*This page introduces the 'Why?' of forecasting to new visitors and budding predictors. If you're curious why someone would start forecasting in the first place, or why aggregated forecasts are hard to beat, you're in the right place.*

***

Have you ever found yourself in a disagreement with someone, where you each expected the future to play out completely differently? Maybe you were discussing whether a new technology would be the next big thing, or if a recession was looming, or the outcome of a presidential election.

These conversations often leave everyone frustrated, talking past one another, and they rarely reach a satisfying conclusion. Often they suffer from:

* **Vague language:** We might say something will "probably" occur or is "unlikely" to happen. But what does that mean, really?
* **No accountability:** No one is checking to see whether those bold claims—made in the heat of the argument—actually came true!
* **Unclear expectations:** What outcomes are we even talking about? Do we expect something to happen in the next two weeks—or the next two years?

## Enter Forecasting

Forecasting is the practice of making explicit predictions about the future. Instead of vague assertions, forecasters share:

* Specific probabilities — like the weather forecast that helps you decide if you should take an umbrella.
* Concrete dates
* Measurable outcomes

Instead of saying:

> “There’s a big risk of recession sometime soon.”

—a forecaster might specify that “there’s a 60% chance the US enters a recession in the next 12 months.” Much clearer! We might agree or disagree with them, but now we know how likely they think it is and when they think it’ll happen. If ultimately they’re right, we know it. If they’re wrong, well, at least they were willing to say what they actually believed with clarity and accountability.

**That’s what Metaculus is all about: questions that matter to people, clearly defined, giving you the ability to make forecasts and follow up on whether they were accurate.**

And when you start regularly using Metaculus, it becomes more than a forecasting platform — it becomes a whole new way of thinking, one that generates more productive disagreements and conversations that are grounded in what will actually happen, and in which pieces of evidence point toward which future.

## Wisdom of the Crowd

You might wonder, “If anyone can join Metaculus, why should I care what the community says about a topic?" 

The wisdom of the crowd is surprisingly—even startlingly—accurate. Metaculus keeps a transparent record of all resolved questions, which [you can explore here.](https://www.metaculus.com/questions/track-record/) That's not to say that the forecast is always right, but it is extremely difficult to beat consistently over time.

Here's an example: In the [Astral Codex Ten 2023 prediction contest](https://www.astralcodexten.com/p/who-predicted-2023), 3300+ participants competed to predict what would happen over the course of the year. Out of those thousands of forecasters+, only *three* individuals outperformed the Metaculus aggregate. That puts the aggregate in the 99.5th percentile! 

In addition to the empirical record on Metaculus's forecasts, there's also a fascinating research literature on the effectiveness of aggregated forecasting more generally. [*Superforecasting: The Art and Science of Prediction*](https://en.wikipedia.org/wiki/Superforecasting:_The_Art_and_Science_of_Prediction) by University of Pennsylvania psychologist Philip Tetlock is a great introduction to the field. It demonstrates how forecasting is a skill that can be developed with practice, feedback, and the right mindset.

**Metaculus is where you can begin honing that skill.** 

**Join Metaculus and help make the future clearer.**