Contest Information

Draft — contest details are not final and will change.

This document explains the Proof paper-trading contest: what it is, who can take part, how your account works, how the leaderboard is scored, and the fair-play rules that keep the result a contest of skill. It is conceptual and rule-focused. It does not cover how the underlying instruments work mechanically — for that, see the market-structure and trading documentation.

Two things to read alongside this page:

• The official contest page holds the figures and dates that change between runs: the exact window, the starting balance, the prize specifics, and the binding official rules. Where this document says "check the contest page," that page is the authority.
• The official rules posted on that page are the binding terms. This document is an explanation; if anything here and the official rules ever differ, the official rules govern.

What the contest is

The contest is a paper-trading competition run on Proof ahead of mainnet. You trade with a fixed amount of virtual funds — no real money is at risk and no real money is deposited — and a public leaderboard ranks participants by how well they trade. It is a hands-on way to learn Proof's instruments, including conditional markets, against live prices and a live order book, while competing for a small set of real prizes and a founding-trader credential.

The point of paper trading is that the full exchange is real — the matching engine, the margin system, the oracles, the settlement logic — but the balances are simulated. You get the real mechanics with none of the financial risk, which makes it the right environment to learn an instrument you have not traded before.

The window

The contest runs for a fixed, short window with a published start and end. The window is deliberately timed so that every event-driven market in the contest set resolves before the contest ends, which matters for how the final standings are scored (see How scoring works). For the exact start and end dates of the current run, check the official contest page.

Eligibility and access

Entry is free, but access is gated — there is no open sign-up.

• Single-use access codes. You join with an access code that works once and then expires. A code admits one participant.
• One account per person. Each participant trades a single account. Multiple accounts controlled by one person are not allowed and are grounds for disqualification (see Fair play and gates).
• Waitlist. If you do not have a code, you can join the waitlist. Codes are issued from the waitlist and through direct invitations.
• Eligibility attestation. Joining requires confirming you meet the basic eligibility terms in the official rules (for example, an age minimum and not being in an excluded jurisdiction). The contest is void where prohibited.
• Identity checks at payout, not at entry. You do not need to verify your identity to join or to trade. Identity and eligibility verification happens only at the prize-payout step, for participants who finish in a prize position. This keeps getting started frictionless while still gating the money. Expect, at minimum, a sanctions screen and the tax paperwork required for cash prizes in your jurisdiction; the specifics are in the official rules.

Your account: a fixed paper balance

Every participant starts with the same fixed balance of virtual funds. That is the entire bankroll for the contest.

• It is the same for everyone. Identical starting balances are what make the leaderboard directly comparable across participants — your result is your percentage return on a shared starting stake, not a function of how much you brought.
• No refills, no resets. The balance is not topped up, and a blown-up account is not reset. You manage one bankroll across the whole window.
• No real money. The balance is virtual. Nothing is deposited, nothing is withdrawable, and the simulated positions are not real financial instruments. Simulated results do not predict real-world performance.

For the exact starting figure in the current run, check the official contest page — it is set per run and is not fixed across contests.

Markets in the contest

The contest runs on a small, focused market set rather than the full exchange. A constrained set is intentional: it keeps the competition legible and pushes participants to actually learn the instruments instead of spreading thin.

The set is built around a single underlying asset and includes:

• A standard perpetual future on the asset — the deepest book and the natural hub for directional trading and for hedging conditional exposure.
• One or more conditional markets on that same asset — each one lets you trade the asset conditioned on whether a real-world event resolves YES or NO, alongside a prediction binary on the event's probability (a $0–$1 instrument whose price is the market-implied probability of the outcome). At least one of these conditions on a recurring event that resolves repeatedly through the window, and at least one conditions on a scheduled one-time event that resolves once.

Contest events always resolve to a definite YES or NO. Every event behind a contest conditional market is chosen and configured so that it resolves cleanly to one side — a contest event will not Void. The general void path that can apply to conditional markets elsewhere on the exchange (for an event that becomes undeterminable, is cancelled, or never happens) does not apply to contest markets: in the contest you can count on every conditional market resolving YES or NO. The void mechanics described in the settlement documentation are therefore out of scope for the contest set. For how voiding works in general, see Settlement, Resolution & Void.

For how these instruments actually work — the underlying perpetual, the conditional perpetuals, the prediction binaries, how they price, margin, and settle — see the market-structure, pricing, and settlement documentation. This page only describes their role in the contest. The specific markets in the current run are listed on the official contest page.

How scoring works

The ranking metric is profit and loss (PnL). Because every participant starts from the same fixed balance, PnL is equivalent to percentage return, which is the most legible possible metric and keeps the leaderboard provably comparable.

There are two distinct numbers, and the difference matters:

Live standings: provisional marked equity

During the contest, the live leaderboard ranks you by your marked equity — your virtual cash plus the current marked value of your open positions, valued against the engine's mark prices. This number moves continuously as prices move and as you trade.

The live number is provisional. It is explicitly labeled as not final, because it includes the marked value of positions that have not yet settled. A conditional position in particular carries value that is contingent on an event that has not happened yet, and that contingent value is not cash until the event resolves.

Final standings: realized cash after settlement

The standings that decide prizes are based on realized cash, computed after every event-driven market has resolved and the value of any remaining open positions has been accounted for.

This is why the window is timed the way it is. Two things have to be true before the final number is clean:

1. All event-driven markets resolve, and the final cash number is taken afterward. Closing a conditional position early does not pay cash directly. It converts the position into prediction-binary tokens, which you then sell on the binary market at whatever the binary book is paying — around the event's implied probability times face. Holding to resolution instead settles the position to cash at the realized outcome. The clean, final cash figure is taken after every event has resolved, so the standings rest on settled cash. (The cash-out mechanism is covered in the settlement documentation.)
2. Any remaining open positions are valued at a clean reference price. Positions still open at the end — such as standing perpetual positions — are valued against the oracle price rather than a contingent mark, so the final figure rests on a clean reference price. The exact end-of-contest handling — how open positions are settled for final scoring — is being finalized, and the precise method will be confirmed on the official contest page before the contest begins.

Once events have settled and remaining positions are valued at a clean reference price, the final number is realized cash, and no contingent mark is left deciding the money.

Tie-break. If two final results are equal, the earlier achievement of that result wins — a deterministic, skill-based tie-break. Ties are never broken at random.

Fair play and gates

The leaderboard is "just PnL," and a set of structural gates is what keeps that fair. The gates fence off degenerate strategies without distorting the metric itself.

• A low leverage cap. Maximum leverage is held to a low, conservative level for the contest. This is the single most important gate: capping leverage shrinks the variance tail that a pure-lottery strategy would chase, so the leaderboard rewards repeatable skill rather than one lucky maximally-levered bet. It is also why no separate drawdown or concentration penalty is needed in the scoring formula.
• An activity floor. To qualify for the final standings and prizes, you must be genuinely active — more than a single trade. The principle is published; the exact threshold is blind during the contest, so it cannot be gamed down to the bare minimum, and it is disclosed in the post-contest recap. The intent is simply to require real participation, so the bar is modest.
• One account per person; no refills. One account per participant, and no balance top-ups or resets — both are necessary for cross-participant comparability.
• Oracle-anchored, liquid instruments only. The contest set is restricted to a liquid perpetual that marks against an external oracle, plus conditional markets anchored to that same oracle-marked underlying at settlement and held in line by arbitrage. There is no thin, self-referential instrument whose mark a participant could move to manufacture PnL.
• Programmatic trading is allowed; gaming is not. Automated, API-driven, or algorithmic trading is permitted — Proof is built for algorithmic traders. The line is between automation and gaming, not between manual and bot. Permitted: trading programmatically. Prohibited: wash trading, self-dealing, collusion, multi-account or Sybil entries, opposing-position straddles across accounts, and any manipulation of marks or the leaderboard. Rate limits exist to protect the infrastructure, not to ban automation.
• Team discretion to disqualify. The contest team reserves the right, at its discretion, to disqualify any participant for gaming, manipulation, or abuse of the system, and runs a post-contest audit before prizes are paid. The specifics live in the official rules.

Prizes and recognition

The headline of the contest is recognition and access, with a small cash component on top.

• A small cash ladder. A modest prize pool is split across the top finishers, weighted toward the top of the leaderboard. The exact pool and the per-rank breakdown are on the official contest page.
• A founding-trader credential. Genuinely-active participants earn a verifiable Founding Trader credential — the identity marker for the cohort that traded Proof before mainnet. There may be associated recognition items for top finishers (skill-selected by rank, never by random draw).
• Future-program eligibility. Genuinely-active participants (gated on meeting the activity floor, not on leaderboard rank) get priority eligibility for future Proof programs.
• A feedback track. Because product feedback is one reason the contest exists, there is a small recognition track for the best structured feedback.

For the exact prize amounts, counts, and any merchandise specifics, check the official contest page. Prizes are skill-ranked; there are no random-draw or percentage-of-winnings payouts.

Schedule and settlement

The contest runs on a fixed schedule with three practical phases:

1. Live trading across the published window.
2. Settlement at the close. Every event-driven market resolves within the window, and any positions still open at the end are valued against a clean oracle reference for the final figure, so standings convert from provisional marks to realized cash. The exact end-of-contest handling for open positions is being finalized and will be confirmed on the official contest page before the contest begins.
3. Audit and payout. After the close there is an audit window to verify standings and screen for gaming before any prize is paid. Identity and eligibility verification for prize-eligible finishers happens here. Prizes are paid after the audit clears.

The exact dates for each phase are on the official contest page.

Changes and maintenance

The contest runs on a pre-release exchange. Two consequences follow:

• Mechanics may change. Engine behavior, market parameters, and contest details can be adjusted before and during the run as the product is finalized. Material changes are communicated through the official channels, and the official contest page always reflects the current rules.
• Scheduled maintenance. As with any pre-mainnet run, there may be scheduled maintenance windows. The team's intent is a fair, comparable result for every participant, and it reserves the right to modify, pause, or cancel the contest as needed — as set out in the official rules.

Where the official rules live

This document explains the contest; it is not the binding terms. The official rules, the exact dates, the starting balance, the specific markets, and the full prize details all live on the official contest page. That page governs. Read it before you enter.