Resource Guide
A comprehensive guide to accessing, understanding, and using prediction market data — for researchers, developers, analysts, and anyone who wants to understand what the collective market knows about future events.
A prediction market is a platform where participants buy and sell contracts on the outcome of future events. The price of a contract reflects the collective probability estimate of that outcome occurring. If a contract on "Team A wins the championship" trades at $0.62, it means the market collectively assigns a 62% probability to that outcome.
Prediction market data captures these prices — and changes in these prices — over time. Unlike polling data (which captures stated opinions) or statistical models (which compute probabilities from historical patterns), prediction market prices reflect the revealed beliefs of participants who have real financial stakes in accuracy. This incentive structure is what makes the data useful for research and forecasting applications.
The data universe includes:
Prediction markets operate across multiple domains: sports events, political outcomes, economic indicators, and other quantifiable future events. The data from these markets has attracted significant academic interest for its properties as a leading indicator of collective knowledge.
Most prediction markets use binary contracts: a contract pays $1 if the event occurs and $0 if it does not. The market price is therefore a direct probability estimate. A contract priced at $0.74 implies a 74% probability of the event resolving YES.
Some markets offer multiple outcomes (e.g., "which team wins the World Series"). Each outcome has its own contract, and the sum of all contract prices should equal approximately $1 (allowing for bid-ask spreads and platform fees).
A growing category: normalized probability estimates computed by aggregating prices across multiple regulated platforms. Aggregated data accounts for liquidity differences between platforms, reduces the noise from any single market's idiosyncratic pricing, and provides a more stable probability estimate. The spread between platforms is itself a signal — high spread indicates market uncertainty or new information being priced in.
Post-resolution records that pair each contract's pre-resolution probability with the actual outcome. Essential for calibration analysis: you can measure whether markets priced at X% resolved YES approximately X% of the time.
Contract market volume and open interest. Higher volume generally correlates with more accurate pricing. Thin markets (low volume) can be moved by individual orders and are more susceptible to noise.
| Data Type | Format | Best For | Availability |
|---|---|---|---|
| Spot prices | JSON / REST API | Real-time dashboards, widgets | Often free |
| Price time-series | JSON / CSV | Research, modeling, backtesting | Varies by platform |
| Aggregated consensus | JSON / REST API | Normalized multi-source signal | Free via Meridian Edge API |
| Settlement outcomes | JSON / REST API | Calibration, accuracy analysis | Free with API key |
| Volume / order book | JSON / WebSocket | Market microstructure research | Platform-specific |
| Historical datasets | CSV / parquet | Academic research, ML training | Starter access |
Multiple sources provide prediction market data at no cost, with varying coverage, depth, and formats. The right source depends on your use case.
Aggregated consensus probabilities across multiple regulated prediction markets. Covers NBA, NFL, MLB, NHL, MLS, US politics, and economics. Updated every 10 minutes. Includes spread, confidence, movement direction, and settlement history.
Many regulated prediction market platforms expose REST APIs with current contract prices. Coverage is limited to that platform's own markets. Rate limits vary. Check each platform's developer documentation for current terms.
Several academic repositories host historical prediction market datasets for research purposes. The Harvard Dataverse and similar platforms host curated datasets from past election markets and economic event markets. Iowa Electronic Markets (IEM) at University of Iowa has maintained public data since 1988.
While not prediction market data, sites like Baseball Reference, Basketball Reference, and Pro Football Reference provide the statistical context needed to interpret prediction market probabilities. Useful as ground truth for calibration analysis.
A note on data licensing: Prediction market price data reflects trades made on regulated platforms. Review each platform's terms of service before using their data in commercial products, publications, or redistribution. Aggregated consensus data from Meridian Edge is licensed per our Terms of Service.
Prediction market data from different sources uses inconsistent schemas. Understanding the common fields and how to normalize them is essential for working with multi-source data.
SPORT:TEAM1-TEAM2:DATE. Allows deduplication across sources that use different naming conventions.Working with raw multi-platform prediction market data requires handling several normalization challenges:
Using an aggregated API like Meridian Edge handles these normalization steps automatically, returning a clean canonical schema regardless of how the underlying platforms represent the data.
Prediction market data is used extensively in academic work on forecasting, information aggregation, and collective intelligence. Common research designs include: calibration studies (comparing consensus probabilities to outcome frequencies), information arrival studies (measuring how quickly prices update after news events), and comparative forecasting studies (market consensus vs. polls vs. statistical models).
See our research access program for free Pro-tier API access for qualified researchers.
Win probability data enriches pre-game analysis, live coverage, and post-game context. A consensus probability from multiple regulated markets provides a more stable estimate than checking a single source. The free embeddable widget makes it trivial to add live probability data to any sports content.
Developers build prediction market data into dashboards, alert systems, data pipelines, and prediction tools. The REST API supports any programming language with HTTP support. Python and JavaScript examples are in the API documentation.
Political events are among the most actively traded prediction market categories. Consensus data shows the collective market view on electoral outcomes, policy decisions, and geopolitical events. This data is used by journalists, political scientists, and policy researchers.
Game outcome probabilities from prediction markets serve as useful reference points for fantasy sports decisions and DFS lineup construction. They reflect real financial stakes rather than model outputs, and update continuously as game conditions change.
The Meridian Edge API is the fastest way to access aggregated prediction market consensus data programmatically. The starter tier (100 calls/day) requires instant activation.
Get a permanent API key at meridianedge.io — instant signup, instant activation.
Full endpoint reference, parameters, and response schemas: meridianedge.io/docs.html
Meridian Edge provides free Pro-tier API access (5,000 calls/day, full historical data, settlement outcomes) to academic researchers studying prediction markets, forecasting, and information aggregation.
Qualifying research areas include: forecasting calibration, market efficiency, information aggregation, collective intelligence, election forecasting, sports analytics, behavioral economics, and adjacent fields.
To apply: email [email protected] from your institutional address with your name, institution, and a brief description of your research. Full details at meridianedge.io/research.html.
Prediction market data consists of contract prices on regulated platforms where participants participate in predictions about the probability of future events. Prices function as probability estimates: a contract at $0.65 implies a 65% probability. Data includes price time-series, volume, spread, and settlement outcomes.
Yes. Many regulated platforms provide public APIs showing current prices. Aggregated consensus data is available via the Meridian Edge API (100 calls/day, instant activation). Historical datasets and higher-volume access typically require a paid plan.
Primarily JSON via REST APIs with fields for event identifier, probability (as decimal), volume, timestamp, and settlement outcome. Meridian Edge normalizes data across platforms into a unified schema with consensus probability, spread, confidence, and sparkline history.
Use it as a probability time-series for calibration analysis, event study designs, or forecasting benchmarks. Meridian Edge offers free Pro-tier API access for academic researchers — see meridianedge.io/research.html.
Prediction market prices reflect financial stakes — participants risk real money, incentivizing accuracy. Polling captures stated opinions without stakes. Academic research finds prediction markets are often better calibrated for high-salience near-term events with active participation.
Individual platform prices update continuously as trades execute. Meridian Edge consensus data updates every 10 minutes during active market hours, 30 minutes during low-activity windows. Full time-series stored for all settled events.
Live dashboard requires no signup. API starter tier: 100 calls/day, instant activation.
Open Dashboard API DocsFor informational purposes only. Not investment advice. Data aggregated from publicly available prediction market sources. © 2026 VeraTenet LLC d/b/a Meridian Edge. See Terms, Privacy, Risk Disclosure.