Sports
Within the rules.
Possiblefutures.
The future has a price.We take a position.
Between what is known and what may come, a market takes shape. Time Seventeen Capital researches and trades risk-event contracts in prediction markets, with our own capital.
The world moves.
Its price follows.
Some forces enter the world long before they find a language in markets. They alter routes, redraw dependencies and rearrange the possible.
We are interested in that interval. The distance between an event being discussed and a claim being priced is a place for research, disagreement and capital.
Prediction markets give different views a shared object: a contract with a defined outcome. Our work begins with the quality of that definition, and continues through everything it takes to trade it.
Follow the questionA contest. A choice.
A world in motion.
We see prediction markets through three broad territories. Our capital is focused on risk: the events that change the conditions under which economic life takes place.
Elections
Through collective choice.
Risk
When the rules themselves move.
A headline contains
many futures.
A conflict is not one future. A disruption is not one scale. Time changes the question.
Limited escalation, sustained disruption and a wider conflict describe different states. A brief interruption and a prolonged loss of critical infrastructure create different consequences. We look for the distinctions that a single narrative compresses.
Which possible world?
The boundary between one outcome and another is where a contract begins.
From a question
to capital at risk.
Research, engineering and trading belong to one continuous process. Follow a question through the decisions that give it consequence.
Begin with the world.
The first task is to decide what deserves attention. A shipping route, a cloud dependency or a policy decision can connect events that appear unrelated. We start with the mechanism: what could change, who would be affected, and which observations would let us distinguish one outcome from another.
What is changing, and what would let us see it?
Give uncertainty a boundary.
A view becomes researchable when its terms are precise. We read the event definition, observation window, evidence hierarchy and resolution process together. An interruption that happens once and an interruption that persists are different claims. So are a threshold reached before a deadline and a condition observed at that deadline.
What exactly must be true for this contract to pay?
Let the view meet the price.
An event can matter greatly and still offer an unattractive trade. We examine the price beside the evidence, alternative scenarios and the costs of being wrong. Related contracts can sharpen the question, but comparisons only become useful after their definitions and observation windows have been reconciled.
Which assumptions make this price worth taking?
Meet the market as it is.
The order book is part of the research. Spread, depth, fees and slippage determine how much of an idea can survive execution. We consider the size that can actually trade, the information revealed by an order, and how the economics change if liquidity disappears before the position can be reduced.
Can this view become a position on acceptable terms?
Stay responsible for the position.
Ownership continues after the fill. Event evidence changes, related exposures accumulate and settlement can take longer than expected. We consider concentration, available capital, platform dependence and the path to resolution together. The ability to hold or reduce a position belongs inside the original decision.
What must remain true for us to keep holding?
Return the result to the question.
A profitable outcome is evidence to examine. We separate the event judgment from the entry price, execution costs and settlement outcome, then ask which parts of the reasoning should survive. A useful review changes the next research question, the next system check or the next decision about risk.
What should we repeat, revise or stop doing?
To hold a view
is to accept
its limits.
A model is a way of asking. A price is an invitation to disagree. A position gives that disagreement consequence.
We look for what would change our minds, and size our convictions for what we cannot know. That discipline connects the research question to the capital behind it.
TradingThe distinctions
we return to.
The work becomes clearer when we keep a few difficult distinctions in view. They shape how we read a contract, evaluate an idea and judge a result.
A forecast and a trade.
Understanding an event and understanding a position are related disciplines. A view of the future must still meet an entry price, an execution cost and a settlement process.
A correct forecast can produce a poor trade if the price already reflects it, the position is too large for the available liquidity, or the contract resolves differently from the event we had in mind. We examine each link in that chain.
A price and a probability.
A contract price is an observation of a market. Its relationship to probability depends on the payoff, timing, costs, liquidity and the constraints faced by participants.
We study the incentives around the price as carefully as the event behind it. The point is to understand what a position means under the terms on which it can actually be held and resolved.
A quiet market and a resilient one.
Calm can coexist with accumulating fragility. Leverage, concentration and reliance on the same exit can make a stable surface a misleading guide to what happens under pressure.
We separate the event that starts a move from the mechanisms that amplify it. That distinction keeps the research attentive to the path of a position, as well as its final outcome.
An outcome and an explanation.
A result leaves questions behind. Was the view well founded? Was the price attractive? Did execution preserve the idea? Which assumptions mattered, and which were merely fortunate?
We want explanations that expose their own limits. A useful account tells us what evidence would challenge it and what, specifically, should change in the next decision.
Observe carefully.
Disagree precisely.
Build continuously.
Research becomes more useful when the distance to implementation is short.
Data, code and judgment meet in the same workflow. We use quantitative tools to make assumptions explicit, automation to reduce repetitive work, and review to examine what the systems cannot settle on their own.
AI belongs in that process as a tool for building and investigation. The responsibility for the evidence, the decision and the result stays with the people using it.
ResearchResearch,
in the open.
A shared record makes room for independent questions.
Our organized and cleaned Polymarket, Kalshi and Polymarket US datasets are publicly available on Hugging Face.
Explore datasetsA different view
is a place to begin.
We value people who question precisely, build thoughtfully and let the work answer.
Independent thinking matters most when it can be shared, challenged and turned into finished work. We look for the curiosity to go deeper and the ownership to carry an idea through.
Find your place