Prediction Markets 101

Understanding Polymarket's CLOB: How the Order Book Works

mBotopoly Team··11 min read

Understanding Polymarket's CLOB: How the Order Book Works

Polymarket's Central Limit Order Book — the CLOB — is the engine that powers every trade on the platform. Whether you're buying YES shares on a political event or selling NO shares on a crypto market, your order passes through this system. Understanding how it works is not optional for serious traders. It directly affects your execution quality, your costs, and ultimately your returns.

This guide explains the CLOB from the ground up: what it is, how orders are matched, what the bid-ask spread tells you, the available order types, how liquidity depth works, and why all of this matters significantly more once you start automating.

What Is a Central Limit Order Book?

A Central Limit Order Book is a system that matches buyers and sellers based on price and time priority. It is the same mechanism used by the NYSE, NASDAQ, and every major financial exchange in the world. Polymarket adopted this model rather than using an automated market maker (AMM) like Uniswap, and the choice was deliberate.

In an AMM system, you trade against a liquidity pool governed by a mathematical formula. The price is determined by the ratio of assets in the pool. This works for fungible token swaps, but it introduces structural problems for prediction markets: impermanent loss for liquidity providers, limited price precision, and poor capital efficiency.

The CLOB solves these problems. Instead of a formula, real buyers and sellers post their prices and quantities. The order book is a live record of every open order, organized by price level. When a new order comes in, the matching engine checks if there's a corresponding order on the opposite side at a compatible price. If there is, the trade executes. If not, the order sits on the book and waits.

Polymarket's Hybrid Architecture

Polymarket's CLOB is technically off-chain — the order matching happens on Polymarket's servers, not on the Polygon blockchain. This is necessary for speed. On-chain order books are too slow and expensive for active trading. However, the settlement is on-chain. Once an order is matched, the token transfer occurs on Polygon, and the trade is recorded on the blockchain.

This hybrid design gives Polymarket the speed of a centralized exchange with the settlement guarantees of a blockchain. Your shares are real on-chain tokens, but the order matching happens at sub-second speed off-chain.

The Bid-Ask Spread

The bid-ask spread is the difference between the highest price a buyer is willing to pay (the bid) and the lowest price a seller is willing to accept (the ask).

Suppose you're looking at a market where:

  • Best bid: $0.62 (the highest price someone is willing to pay for YES shares)
  • Best ask: $0.64 (the lowest price someone is willing to sell YES shares for)
  • Spread: $0.02
This $0.02 spread is the cost of immediacy. If you want to buy YES shares right now, you'll pay $0.64 (the ask). If you're willing to wait, you can place a limit order at $0.62 (the bid) and hope someone sells to you.

What the Spread Tells You

Tight spreads ($0.01-$0.02) indicate a liquid, competitive market. Multiple market makers are actively quoting, and the cost of trading is low. This is typical of high-volume markets like major political events. Wide spreads ($0.05-$0.15+) indicate a thin market. Few participants are providing liquidity, and the cost of trading is high. You'll see this in niche markets with low volume. Wide spreads can also indicate uncertainty — if a market-moving event is imminent, liquidity providers often widen their spreads to protect against adverse selection. Asymmetric spreads — where the bid-ask spread is tighter on one side than the other — can indicate directional bias among market makers.

For traders, the spread is a direct cost. Every time you cross the spread (by placing a market order), you're paying the spread as an implicit fee. This is why limit orders are almost always preferable for cost-conscious traders.

Order Types on Polymarket

Polymarket supports several order types through its CLOB. Understanding each one gives you more control over your execution.

Limit Orders

A limit order specifies the price and quantity you want to trade. It only executes at your price or better.

  • Buy limit: "I want to buy 100 YES shares at $0.60 or less."
  • Sell limit: "I want to sell 50 NO shares at $0.45 or more."
Limit orders that don't immediately match are placed on the order book, where they add liquidity. These are "maker" orders, and Polymarket charges 0% fees on them. This is a significant incentive — you're being rewarded for providing liquidity to the market.

Market Orders

A market order executes immediately at the best available price on the order book. You specify only the quantity (or the amount of USDC to spend), and the matching engine fills your order against the existing resting orders.

Market orders are "taker" orders because they remove liquidity from the book. Polymarket charges a small fee on taker orders.

The risk with market orders is slippage. If you want to buy 1,000 YES shares and only 200 are available at the best ask, the remaining 800 will fill at progressively worse prices as they consume deeper levels of the order book. In thin markets, this can result in an average execution price far worse than the quoted price.

Good-Til-Date (GTD) Orders

A GTD order remains on the book until a specified expiration date. If it hasn't been filled by that date, it is automatically cancelled. This is useful for positions where your thesis is time-sensitive:

  • "I want to buy YES at $0.45, but only if it gets there within the next 48 hours. After that, the information environment will have changed."
GTD orders prevent stale orders from sitting on the book indefinitely, which is important for risk management.

Fill-or-Kill (FOK) Orders

A FOK order must be completely filled immediately or it is cancelled entirely. There is no partial execution. This order type is used when partial fills are unacceptable — for example, when you need a specific position size for a hedging strategy.

If you place a FOK order for 500 shares at $0.55 and only 300 shares are available at that price, the entire order is cancelled. Zero shares are bought.

Good-Til-Cancelled (GTC) Orders

A GTC order remains on the book until it is either filled or manually cancelled. This is the default behavior for most limit orders on Polymarket. GTC orders are appropriate when you have a price target and are patient enough to wait for the market to come to you.

Liquidity and Depth

Liquidity depth describes how many shares are available at each price level in the order book. A "deep" order book means large quantities are available at prices near the current market price. A "shallow" order book means small quantities, and larger orders will experience significant slippage.

Reading the Order Book

Visualize the order book as two columns:

Bids (buy orders):

| Price | Quantity | |-------|----------| | $0.62 | 5,000 | | $0.61 | 8,000 | | $0.60 | 12,000 | | $0.59 | 3,000 |

Asks (sell orders):

| Price | Quantity | |-------|----------| | $0.64 | 4,000 | | $0.65 | 7,000 | | $0.66 | 10,000 | | $0.67 | 2,000 |

If you place a market buy order for 4,000 shares, you'll fill entirely at $0.64. If you want 11,000 shares, you'll fill 4,000 at $0.64 and 7,000 at $0.65, for an average price of approximately $0.647. If you want 25,000 shares, you'll sweep through multiple levels, paying progressively more.

Depth as a Signal

Order book depth conveys information beyond just execution cost:

  • Thick bid support: Large quantities on the bid side suggest strong buying interest. Traders believe the current price is fair or cheap.
  • Thick ask resistance: Large quantities on the ask side suggest sellers are lined up. This can cap upward price movement.
  • Imbalanced books: Significantly more depth on one side than the other often precedes a price move toward the thicker side (counterintuitively — the side with more resting orders can absorb more flow in that direction).
However, order book depth can be misleading. Orders can be cancelled instantly, and sophisticated participants sometimes place large orders they intend to cancel before execution (a practice known as spoofing, which is illegal in regulated markets but difficult to enforce in crypto markets). Treat order book depth as one signal among many, not as ground truth.

How Bots Interact with the Order Book

Automated trading systems are the dominant participants in Polymarket's CLOB. Understanding how they operate reveals the dynamics that manual traders are competing against.

Market Making Bots

Market making bots continuously post bid and ask orders on both sides of the order book. Their goal is to capture the spread — buying at $0.62 and selling at $0.64, pocketing $0.02 per share minus any adverse selection costs.

These bots typically:

  • Quote on dozens or hundreds of markets simultaneously
  • Adjust prices in real time based on market conditions, volatility, and inventory
  • Cancel and replace orders within milliseconds when new information arrives
  • Maintain tight spreads in normal conditions and widen them during volatility
Market making bots are generally beneficial for retail traders. They provide the liquidity that allows you to enter and exit positions efficiently. Without them, spreads would be wider and execution would be worse.

Arbitrage Bots

Arbitrage bots look for pricing inconsistencies across related markets. For example:

  • If "Event X by June" is priced higher than "Event X by December," there's a logical inconsistency (the December probability must be at least as high as June). An arb bot would buy December and sell June.
  • If the YES and NO prices on a single market don't sum to approximately $1.00, an arb bot will trade both sides to capture the guaranteed profit.
These bots improve market efficiency by eliminating obvious mispricings. For more on how this works, see our guide on how prediction market arbitrage works.

Signal-Based Bots

Signal-based bots trade directionally based on quantitative models. They might ingest news feeds, social media sentiment, polling data, or on-chain activity to generate probability estimates, then trade when their estimate diverges from the market price.

These bots are your competitors. When you spot a mispricing, a signal-based bot may have already traded on it milliseconds ago. Understanding this dynamic is important for setting realistic expectations about the opportunities available to manual traders.

Execution Bots

Execution bots optimize how large orders are filled. Instead of placing a single large market order (which would cause slippage), an execution bot might:

  • Break the order into smaller pieces executed over time (TWAP — Time-Weighted Average Price)
  • Place limit orders and wait for natural flow to fill them
  • Monitor the order book and opportunistically hit favorable price levels
This is the category where mBotopoly operates — optimizing execution to minimize costs and maximize fill rates.

Why This Matters for Automated Trading

If you're considering automated prediction market trading, the CLOB is your primary interface. Every strategy — market making, directional trading, arbitrage, execution optimization — interacts with the order book. Understanding its mechanics determines whether your strategy works or leaks money through poor execution.

Key implications:

Execution quality is a source of edge. Two traders with identical signals can have vastly different returns based on how they execute. Placing limit orders instead of market orders, sizing appropriately for the available depth, and timing entries to coincide with favorable order book conditions all contribute to better outcomes. Latency matters, but not as much as in traditional markets. Polymarket is not a high-frequency trading venue in the traditional sense. Sub-millisecond speed is not required. But sub-second responsiveness to new information and order book changes does provide an advantage over manual traders who might take minutes to react. Understanding order book dynamics improves your models. If your trading model generates a signal to buy YES at $0.60, but the order book shows only 500 shares available before the price jumps to $0.68, the actual expected execution price is much higher than $0.60. Your model needs to account for market impact.

The order book is the terrain. Your trading strategy is your plan. You cannot plan effectively if you do not understand the terrain.

To build your understanding of the broader context, read our guides on what prediction markets are and how prediction market odds work. Both connect directly to the order book concepts covered here.


mBotopoly handles order book execution for you — optimizing for speed and fill rate. See how →

Ready to automate your trading?

Join traders using mBotopoly to execute strategies on Polymarket around the clock.

Start trading with mBotopoly