# How Predictions work?

## Think it's going up?

In Predictions, players choose whether the price of an asset will be higher or lower than the mark price at the start of the epoch. If the player predicts correctly they win the payout based. If the prediction is incorrect, players loses the amount they added into the pool. Predictions offers a fast-paced and engaging way to speculate on short-term market movements.

## How Does It Work?

Predictions is  built on top of an Automated Market Maker (AMM) pool using Pyth's real time market data oracle. Players place their bets on whether the price of an asset will be higher or lower than a predefined mark price at a future time. The payout structure is dynamic and determined by the liquidity in the pool on either side of the bet. As more participants choose one side, the potential payout for the opposite side increases, creating a balanced risk-reward system driven by the distribution of funds in the AMM pool.

## About Results

Once the 5 minute epoch ends, the actual price of the asset is compared to the mark price. The results will be clearly displayed over the price graph indicating whether the final price was higher or lower. For each side, the payout ratio is calculated based on the total amount staked on both outcomes. Winning participants be able to claim their earnings in the history section by selecting claim or claim all.

## Not Working or Round not Executed, What Should I Do?

In the event that a round in Predictions is not executed or the system stops working, rest assured your funds are safe. Participants in the round must manually claim their refunded tokens in the history section.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.dragonswap.app/dragonswap/guides/predictions/how-predictions-work.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.