Prediction‑market scoring gaps

Prediction markets turn real-world questions into $1 contracts: a winning share pays $1, a losing share pays $0, and the trading price is the crowd’s odds. On Polymarket US, the YES and NO sides of one market are designed to add up to $1.00. (docs.polymarket.us) That arithmetic is why traders pay close attention when related markets drift apart. In examples circulating on X this week, traders highlighted setups where a YES contract traded near $0.60 while a linked NO contract elsewhere sat near $0.35, leaving a gap that did not line up cleanly with a one-dollar payout. (x.com) On Polymarket, each market is a binary question with separate YES and NO outcome tokens, and related questions can also be grouped inside a larger event. A presidential event, for example, can contain separate yes-or-no markets for several candidates rather than one single winner-take-all contract. (docs.polymarket.com) On Kalshi, contracts also trade between 1 cent and 99 cents and settle at $1 if the trader’s side wins. Kalshi says most markets settle within a few hours after the outcome is known, while Polymarket resolves markets through the UMA Optimistic Oracle process before winners can redeem tokens for $1. (news.kalshi.com) (help.kalshi.com) (docs.polymarket.com) Those mechanics create room for temporary mismatches even when two contracts look similar on screen. Traders are often comparing different venues, different rulebooks, different settlement clocks, or different market structures that are only approximately linked rather than perfectly opposite. (help.kalshi.com) (docs.polymarket.com 1) (docs.polymarket.com 2) The spread is not automatically free money. Polymarket charges taker fees in many categories, Kalshi charges transaction fees on expected earnings in many markets, and both platforms use order books where the posted bid and ask can move before a trader gets filled. (docs.polymarket.com) (help.kalshi.com 1) (help.kalshi.com 2) Some traders build software specifically to watch for these gaps. Public tools and code repositories now advertise scans for cases where equivalent contracts diverge across Polymarket and Kalshi, or where bundled prices inside one platform do not add up the way a fully efficient market would imply. (github.com) The alerts matter because prediction markets have grown into a larger, faster retail trading niche in 2026, with Kalshi expanding as a federally regulated U.S. exchange and Polymarket continuing to market itself as the world’s largest prediction market. Bigger participation can deepen liquidity, but it also means pricing errors are spotted and traded more quickly. (news.kalshi.com) (docs.polymarket.com) What traders are really measuring in these moments is not just a bad quote, but disagreement: between two books, two user bases, or two ways of defining the same event. Until the contracts resolve to $1 or $0, those gaps are the market’s clearest sign that the crowd has not settled on one price. (docs.polymarket.us) (docs.polymarket.com)