BOUT / Docs

Readings and the decision

Trace the public readings behind each round. Check the source, the timing, the middle reading, and the no contest rule.

Prelaunch. Live readings. Simulated tickets.

The public sources

The board reads two public endpoints. It uses the leaderboard, then requests each listed account's portfolio.

The portfolio request uses POST with type set to portfolio and user set to the public account address.

These are source endpoints, not partners. BOUT is independent.

The read cadence is 5 minutes. The board caches successful responses on that cadence.

The published decision rule takes the median of three eligible readings at each end of a round.

Sort the three spread values. Use the middle value at open, then the middle value at close.

A reading more than 20 minutes from the decision moment is ineligible. Missing or duplicate timestamps cannot supply extra readings.

A wins when the closing spread exceeds the opening spread. A flat result goes to B.

Either account leaving the board makes a no contest. So does having fewer than three eligible readings at either end.

Stale readings cannot decide the round. If they leave too few eligible readings, every stake returns at cost.

Each successful board read carries a SHA-256 hash. It covers readAt, accounts, and matchups in that order.

Open /api/board. Save the response, including its timestamp and hash.
Build an object containing readAt, accounts, and matchups, in that order. Keep the values and array order unchanged.
Serialize that object with JSON.stringify. Use Node's crypto.createHash("sha256").update(text).digest("hex") on that string. Compare the result with hash.

Matching hashes confirm the saved payload is consistent. They do not prove the source figures are correct.

To inspect the source, compare the leaderboard and portfolio responses. Later responses may differ as public data changes.