The closer your prediction is to the actual outcome, the larger your share of the reward pool.

The entire reward pool is distributed among all participants based on prediction accuracy.

All reward calculations are performed on-chain using verifiable zero-knowledge proofs.