# Payout system Torch’s payout model rewards high-quality predictions without relying on fixed market deadlines or pooled prize funds. The winning bets are rewarded based on a dynamically computed multiplier that reflects the quality and timing of the prediction. Each token maintains its own isolated liquidity reserve, funded by forfeited losing bets. This reserve is the sole source of payouts, avoiding payout debt or delays. The size of a winning base payout is set in advance at the moment of betting and is released at the moment of resolution unless it exceeds the reserve. It is calculated based on two core factors: 1. **Prediction quality**\ Bets are evaluated for their 1. Sharpness (how narrow the predicted range is) 2. Boldness (how far the prediction deviates from the current probability distribution at the time of placement) 3. Lead time (how far in advance the prediction was made) Together, these dimensions reflect how risky, useful, and information-rich the prediction was. 2. **Reserve liquidity**\ All payouts are capped based on the current size of the token's liquidity reserve to ensure solvency Additionally, the payout might be boosted with a bonus, should reserve accumulate sufficient funds from the recent betting activity: 3. **Reserve bonus**\ A portion of the surplus reserve funds that exceed its Target Level. To prevent spam and abuse, especially from trivial predictions like betting on extremely wide ranges, all bets are subject to a protocol fee. This guarantees that even bets with a theoretically positive expected return will be unprofitable if they offer no meaningful insight. The system naturally filters for high-confidence, information-rich predictions. Torch’s payout architecture strikes a balance between personal incentive (to earn by predicting accurately) and systemic value (to contribute useful signals). It rewards those who identify significant movements early with precision and conviction, especially when doing so against the grain. --- # 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://torch-1.gitbook.io/litepaper/payout-system.md?ask= ``` 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.