Module 2 · Lesson 1

Order books vs. automated market makers

Key takeaways

• There are two ways to run a market: match people against each other (an order book), or match everyone against a formula (an automated market maker, or AMM). You already know the first one — Module 1 Lesson 3 was entirely about it
• An order book needs two willing people to make a trade. When no one’s there, it stalls. An AMM is always willing to trade with you, because it’s a formula, not a person — so it never stalls, but you pay for that certainty
• The deepest practical difference: in an order book, prices can move without a single trade (traders revise their offers when news breaks). In an AMM, the price only moves when someone actually trades
• Neither is “better.” Order books dominate the big, liquid markets; AMMs exist to give life to the small, thin ones no human would bother quoting. Knowing which one you’re standing in tells you what to expect from the price

Two ways to build a market

In Module 1 you learned to read a market — what a price means, why a crowd’s price can be trusted, how contracts change hands through an order book. This module goes one level deeper: not what the price says, but how the machine that produces it is built — and it turns out there’s more than one design.

Every market has the same job: bring buyers and sellers together and produce a price. But there are two fundamentally different ways to do that job, and prediction markets use both. Understanding the difference is the whole of this lesson, because the structure you’re trading in quietly shapes everything — how fast the price reacts, what it costs you to trade, and whether you can trade at all.

The one you already know: the order book

Module 1 Lesson 3 taught this in full, so here’s just the one-sentence version to anchor against: an order book is a live, two-sided list of what buyers are willing to pay and what sellers are willing to accept, and a trade happens when one side agrees to meet the other.

The crucial feature — the one this lesson turns on — is that an order book needs two people. For you to buy, someone else has to be willing to sell to you at a price you’ll accept. If nobody is offering, you wait. In a busy market that’s invisible: there’s always someone there. But in a quiet market — a niche question, an obscure event, a contract nobody’s paying attention to — the book can be nearly empty, and you simply can’t trade, because there’s no one on the other side.

That gap is the problem the second design was invented to solve.

The other one: the automated market maker

An automated market maker — an AMM — throws out the idea of matching you against another person. Instead, it matches you against a formula.

Picture a pool holding both sides of a market — “yes” shares and “no” shares — governed by a mathematical rule. When you want to buy “yes,” you don’t wait for a seller. You trade with the pool directly: you put money in, the formula hands you “yes” shares, and it recalculates the price as it does so. Buying “yes” makes “yes” more expensive and “no” cheaper, automatically, by the rule. There’s always a price, and there’s always someone — something — to trade with. The market never stalls, because a formula never goes home.

This is the key inversion. An order book waits for a human counterparty. An AMM is the counterparty, always.

Why buying pushes the price up

If you’ve watched the price climb as you spend more, you may have wondered why — why isn’t every contract the same price? The answer is the heart of how any market works, and it’s worth getting clearly.

Start with what you already know: the price is a probability. A contract at 50¢ means the market thinks the event is 50% likely. So when you buy “yes,” you’re not just acquiring shares — you’re making a claim: “I think this is more likely than the current price says.” A market’s whole job is to move toward where the money goes, so your buying nudges the probability up. Buy a little, you’ve nudged it a little. Buy a lot, you’ve said “I’m very confident,” and the price climbs to match — because now your money is part of what the market believes. This is true of every market, order book or AMM: demand moves the price, because the price is a belief and buying is a vote.

The AMM just makes this happen mechanically, inside your own order, and there’s a simple picture for it. Think of the pool as a balance between “yes” shares and “no” shares — it starts even, so the price starts at 50¢. When you buy “yes,” you take yes-shares out of the pool. Now yes-shares are scarcer and no-shares are relatively more plentiful — and scarcer things cost more. So the next yes-share costs a little more than the one before it, because each purchase you make leaves fewer behind. You are your own demand, eating through the supply, lifting the price as you go.

That’s also exactly why slippage happens, and why it’s brutal in a thin pool. A thin pool has few shares to start with, so each contract you buy removes a big fraction of what’s there — scarcity spikes fast, and the price lurches. A deep pool is stocked with shares, so the same purchase barely dents it, and the price only creeps. Slippage is nothing more than the price of making your own shares scarcer by buying them — and it’s the reason the amount you spend, and the depth of the pool you spend it in, change how many contracts your money actually buys.

There’s a neat way to feel this against something you’ve already seen. The calculator back in Module 1, Lesson 1 let you spend $100 at a fixed price of 65¢ and get about 154 contracts. Spend that same $100 in a deep AMM pool and you get almost exactly the same thing — the pool is so well-stocked that your buying barely moves the price, so it behaves like the fixed-price calculator. Now spend it in a thin pool and the result falls apart: the same $100 buys far fewer contracts, at a worse average price, because your own order shoved the price up as it filled. Same money, same question — the only thing that changed is how much liquidity was there to absorb you. That gap is the lesson.

The catch: certainty isn’t free

The AMM’s guarantee — always a price, always a counterparty — comes at a cost, and someone pays it. The formula has to hold real money to pay out whichever side wins, and that money comes from whoever funds the pool: the operator, or liquidity providers who put up capital. They take on real risk — if the event resolves against the shares they’re left holding, they lose. That funding, a subsidy staked to keep the market open, is the quiet price of a market that never stalls. An order book pushes that risk onto the traders who choose to post offers; an AMM concentrates it in whoever seeds the pool. Neither makes the risk disappear — they just put it in different hands.

How each one sets the price

This is the difference that actually changes how you trade, so it’s worth being precise.

In an order book, the price is simply the last point where two people agreed — and, just as importantly, the best offers waiting on each side. Those offers are placed by people, and people can change their minds without trading. News breaks, and the sellers yank their offers and repost higher — the price moves, and not a single contract changed hands. The order book can re-price on information alone, because it’s made of live human intentions that update in real time.

In an AMM, the price is output by the formula, and the formula only knows one thing: how many shares have been bought and sold. So the price moves only when someone trades. If news breaks and nobody has traded yet, the AMM is still quoting yesterday’s price — it has no way to know anything changed until a trader acts on it and shifts the pool. The price updates through transactions, not through intentions.

That’s a genuinely useful thing to carry into any market you look at. An order-book price is a live consensus that can jump on news; an AMM price is a lagging record of the last trade that moved it. When you see a price, knowing which kind of machine produced it tells you how much to trust that it’s current.

Not one formula, but a family

“An AMM uses a formula” is true, but it hides something worth knowing: there isn’t one formula. There’s a family of them, and the choice shapes how the market behaves. Three are worth recognizing.

The logarithmic market scoring rule (LMSR) is the one built specifically for prediction markets, by the economist Robin Hanson. Its defining virtue is that the market maker’s worst-case loss is bounded and known up front — the operator can calculate the most they could ever lose before opening the market. That safety is why LMSR became the classic choice for event contracts, and its prices always sum neatly to $1, so they read cleanly as probabilities.

The constant-product rule is the one that runs most of decentralized finance — the engine behind Uniswap and its kin. It’s elegant and battle-tested: a pool holds both sides, and their product stays constant, so buying one side automatically prices it up. But it wasn’t designed for $1/$0 event contracts, and it bleeds value badly when a price sits near 0 or 1 — exactly where prediction markets often live — which is why it’s more common in general crypto trading than in serious prediction markets.

The parimutuel model is the odd one out, and recognizing it is the point. It’s the system behind horse-track betting: there’s no live price at all while the market is open. Everyone’s money goes into one pool, and after the event the whole pool is split among whoever backed the winning side. Your “odds” are just your side’s share of the pool, and they aren’t final until the market closes — you never know your exact payout until the end. It guarantees the book always balances, at the cost of the continuous, tradeable price the other two give you.

You don’t need the math behind any of these. What matters is the recognition: when someone says “it’s an AMM,” the next question is which kind — because a bounded-loss LMSR, a DeFi constant-product pool, and a no-live-price parimutuel behave like three different animals, even though all three replace the human counterparty with a rule.

Order types: the thing an order book lets you do

There’s one more order-book concept worth naming here, because it only makes sense against the AMM contrast — and Module 1 pointed forward to it.

In an order book, you get a choice about how you trade. A market order says “fill me now, at whatever price the book offers” — you’re guaranteed to trade, but you accept whatever price is there, including the slippage if your order is big. A limit order says “here’s my price; wait until someone meets it” — you’re guaranteed your price, but not that you’ll trade at all. Size or price: you can lock one, never both. That choice exists because an order book is made of resting offers you can either take or add to.

That “take or add to” is worth sitting with, because it’s the whole shape of trading in an order book. A market order takes a price already resting in the book — you hit someone else’s offer. A limit order adds one — you post your own price and wait, which means you become part of the book: your order sits there as a resting offer until someone else takes it. And this works on both sides. When you’re buying, a limit order posts a bid — you become the bid someone can sell into. When you’re selling a position you hold, you have the same choice: take the best bid right now with a market order, or post your own ask with a limit order and wait — you become the ask someone can buy from. So the bid and ask you read on the screen aren’t the market’s prices handed down from above; they’re just the best offers other traders have posted, and every time you place a limit order, yours joins them.

An AMM barely has this distinction. You trade with the formula at the price the formula gives — there’s no book of resting orders to post a patient limit into and wait. Some venues layer limit-order-like features on top, but the native AMM experience is closer to “market order only.” The very existence of order types is a feature of the order-book structure.

Which one you’re actually trading in

Here’s the part that keeps this honest and current. It would be tidy to say “regulated dollar platforms use order books, crypto platforms use AMMs” — but that’s not how it actually shook out. The largest prediction markets, including the major crypto-native one, mostly run on order books. AMMs turned out to matter most for the long tail — the thin, obscure, just-launched markets where no human would bother posting quotes, and where a formula that’s always willing to trade is the only thing keeping the market alive at all.

So the two structures aren’t rival camps you pick a side in. They’re tools suited to different jobs: order books for deep, active markets where there are plenty of humans to match; AMMs for thin or new markets that need a guaranteed counterparty to exist at all. A single platform can use one, the other, or both in different corners.

What this means for you as a reader of prices: when you land on a market, one useful question is which machine is running underneath. A deep order book with a tight spread is a live, trustworthy consensus. A thin AMM pool is a price that will lurch if you lean on it and may be stale if news just broke. Same $1/$0 contract, same “price is a probability” idea from Module 1 — but a different engine, with different behavior you can now anticipate.

The bottom line

Two designs, one job. An order book matches people against each other: it re-prices instantly on news, it lets you choose market or limit orders, and it’s unbeatable in deep markets — but it stalls when no one’s there. An automated market maker matches you against a formula: it never stalls and it gives thin markets a pulse they’d otherwise lack — but it only moves on trades, it charges you slippage from the shape of its pool, and someone has to fund it.

You now know not just how to read a prediction-market price, but how the two machines that produce one actually work — and how to tell, from the behavior of the price in front of you, which one you’re looking at.

Next, we go from the machinery to the thing you actually walk away with: what you own when you take a position, and how it pays out — shares, contracts, and settlement.

Related lessons

Previous:

• Module 1, Lesson 3: How contracts work — The order book this lesson builds on: the bid-ask spread, liquidity, depth, and the market-versus-limit choice.

Next:

• Module 2, Lesson 2: Shares, Contracts, and Settlement — What you actually own when you take a position, and how it pays out.