How prediction market pricing works: from shares to resolution

Key takeaways about predicition market pricing

• A prediction market share priced at $0.65 implies a 65% probability—winning shares pay $1.00, losing shares expire at $0.00.

• Prices move through a central limit order book (CLOB) where buy and sell orders are matched by price and time priority.

• Market depth determines execution quality—deep markets have tight spreads, while thin markets are prone to slippage and sharp price swings.

• In multi-outcome markets, all share prices should sum to approximately $1.00—deviations signal potential arbitrage.

• Resolution relies on oracle infrastructure with built-in dispute mechanisms, and ambiguous resolution criteria create risk independent of the underlying event.

• Information cascades, liquidity constraints, cognitive biases, and identity-driven trading can all push prices away from fundamental probabilities.

Prediction market shares represent probability claims that an event will happen in a certain way—for example, a $0.42 share means roughly a 42% implied probability. Prices shift constantly as traders buy and sell, reflecting new information and changing expectations. When a market resolves, winnings can be kept in the wallet and may be cashed out or withdrawn. Understanding predicting market pricing mechanics helps to identify when markets may be mispriced and why resolution design matters as much as the trade itself.

Disclaimer: This guide is for educational purposes only. It is not financial advice, not a solicitation, and not for UK audiences. Prediction market shares are risky and not suitable for all users.

What prediction market shares represent

A prediction market share's price reflects its implied probability. A share trading at $0.65 on a $0.00–$1.00 scale means the market collectively assigns a 65% chance to that outcome. This relationship is structural, not arbitrary—buyers at $0.65 believe the true probability exceeds 65%, while sellers at that price believe it's lower.

The payout math is straightforward. If the outcome resolves true, each share pays $1.00. A position of 100 shares purchased at $0.45 each costs $45 and returns $100—a $55 gain. If the outcome resolves false, the shares expire at $0.00. The entire $45 is lost.

This binary structure means prediction market prices distill collective probability estimates into a single, continuously updating number. For a deeper look at how binary contracts, oracles, and resolution criteria work on Polymarket specifically, see What is Polymarket? A guide to decentralized prediction markets.

How prices move: order flow and the CLOB

Prediction market prices move when new information arrives or new traders enter. A political poll, earnings report, injury announcement, or legislative development triggers immediate repricing as traders act on what they believe the data means for the underlying probability.

This price discovery happens through a central limit order book, or CLOB—the same matching engine used by traditional financial exchanges. Polymarket's CLOB matches buy and sell orders by price and time priority.

When a trader places a buy order for Yes shares at $0.65, the CLOB simultaneously displays this as a sell order for No shares at $0.35, since the two outcomes form a complete probability set that must sum to $1.00. This mirroring creates deeper liquidity and tighter spreads than separate order books would allow.

The depth of a market's order book determines how smoothly prices adjust. In deep markets—those with high volume and many participants—individual trades barely move the price, and the gap between the best buy and sell prices (the spread) stays tight. In thin markets with fewer participants, a single large order can push the price several cents, and wider spreads make entering and exiting positions more expensive. On Polymarket via MetaMask, trades settle on Polygon—keeping network fees low and execution fast relative to Ethereum mainnet—and every fill creates a permanent onchain record under self-custodial control.

For a full glossary of terms like spread, slippage, and order book mechanics, see Essential prediction market concepts and key terms.

For a comparison of how centralized platforms like Kalshi differ in structure and regulation, see Prediction market trends in 2026.

Multi-outcome markets and share pricing

Not every prediction market is binary. Many markets cover mutually exclusive outcomes where exactly one must be true at resolution—for example, the winner of a 32-team tournament or the country hosting a future Olympics.

In these markets, each possible outcome gets its own set of shares. A $0.08 share in one team implies roughly an 8% probability. A $0.12 share in another implies 12%. In a well-functioning multi-outcome market, all outcome shares should sum to approximately $1.00. When they don't, the discrepancy signals something worth investigating.

If all shares in a 32-outcome market sum to $0.98, a trader could theoretically buy one of each for $0.98 and guarantee a $1.00 payout regardless of which outcome wins—a riskless 2% return. In practice, these imbalances are corrected quickly as arbitrageurs exploit them. That self-correcting tendency is one reason prediction market prices stay well-calibrated.

What happens at resolution

Every prediction market has defined resolution criteria—specific conditions, sources, and a timeframe that determine the outcome. A well-designed market leaves no room for interpretation: "Will [candidate] be certified as the winner by [authority] by [date]?" is precise. "Will leading indicators suggest a recession?" is ambiguous and creates resolution risk separate from the underlying event.

When the outcome is known and verified, resolution is binary. Winning shares convert to $1.00. Losing shares expire at $0.00.On Polymarket, payouts are generally handled automatically after market resolution, and users may still need to withdraw or cash out funds depending on the platform flow. From there, winnings can be swapped for other assets through MetaMask Swaps or bridged to other blockchains—no withdrawal process or waiting period. Approximately 93% of Polymarket markets resolve within two hours.

Polymarket's resolution system uses predefined resolution criteria and designated data sources to settle markets, and some markets may use automated oracle-based data feeds for fast price verification.If disputed, the resolution escalates to a token-holder vote that typically takes 48–96 hours. For fast-resolving markets like 15-minute crypto price contracts, Polymarket uses Chainlink Data Streams for low-latency, automated price verification—pulling directly from Chainlink's decentralized feed infrastructure rather than relying on human proposers.

The system is trust-minimized rather than fully trustless. Economic incentives—proposers and disputers both risk bonds—are designed to make honest reporting the most profitable strategy. But resolution risk is real. High-profile disputes, including contested markets on UFO declassification and geopolitical events, have shown that ambiguous resolution criteria can lead to outcomes some participants consider incorrect. Reading the resolution rules (not just the market title) before entering a position is important for exactly this reason.

When markets misprice: biases, liquidity, and information cascades

Prediction markets have demonstrated strong forecasting accuracy in aggregate—often matching or exceeding expert predictions and statistical models. But individual markets can and do misprice, sometimes significantly, for reasons that fall into a few recurring patterns.

Information cascades occur when early traders establish a price and later participants follow it, assuming the first movers had superior information. This self-reinforcing pattern can push prices away from the underlying probability, especially in low-attention markets where few independent analysts are participating.

Liquidity constraints prevent large traders from correcting mispricings. A trader who believes a $0.55 share should be priced at $0.70 may be unable to deploy enough capital without moving the price dramatically against themselves. The mispricing persists not because the market disagrees, but because the order book can't absorb the corrective trade.

Cognitive biases are individual but systematic. Recency bias (overweighting recent events), confirmation bias (favoring outcomes that align with existing beliefs), and overconfidence (overestimating one's own accuracy) all appear in prediction market pricing patterns. When these biases cluster around a culturally or politically significant outcome, they can tilt the market price away from the fundamental probability.

Structural incentives also play a role. In some markets, buying a particular outcome functions as a cultural or political signal rather than a probability assessment. When buying activity is driven by identity rather than analysis, prices reflect sentiment rather than information.

Recognizing these dynamics doesn't guarantee profitable trades—but it does help distinguish between markets where the price reflects genuine collective intelligence and markets where structural factors may be distorting the signal. For strategies used by experienced traders to identify and act on these mispricings, see Advanced prediction market trading strategies.