A prediction-market headline number — “Newsom 26%,” “ceasefire 8%,” “recession 25.5%” — looks like a poll. It is not. A poll measures stated preference today; a prediction market measures the equilibrium price between buyers and sellers staking real money on a specific resolution. Reading one well takes about five minutes once you know what to look at.
This guide is the five-minute version. By the end, you will be able to look at any market on Polymarket, Kalshi, or Manifold and pull the four signals that matter.
Signal 1: the price is the probability
A prediction-market contract pays exactly $1 if its event happens and $0 if it does not. The price you pay for the contract is the market-implied probability of that event.
A 38¢ price means the market thinks there is roughly a 38% chance the event happens. A buyer pays 38¢ now to win $1 if “Yes” resolves; a seller is effectively betting $0.62 to win the same $1 if “No” resolves. The two prices always sum to $1. That mechanical identity is the entire foundation.
When a headline says “Polymarket gives Newsom a 26% chance,” it means the YES contract trades at 26¢. Nothing more, nothing less.
For more on the underlying mechanics, our Polymarket explained primer walks through how shares are priced, sold, and redeemed.
Signal 2: the resolution rule is the contract
Markets do not resolve on vibes. Every contract has a written resolution criterion and a designated source. “Will Trump be impeached?” might resolve on the House passing articles. “Will the Fed cut?” resolves on the FOMC statement language. “Will Kane win the Ballon d’Or?” resolves on the France Football announcement.
The resolution rule is the contract. If you do not read it, you do not know what you are pricing.
Common resolution traps:
- “Win the nomination” vs “win the convention.” Different rules. Different prices.
- “Recession” vs “two negative quarters.” NBER definition is broader and slower.
- “Ceasefire” vs “peace deal.” A ceasefire is bilateral and announced; a frozen conflict is not.
- “By March 31” vs “before April 1.” Edge dates trip up traders constantly.
If the rule is ambiguous, the platform’s dispute resolution process — UMA optimistic oracle on Polymarket, internal CFTC-compliant adjudication on Kalshi — kicks in. That can take days or weeks. Markets sometimes hold a non-trivial price gap during disputes.
Signal 3: order-book depth tells you fragility
A 26% price on a market with $50,000 of liquidity is fragile. A 26% price on a market with $5 million of liquidity is robust.
Two markets with identical headline prices can behave very differently when news hits. The fragile market jumps 8 points on a single $20,000 trade; the deep market absorbs the same trade in 2 points and snaps back. When you cite a market price, glance at the order-book depth and the recent volume. If both are thin, treat the headline as a soft signal, not a hard one.
For Polymarket, the volume figure shown on each market page is cumulative. The more useful number is the daily volume — how much traded in the last 24 hours. Markets that haven’t traded in days are essentially stale prices.
Signal 4: time to resolution shapes the price
A market that resolves in 18 days behaves differently from a market that resolves in 18 months. Long-dated markets carry compounding uncertainty: a single low-probability catalyst — a leadership change, a sanctions package, an injury, a policy reversal — has more time to occur. That is why our Russia-Ukraine ceasefire odds page shows 0.5% by April 30 but 26% by year-end on the same underlying conflict.
The shape of a price curve across resolution dates is itself a forecast. A steeply rising curve means traders expect change over time; a flat curve means traders expect the current state to persist.
Common misreadings to avoid
Once you know the four signals, the most common errors become easy to spot.
“The market is wrong.” Sometimes it is. More often, you are reading a different question than the contract resolves on. Re-read the rule before betting against the market.
“Volume = consensus.” Volume tells you depth, not consensus. A market can have $20M of volume and still be controlled by three large traders.
“The price is a forecast.” It is a clearing equilibrium. It is the best public guess of the probability, weighted by capital at risk. It is not a guarantee, not a prophecy, and not an oracle.
“Markets always beat polls.” Sometimes they do. For binary, well-defined, source-resolvable events with deep liquidity, prediction markets have historically matched or outperformed individual polls and roughly matched the best aggregators. For thin markets, ambiguous resolution rules, or distant resolution dates, treat the price as one input — not the answer.
For a deeper look at how this has played out empirically, see our prediction-market accuracy breakdown.
A worked example
Take Polymarket’s “Will the Fed cut at the April meeting?” — currently trading at 100% No. What that price actually tells you:
- Probability: essentially 0% Yes, 100% No.
- Resolution rule: FOMC statement language on the April 28–29 meeting.
- Depth: $186M cumulative volume, $32M live liquidity. Robust.
- Time: resolves in days. Almost zero room for catalyst risk.
Combined read: this market is essentially priced. The signal is so strong that even a $1M trade would barely move the price. If you saw a 5% Yes price on this market, you would not be reading a probability — you would be reading a structural mispricing or a misunderstanding of the rule.
Compare that to “First Fed rate cut at the December meeting?” at 61%. Same underlying central bank, very different market: distant resolution, multiple FOMC meetings between now and December, more room for the price to drift on data prints. A 2-point daily move on the December contract is normal noise; a 2-point move on the April contract would be a major event.
For the full breakdown, see our next Fed rate cut odds analysis.
When to use a prediction market in your reading
Prediction markets are most useful as a second opinion on news cycles. When pundits say “war is imminent,” the market price is the cheap counter-read. When polls show one candidate up, the market price tells you whether traders are willing to bet on the polling story. When economists disagree on a recession, the recession contract aggregates the disagreement into a single number.
They are less useful when the question is fuzzy, the resolution source is contested, or the market is illiquid. In those cases, treat the number as decoration, not data.
What to do next
If you are new to prediction markets, the natural follow-ups are: our Polymarket explained primer, our Kalshi vs. Polymarket comparison, and the is Polymarket legal page if you are a US resident.
Common questions
What does a 38% price on a prediction market mean?
It means a YES contract is trading at 38¢ on the dollar — and therefore the market is implying a 38% probability that the event will happen. The matching NO contract trades at 62¢. The two prices always sum to $1.
How are prediction markets resolved?
Each contract has a written resolution criterion and a designated source — typically a regulated outlet for elections, an official scoreboard for sports, or a designated index for economic markers. When the source declares the outcome, the contract resolves and YES or NO holders are paid out.
Is a high-volume market always a reliable price?
Volume helps, but it is not sufficient. A high-volume market can still be controlled by a handful of large traders, and a thin market with clear public information can still be reasonably priced. Always check the *recency* of volume — a market with no trades in 48 hours is a stale order book.
Are prediction markets more accurate than polls?
For binary, well-defined, source-resolvable events with deep liquidity, prediction markets have historically matched or outperformed individual polls and roughly matched the best aggregators. For ambiguous events, thin markets, or long resolution horizons, the comparison is less favorable.
What is the most common mistake new readers make?
Reading the price without reading the resolution rule. "Will X be the nominee?" and "Will X win the convention's first ballot?" are different questions and they price differently. The rule is the contract; the price is just the rule's shadow.
