SPOTR: Conviction Markets

Polymarket proved you can build liquid markets around real-world events. That is a genuine breakthrough. But the mechanism has a structural ceiling: the loudest wallet wins. A whale who stakes $500K on “Yes” moves the price more than a thousand informed bettors who collectively know more. Capital is the signal. Knowledge is noise.

This creates a market that rewards information access, not intelligence. If you know the answer before everyone else, you profit. If you merely have a better argument for why the answer should be what it is, you are irrelevant. The market does not care what you think. It cares what you can afford to bet.

Prediction markets price information. Conviction markets price the quality of your reasoning about that information.

SPOTR introduces a second variable into the payout formula: conviction score. Your share of the pool is not determined by stake alone. It is determined by stake multiplied by how well you argued your position, as scored by an AI persona with a distinct intellectual lens. Write a better take, earn a disproportionately larger share.

SPOTR is live on Solana mainnet. Five pilot markets are open. The mechanism works as follows.

Markets are persistent, async, and multi-participant. Anyone can enter any open market at any time. There is no matchmaking, no session, no opponent. You stake spotrUSD on a bucket or side, write a conviction take (or inherit one), and your position earns returns based on your conviction score when the market resolves.

The flow for any participant:

The AI does not resolve markets. It scores takes. Resolution comes from verifiable external facts entered by the platform administrator: a tweet count, a tournament result, a confirmed outcome. The AI is a quality signal, not an oracle.

SPOTR uses a conviction-weighted pari-mutuel system. Every participant on every side receives a share of the total pool. Winners receive more. Losers receive less. But within each side, the distribution is not equal. It is weighted by conviction score multiplied by stake.

Where:

This is the core insight. Losers still receive a payout (at 0.5x multiplier), but their share is compressed. Winners with high conviction scores earn disproportionately more than winners with low scores, even at the same stake size. A $5 stake with a conviction score of 9 earns more than a $10 stake with a conviction score of 3. Skill in writing directly translates to financial return.

Capital gets you in. Conviction determines how much you take out.

Each take is scored 0 to 10 by a single AI persona. The same persona scores ALL markets within a one-hour window, then rotates to the next. Seven personas, seven hours, then the cycle repeats.

The scoring is not generic. Each persona values fundamentally different things. The same take receives different scores from different personas. This is by design. Timing your take to align with the active persona is part of the strategy.

Difficulty scales with participation. The first three takes on a side receive lenient scoring (floor of 4 for genuine effort). After that, standards tighten. A crowded side requires a stronger take to stand out. This prevents early movers from being penalized for being first while maintaining competitive pressure as the market fills.

Not every participant needs to write. SPOTR introduces a take inheritance system. If someone on your side has already written a strong take, you can inherit it instead of writing your own. Your position activates with a decayed version of the original author’s conviction score.

Each take supports a maximum of 7 inheritors. The first inheritor receives the full core CCS. Each subsequent inheritor receives 3.5% less. Inheritor #7 receives 79% of the original score. The author always keeps their full score.

This creates a market for takes themselves. A high-scoring take becomes an asset. The author can set two parameters when writing: a minimum stake to inherit and a profit-take percentage. If an inheritor closes with positive PnL, the author receives that percentage of the profit. Authors are compensated for writing well. Inheritors are compensated for choosing wisely.

Write once. Get paid every time someone inherits your conviction.

Every position can be closed before market resolution. A pending position (no take written or inherited yet) receives a full refund. An active position exits at the 0.5x outcome multiplier, equivalent to being on the losing side at resolution.

This is the cost of leaving early. You always receive something back, proportional to your conviction score, but you are guaranteed less than if you hold and your side wins. Conviction has a price. Leaving is expensive.

One currency: spotrUSD. One on-chain touch point in. One out. Everything between is off-chain ledger arithmetic.

The Solana program is a dumb 1:1 USDC vault. Deposits transfer SPL tokens to the vault; withdrawals are admin airdrops. The program does not know about markets, positions, or conviction scores. All game logic is off-chain. This keeps the on-chain surface area minimal and the security posture tight.

Sponsor-pays-gas: users never need SOL. The platform sponsors all transaction fees. The user experience is USDC in, play, USDC out.

Points measure participation quality, not speculation volume. They are earned through action: staking positions, writing takes, and producing takes that others inherit. The incentive structure rewards the behavior that makes the market better: producing high-quality conviction arguments that others want to adopt.

Writing earns significantly more than entering. Getting inherited earns the most. Self-inheritance does not earn points. The leaderboard tracks accumulated engagement across all markets.

Season 1 launches with five fixed markets. Two multi-bucket prediction markets on Elon Musk tweet volume. Three binary World Cup 2026 markets.

All markets resolve from external facts entered by the administrator. Tweet counts from public trackers. Tournament results from official records. The AI never resolves an outcome. It only scores conviction.

A conviction market with persona-specific scoring, inheritance economics, and early exit pricing creates genuine strategic diversity.

SPOTR creates a new professional class: the takeswriter.

A takeswriter is someone who writes conviction takes on behalf of others. They study the seven personas. They know The Archivist rewards historical lineage. They know The Contrarian punishes safe takes. They adapt their voice to the active persona and set their own profit-share terms on their takes.

Their income model: write a take that scores high, set a profit-share percentage, let inheritors ride the score. Every inheritor who profits pays that percentage of their gains to the author. One great take, up to seven inheritors, compounding returns. The takeswriter is to SPOTR what a market maker is to a trading venue: essential infrastructure that makes the market function better.

The first profession where your income is determined by how well you persuade an AI.

Prediction markets proved you can price information. SPOTR proves you can price the quality of reasoning about information. That is a fundamentally different primitive.

The dataset SPOTR generates is unprecedented: financially incentivized human persuasion attempts, each scored 0 to 10 by a distinct AI persona with a published intellectual lens, with written rationales explaining the score. This is not crowdsourced annotation. It is conviction under financial pressure, judged by consistent characters with known biases.

Every market produces signal on two axes simultaneously. It produces a prediction (which side has more capital). And it produces a conviction map (which arguments scored highest, with which personas, at what stake levels). No other market produces both.

The strongest argument is the one the judge did not expect to agree with.

SPOTR trains humans to reason more precisely under pressure. The AI personas develop increasingly calibrated models of cultural argumentation. Both sides improve. That is the flywheel. That is the product.