01 / 04Research
Publications
Working papers written to be argued with, not admired. Nothing carries the Institute’s name until it survives review — and dissents publish alongside the paper.
ExploreIndependent research institute
What if tomorrow
was predictable?
Prediction markets are becoming public infrastructure — pricing elections, policy, technology and risk. The Institute makes that infrastructure legible: methods published before results, tools you can use today, and a public record of what we have and have not done yet.
The Institute is a research and professional body. It is not an exchange, a broker-dealer or a regulator, and it does not provide investment advice.
Live ensemble · binary settlement Running
- Resolved
- 0
- Settled yes
- 0
- Settled no
- 0
Each path is a martingale: p(t) = Φ(X / √(T−t)). No drift, no edge — so the yes share converges on — as the sample grows. That is what a price with no information advantage looks like, and it is the baseline every claim about forecast accuracy has to beat.
Illustrative contract wordings with simulated prices. Not live market data, not quotes, and not a forecast of any outcome.
Free tool · runs in your browser
Test our judgement before you trust it.
Paste a market question. It is checked against published wording-failure patterns from the standing research agenda — the ones that reliably produce disputed resolutions. Every rule is named and fixed in advance. Not AI, no scoring model, no network call: nothing you type here leaves this page.
How this works
- Everything above stays in your browser. Nothing is sent to a server or stored anywhere.
- Each check is a named pattern, listed alongside the result — not a model's judgment.
- A clean result does not mean the wording is safe. It means no listed pattern matched.
- This is not legal advice, and it is not a determination the Institute stands behind for any specific contract.
What gets checked
- · Subjective language
- · Compound condition
- · No resolution source named
- · No resolution deadline stated
- · No dispute or tie-break procedure
- · No halt or suspension criteria
- · Numeric threshold without a rounding rule
Start typing, or load an example, to see a result here.
A price is the most honest forecast we have
Four bodies of work,
one standard of care.
One organisation, two names: PMI is the Institute; PMA-CICP is the credential it is building — not a second body.
02 / 04Data
Methods & measurement
Scoring rules, calibration measurement and sample thresholds — all fixed before a single result is published, so a flattering early sample can never become a headline.
Explore03 / 04People
Fellows
A working network, not a masthead. Research, practice and policy fellows carry the agenda; reviewers may publish a dissent against work they were not persuaded by.
Explore04 / 04Standing
Membership & credential
A professional home for people who advise, design, operate and govern these markets — plus PMA-CICP, an assessed credential in development.
ExplorePublic record
What exists, and what doesn't.
Founding institutions usually describe the finished version. This is the actual state, dated. If a row moves you will be able to see when — and if it doesn't move, you can hold us to that too.
2 live · 7 in progress · 3 not started — 12 of 12 rows computed from the site's own data, 1 resting on an attestation
| Commitment | Status | Last changed |
|---|---|---|
| The Wording Clinic autoFree, rule-based and client-side. Seven published patterns, fixed in advance and citable by identifier, each tied to a standing research track. Embeddable on other sites. | Live | Recorded 17 August 2026, from CLINIC_PATTERNS. |
| Standing research agenda autoSix open tracks, published with the question each one is trying to answer. | Live | Recorded 16 July 2026, from RESEARCH_AGENDA. |
| Adversarial review memos auto1 memo drafted, 0 published. Current questions argued from named opposing positions. Shown while still draft — pending review, not yet an Institute position. | In review | Recorded 17 August 2026, from pmi_adversarial_memos. |
| Published working papers autoNone yet. A research index padded with work that does not exist is worse than an empty one. | None | — |
| Flagship report auto“The resolvability of event contracts” — a working title. Clears internal review by fellows and outside readers, with any dissent published alongside it. | In preparation | — |
Rows marked auto are computed from the data that already governs this site — the pattern registry, the research agenda, the publications list, the fellows and council rosters, exam scheduling, the entity note, and the memo and credential tables — so they change on the day the fact does, not on the day someone remembers to edit this page. A row marked attested rests on a person's claim about the present that nothing here can check; after 90 days without re-confirmation its date is replaced by unverified since, rather than going on looking like a date we could prove.
Who it's for
Three doors, three different asks.
The Institute is not selling the same thing to everyone. Start where your problem actually is.
01 — Venues & operators
Wording that survives a contested outcome.
Your resolution risk is written into the contract months before it bites. The failure patterns are known, catalogued, and testable in about fifteen seconds.
- Pre-launch wording checked against published failure patterns
- Settlement governance and appeal-window design
- Halt criteria and surveillance escalation patterns
02 — Institutions & funds
A baseline for judging anyone’s forecast.
Before you can price an edge you need a defensible way to measure one. The methods are published in advance, so a counterparty’s accuracy claim can be checked rather than accepted.
- Scoring, calibration and venue-comparison methodology
- Definitions fixed first: resolved, stale, comparable
- Funding, margin and liquidation mechanics in 24/7 venues
03 — Enterprises & supervisors
Competence your compliance function can point at.
Clients are asking your advisers about event contracts now. The credential is an assessed competence standard — explicitly not a licence, and never a substitute for your own supervision.
- Assessed professional competence, renewable
- Conduct, conflicts, records and AI-governance standards
- Team access to standards drafts and the comment process
Founding phase
Help set the standard before it is set without you.
The founding cohort is open. Applications are read by a person, every one gets a reply, and no fee has been set or published. Admission to the cohort is not admission to a credential.