PolytoolhubWhat is Metaculus?
Metaculus records forecasts directly rather than asking users to buy an outcome share. It supports binary, multiple-choice, numerical and date questions, making it useful for building a forecasting record and studying disagreement on questions beyond a trading venue’s catalog.
Who it is for
- Users practicing calibrated forecasting
- Researchers comparing human and AI forecasts
- Teams running structured forecasting questions
Features and coverage
- Available on
- Web, API
- Markets & venues
- Metaculus
- Research topics
- Probability forecasting, Forecast evaluation
- Binary probabilities
Enter a percentage through a slider or number field.
- Multiple choice
Allocate probabilities across mutually exclusive outcomes.
- Distribution forecasts
Express uncertainty over a numerical range or date.
- Mixture controls
Combine up to four weighted components for a distribution.
- Community comparison
Read the aggregate alongside individual reasoning.
- Scoring
Separate performance against a fixed baseline from performance against peers.
- Tournaments
Participate in question sets with their own rules and scoring.
- AI-bot programs
Explore the separate automated-forecaster tournament track.
Workflows with Metaculus
Form a probability
Read the resolution criteria, record an initial estimate and state what evidence would move it.
Update a distribution
Use a range that reflects uncertainty instead of only a point estimate.
Audit your record
Review many resolved questions and the chosen score definition, including when forecasts were standing.
Pricing and total cost
Public forecasting and research are accessible through the community platform. Private instances, commissioned tournaments and professional forecasting are separate service paths; their fees require a service-specific quote or terms.
Getting started
Before you start: Forecast scores are not returns on invested money.
- Choose a question and read the resolution criteria.
- Create an account to submit a forecast.
- Enter probabilities or a distribution.
- For multiple choice, ensure the total is 100%.
- Update when evidence changes.
- Review the resolved score and reasoning.
Useful links
What to check
- Forecast scores are not returns on invested money.
- Baseline and Peer scores answer different questions and should not be mixed.
- Time-averaged scoring considers how long each forecast stood; a last-minute update does not erase earlier calls.
- Tournament prizes depend on the tournament’s own rules.
A closer look
Ratings & reviews
User reviews 0
The main score switches to user ratings after 5 published reviews.
A specific example helps the next person choose the right tool.
Write a review
How was your experience with Metaculus?
Share what you used, what worked and what could be better. We confirm your email and review the submission before publishing.
To submit a review, email info@polytradingbot.com with the tool name, your rating and your experience. Enable JavaScript to prepare a draft here.
Related tools
Browse category
Create and trade play-money prediction markets, discuss the reasoning and build a public forecasting record with virtual mana.
- Access Web · Mobile app
Compare AI agents’ forecasts, reasoning and revisions against a resolved record, with APIs and divergence alerts on paid plans.
- Access Web · API
Turn a market link, question or screenshot into an AI forecast and compare it with market odds and news context.
- Access Web
Frequently asked questions about Metaculus
Do I buy shares to forecast?
The described workflow records a probability directly.
Can I forecast a number or date?
Yes, using a distribution.
What does Peer score compare?
Your performance with other forecasts on the same question.
Does only my final forecast count?
The standard time-averaged method considers earlier standing forecasts too.