How Does Polymarket Work?

Conditional tokens (CTF): binary outcome tokens representing YES/NO positions on future events, traded via order book

Polymarket uses the standard CTF framework where each market has complementary YES/NO tokens that sum to $1. Users trade these tokens on a CLOB (central limit order book) to speculate on event outcomes. CTF is a battle-tested standard from Gnosis.

UMA Optimistic Oracle: decentralized oracle resolving market outcomes via economic incentive system

Polymarket uses UMA's Optimistic Oracle V3 for market resolution. Proposers submit outcomes, which are accepted unless disputed within challenge period. Disputes escalate to UMA tokenholders. This creates a trust-minimized resolution mechanism but introduces oracle manipulation and delay risks.

Order book market making: professional market makers provide liquidity via limit orders on centralized order book

Polymarket operates a traditional CLOB where market makers post bid/ask quotes. Polymarket processes $786M weekly volume with tight spreads on major markets. The order book is centralized (off-chain matching) but settles on Polygon for final execution.

Polygon L2 settlement: all trades settle on Polygon PoS for low-cost, fast finality

Polymarket chose Polygon for its low transaction costs and fast block times, critical for high-frequency prediction market trading. This makes Polymarket Polygon's largest application by user activity, but also creates dependency risk.

Binary market resolution: each market resolves to exactly YES (1) or NO (0) based on real-world outcome

Standard prediction market mechanism where markets resolve to single outcome. Winners receive $1 per share, losers receive $0. Clean economic model but creates winner-take-all dynamics.

Market maker incentives: Polymarket subsidizes market makers to maintain tight spreads and deep liquidity

Polymarket reportedly pays market makers directly to maintain liquidity, differentiating from pure fee-based models. This enables tighter spreads than competitors but creates platform cost structure and potential market maker dependency.

In markets with $50M+ open interest, the economic incentive to manipulate oracle resolution exceeds UMA's dispute bond requirements, enabling profitable oracle attacks that incorrectly resolve markets. This risk materialized June 2026: UMA whale tokenholders (two wallets earning $300-370K each) controlled resolution of the Strategy Bitcoin sale market ($80M+ volume) against majority expectations, with resolution based on narrow rule interpretation disputed by the losing side. A separate $16M dispute in April 2026 and over 1,150 disputed markets YTD in 2026 confirm this is a structural, recurring failure mode — not isolated incidents.

When correlated markets (e.g., US election markets) resolve simultaneously, winners rushing to redeem and exit create liquidity crunch, while losers may dispute multiple resolutions, freezing capital

Polymarket's off-chain order book and frontend infrastructure are centralized components; if the matching engine goes down or a third-party frontend vendor is compromised, the platform becomes unusable or user funds can be drained. This risk partially materialized June 25–27, 2026: a compromised third-party vendor injected malicious code into Polymarket's live site, draining approximately $3.1M from ~15 user wallets before detection and containment.

All Polymarket funds are locked on Polygon; a Polygon bridge exploit, network halt, or censorship event would freeze all user funds. Polygon represents single point of failure for entire $414M TVL

If Polymarket cuts market maker incentives or faces regulatory pressure, professional LPs may exit, causing spreads to widen dramatically and platform becoming unusable for price discovery