2,222
Seats in a season
Every phase sells against the same audited contracts, the same settlement engine and — where possible — the same users. Only phase one is built. The rest are ordered by how much of phase one they reuse.
A weekly contest on tokenized shares. Participants take a seat, run a notional $100,000 book on live prices, and set their weights once a week. Books lock, the market runs, and the worst performers are eliminated every Friday until a final field remains. The winner takes the majority of a pot funded entirely by optional participation fees.
Seats in a season
Weekly cycles
Instruments to choose from
Held in custody, ever
Three contracts deployed. Settlement engine scoring real weeks from a recorded price tape. Full participant experience — wallet, weight editor, live board, claim. 234 tests green, two adversarial audits, 76 findings closed.
No mint has run. No revenue has been earned by this product. External audit and a jurisdiction-specific legal opinion are the two items gating launch.
It is the hardest version of the mechanic. Everything after it reuses the settlement engine, the payment verification and the price infrastructure without re-earning them.
The same seats and the same ladder, opened to instrument classes with more expression in them. Sold as a recurring access pass each season rather than as a new audience to acquire.
Settlement engine, price tape, seats, contracts, payment rails. Effectively all of phase one.
Instrument-specific scoring and a second pot running alongside the first.
Phase one producing real cohort data on how often people pay to keep playing.
Prediction markets traded $111B of notional in a single quarter of 2026, up more than 1,700% year on year, and the infrastructure for participating in them programmatically is thin. We already run the price and settlement machinery that the problem needs.
The same engine that scores a weekly contest can price and settle an event contract. The feed changes; the machinery does not.
Researched, not built. No revenue is attributed to this phase anywhere in our materials.
Whether to participate in existing venues or to run our own contests on top of them.
$32.2B of real-world assets are already on-chain, close to triple a year earlier — and almost all of the attention has gone to bringing assets on rather than to distributing them once they arrive. Phases one to three make us a distribution channel before we ever become an issuer.
Build the audience on someone else's assets first. Issue only once there is somewhere for the asset to go.
The user base and the compliance groundwork built across the first three phases.
Direction only. Deliberately unscoped until the earlier phases produce data.
The least liquid things people own meeting the most liquid settlement rail ever built. Fractional exposure, continuous pricing, and a secondary market that does not close — applied to categories where the current holding period is measured in years and the current exit is a phone call.
It carries the most regulatory weight and the least reuse. It only makes sense once the earlier phases have paid for the compliance work.
A jurisdiction where the issuance is clean, and a distribution base large enough to clear the first offering.
A thesis, not a plan. Listed because it is where the chassis leads, not because it is scheduled.