The short version
Polymarket and sportsbooks both price the likelihood of outcomes — but they do it in fundamentally different ways. Polymarket is a prediction market where you buy and sell shares; sportsbooks set odds based on oddsmaker models and market movement.
When you see a price difference between the two, it doesn't automatically mean an opportunity. It might mean the markets are pricing different events, charging different fees, or reflecting different participant pools.
Side-by-side comparison
| Dimension | Polymarket | Sportsbook |
|---|---|---|
| Price format | Share price (0–1) | American or decimal odds |
| Probability source | Market consensus (traders) | Oddsmaker + market action |
| Vig / margin | ~1% on-chain taker fee | Embedded in odds (5–10%) |
| Settlement | On-chain, rules vary by market | Standard sportsbook rules |
| Liquidity | Concentrated in popular markets | Deep for major events |
| Access | Global (non-US mostly) | US states vary by operator |
| Market types | Political, cultural, sports | Sports outcomes primarily |
| Fee transparency | Visible on each trade | Hidden in odds spread |
Fee structures
This is one of the most important differences and the most commonly overlooked.
Polymarket fees:Every trade on Polymarket pays an on-chain taker fee of approximately 1% of the notional value (subject to market conditions and CLOB parameters). This is deducted from your payout when the market resolves. There's no vig baked into the price — the fee is explicit and visible.
Sportsbook vig:The margin is embedded in the odds themselves. When DraftKings shows −180 on Brazil to win Group A, the implied probability is 64.3% — but removing the vig (via no-vig normalization) reveals the “true” implied is closer to 61.5%. The ~2.8% difference is the margin, spread across all outcomes in the market.
Settlement rules
Prediction markets and sportsbooks don't always settle on the same event definition. This is the most important reason a price difference might be real — but not exploitable.
Common mismatches for World Cup markets:
- 90-minute vs. full-time:A prediction market on “Brazil wins World Cup” may settle on the trophy being awarded, which includes extra time and penalties. A sportsbook outright may do the same, but not always — some sportsbooks use 90-minute result only for certain markets.
- Outright vs. progression:“Brazil wins Group A” on a sportsbook may mean finishing top of the group outright. A prediction market with the same name might settle on any qualification scenario where Brazil finishes first by any method.
- Draw vs. tie-break: Head-to-head markets on two teams may have different definitions of what happens if they draw in a group stage — some sportsbooks void the market, others grade on goal difference.
The market diagnostics dashboard flags which mapped markets have settlement-rule differences. Markets where the rules don't match are labeled “not comparable” — the price difference is real, but it doesn't reflect a view on the same outcome.
Liquidity and price impact
Sportsbooks have decades of market-making infrastructure. DraftKings and FanDuel can offer deep liquidity on World Cup matches because they have massive balance sheets and established market-maker relationships.
Polymarket liquidity is more concentrated. The top World Cup markets — like World Cup winner outrights — may have $100k+ in open interest. Smaller markets, like individual group stage matches weeks ahead of time, may have $5k–$20k. A large trade in a thin market can move the price significantly.
Thin liquidity on Polymarket means:
- Wider bid-ask spreads (paying more to enter, getting less to exit)
- Higher sensitivity to large trades and news events
- Prices that may not reflect the underlying probability as accurately
The dashboard marks markets below $10k in Polymarket liquidity as “low liquidity” — treat price differences in those markets with more caution.
Participant pool differences
Sportsbook bettors are predominantly US-based (for DraftKings/FanDuel) and tend to be fans with skin in the game on game day. Prediction-market traders tend to be more globally distributed, more crypto-native, and may have different information and incentives.
These differences can create genuinely different probability estimates — not because one is wrong and one is right, but because they're responding to different information and participant bases.
How to read both together
The most useful framing is not “which market is right” but “what does the disagreement tell me about how these markets work.”
- A large, persistent gap after no-vig normalization, with comparable settlement rules and deep liquidity = worth understanding the reason behind the disagreement.
- A gap that disappears after removing sportsbook vig = the markets are actually aligned; the apparent gap was just the sportsbook margin.
- A gap in a thin, illiquid market = likely noise, not a durable read; the price could move back quickly once volume enters.
- A gap in markets with different settlement rules = not a comparison problem, it's a rule-definition problem.
The dashboard is built to surface exactly these distinctions. It normalizes for vig, flags settlement-rule mismatches, and marks liquidity quality — so you're reading the market structure, not just the price.