As Stars accumulate, the same position moves it less. The market gets harder to shove around exactly as it earns the right to be taken seriously.

Slippage, and why your big call costs more than the sticker price

Because each Star you commit moves the price, a large position does not execute at the price you saw. It executes across a range, starting at the quoted price and ending higher. Your average is somewhere in between.

This surprises people, so it is worth being blunt about it: slippage is not a fee. No one collects it. It is the arithmetic of asking a market to give you more of one side than it currently holds cheaply. The deeper the market, the smaller the effect.

Two practical implications:

• In thin markets, commit in pieces. Two positions taken hours apart on a new take will usually average better than one large position taken at once, because the market has time to attract the other side in between.
• Depth is a feature you can shop for. A heavily traded take will absorb a large call at close to the quoted price. A brand-new one will not. If you have a strong view and a lot of Stars, the deep market is worth more to you than a slightly better headline number on a thin one.

Why one player cannot simply buy the outcome

An obvious attack on any market maker: what stops someone with an enormous Star balance from pushing a price to 99 and sitting there?

Three things.

It costs more the further you push. Moving a price from 50 to 60 is cheap. Moving it from 90 to 99 is brutally expensive, because you are buying an increasingly thin slice of remaining probability. The cost curve is convex on purpose.

You are wrong out loud. A price pushed far from reality is an invitation. Anyone who disagrees now gets to take the other side at a price better than it was before you arrived. In a market with active players, a distorted price is a gift to everyone else.

There are caps. We limit how much any single account can commit to one side of one take. This is a deliberate constraint on the pure mathematics, and we accept the small loss of theoretical elegance for it: a market that one very large account can dominate is not a market, it is a scoreboard for that account. The cap keeps prices meaningful for everyone else.

We have also had to close specific exploits along the way — patterns where the sequencing of positions, rather than any view about the world, produced a guaranteed gain. Every one of those we found got fixed, and the general rule we apply is simple: if a strategy makes Stars without expressing any belief about the outcome, it is a bug in our market and not a clever play.

Bounded loss, and what the platform is actually risking

A market maker that always quotes a price is, in principle, exposed. The saving grace of this family of formulas is that the maximum it can lose across a market is bounded and known in advance, as a function of the market's depth. It cannot spiral.

That property is why this design is used for prediction markets rather than something simpler. It lets us open a market on a question nobody has ever traded, with no idea whether it will attract eleven players or eleven thousand, and know precisely what the worst case looks like.

What this means when you play

Four things worth internalising:

1. The price is the crowd, not the house. Nobody is setting a line against you. If you think it is wrong, the disagreement is with other players.
2. There is no built-in margin. Yes and No sum to 100. Your edge is entirely about whether your probability estimate beats the crowd's.
3. Early positions in a market you understand are the highest-leverage ones — the price is most movable when the market knows least, which is also when you are most likely to know something it doesn't.
4. Size with the market's depth in mind. The same number of Stars is a small nudge in a busy market and a shove in a quiet one.

Understanding the mechanism will not tell you which way a take resolves. It will stop you from misreading your own results — from blaming a bad fill on a rigged market when it was slippage in a thin one, or from mistaking a price you moved yourself for a price the crowd agrees with.