Plain-English glossary
Prediction Markets in Plain English
A compact guide to the terms that appear in prediction markets and in Prediction Market Radar, with the emphasis on what each term means in practice.
Prediction markets borrow vocabulary from trading, statistics, databases and event settlement. That can make a simple question sound more complicated than it is. This glossary keeps the definitions short and focuses on the distinctions that matter when comparing markets across platforms.
If you are new to the subject, start with Prediction Markets, Explained.
A
- All Offers
- The broad searchable catalogue of individual prediction-market offers gathered from supported sources. It is intentionally wider than the smaller set of cross-platform Super Matches.
- Arbitrage
- A situation where differences between genuinely equivalent markets may imply an exploitable price relationship. Prediction Market Radar is not an arbitrage execution service; equivalence has to be established before any price difference is meaningful.
- Ask
- The price at which someone is currently willing to sell a contract. It is one side of an order book.
B
- Bid
- The price at which someone is currently willing to buy a contract. The distance between bid and ask is the spread.
- Binary market
- A market with two principal outcomes, usually YES and NO. The apparent simplicity can hide complicated settlement rules.
C
- Canonical identity
- The normalized representation used to decide what real-world event, question and outcome a source contract refers to. This is one of the core data problems behind cross-platform comparison.
- Category
- A broad subject area such as Politics & Elections, Sports, Economics, Technology or Entertainment. Being in the same category does not mean two markets are equivalent.
- Compatibility
- The degree to which two source contracts refer to materially comparable outcomes. Prediction Market Radar checks event, outcome, timing and settlement conditions rather than relying on title similarity alone.
- Contract
- The tradable proposition attached to an outcome. Different platforms structure contracts differently, which is one reason normalization is necessary.
- Cross-platform comparison
- A comparison between markets on different venues after checking that the underlying questions are materially compatible.
D
- Data lineage
- The trace from a derived comparison back to the original source records. This matters when a comparison is questioned, audited or licensed.
- Deadline
- The date or time boundary built into a market. Two questions with the same subject but different deadlines may be different markets.
E
- Event
- The real-world occurrence behind a market, such as an election, match, policy decision, product launch or economic release.
- Event identity
- The answer to “what real-world thing is this actually about?” Names alone are not enough; competition stage, office, jurisdiction, date and other qualifiers can be part of identity.
I
- Implied probability
- A common way of translating a contract price into a rough probability interpretation. A YES price near $0.72 is often described as about 72%, but that is a market price, not a scientific certainty.
L
- Liquidity
- How easily a market can absorb buying and selling without large price movements. Two equivalent markets can have different prices partly because one has much more liquidity.
M
- Market maker
- A participant or system that provides buy and sell prices, helping other participants trade. Market-making structure can affect spreads and observed prices.
- Market question
- The proposition being resolved. The question should be read together with its rules rather than relying on the headline alone.
N
- Normalization
- The process of converting different source structures into a consistent form so they can be compared. This is necessary because platforms use different IDs, names, outcomes and data models.
O
- Outcome
- One possible resolution of a market. YES and NO are binary outcomes; elections and awards may have multiple named outcomes.
P
- Platform
- A source venue on which prediction markets are listed. Prediction Market Radar preserves platform identity rather than flattening all sources into one anonymous feed.
- Price
- The current trading value of a contract. It can be interpreted as a probability signal, but its meaning depends on the exact contract.
- Provenance
- The record of where data came from. For Prediction Market Radar this includes the source platform, original wording, source link, timestamps and other evidence needed to audit a comparison.
R
- Reference data
- The structured information that tells us what an object is, how it is identified and how it relates to other objects. In prediction markets this includes event identity, outcome identity, deadlines, settlement rules and source provenance.
- Resolution
- The final determination of a market’s outcome. It is closely related to settlement.
- Resolution source
- The authority, publication or data source used to decide the result. Two apparently similar markets can differ if they rely on different resolution sources.
S
- Settlement
- The process by which a contract is finally determined as YES, NO or another allowed result. Settlement rules are part of the market’s identity.
- Settlement compatibility
- Whether two source contracts would resolve consistently under materially the same real-world circumstances. This is one of the most important checks in a cross-platform match.
- Snapshot
- A recorded view of market data at a point in time. Snapshots make it possible to distinguish current information from historical observations.
- Source evidence
- The original details retained so a reader can inspect and challenge a comparison rather than simply trusting a derived percentage.
- Spread
- In market microstructure, the difference between the best available buying and selling prices. Prediction Market Radar also uses structured contract-spread information when presenting comparable sides.
- Super Match
- A Prediction Market Radar grouping of materially compatible market outcomes from different sources. It is not simply a keyword match.
T
- Threshold market
- A market asking whether a value will be above, below or reach a specified level. The exact threshold, measurement source and time boundary must match before two threshold markets are treated as equivalent.
- Timestamp
- A record of when data was observed or refreshed. Freshness matters because market prices can move quickly when new information arrives.
- Topic
- A navigational grouping of markets around a subject. Topic similarity is useful for discovery but is much weaker than contract-level identity.
V
- Venue disagreement
- A difference in prices between platforms for materially comparable outcomes. The difference may be interesting, but only after compatibility has been established.
W
- Withheld match
- A potential relationship that Prediction Market Radar deliberately does not publish because the evidence is not strong enough. Withholding is a quality-control outcome, not necessarily a system failure.
Y
- YES / NO
- The two sides of a binary market. YES represents the proposition occurring under the contract rules; NO represents it not occurring under those same rules.
A useful distinction
Similarity is not identity
A shared topic, person or keyword can help discovery, but it is not enough to prove that two contracts resolve the same way. Event, outcome, timing and settlement conditions still have to align.