DEMO PREVIEW — Markets and probabilities shown are sample data, not real model output. Our ML engine is in development.

Get notified at launch →
Get notified at launch
← Back ABOUT MISPRICEHQ

Quant research
for prediction
markets.

MispriceHQ is building independent models to score event contracts on Kalshi, Polymarket, Manifold, and other venues against an independent fit — then publish the gap, sized and signed, calibrated against what actually happens. The engine is in development; the site today is a design preview with sample data.

WHAT WE DO

We're building independent statistical and ML models for prediction-market questions across macro, crypto, technology, politics, sports, and culture. Once live, each model will be graded against realized outcomes — Brier scores, reliability diagrams, and per-category calibration — using the approach on the Methodology page.

Where we have edge, the UI will say so in blue. Where we don't, it will say so in grey. That honesty is the product. Until the engine ships, everything on the site is clearly-marked sample data.

WHAT WE DON'T DO
· We don't take positions. No house book, no proprietary trading, no skin in the game.
· We don't route orders. We link out to platforms — those are regulated derivatives venues and your relationship is with them, not us.
· We don't dress up uncertainty. If a category is noisy, the calibration page admits it.
· We don't sell signals as advice. Every page carries a disclaimer for a reason.
DEVELOPMENT ROADMAP

Where the product is and what ships next. Dates are targets, not promises — we'll update this page as milestones land.

Now
Design preview + research library The site, the published methodology, and sourced explainers are live. Market data shown is sample data, clearly marked. No model output is published.
Next
Data ingestion + first model fits Live price ingestion from public platform APIs, the first category models, and a backtest harness scored on held-out resolved markets.
Later
Live, calibrated probabilities Public model output replaces sample data — only for categories where we beat the market baseline out-of-sample. Calibration is published with every number.
EDITORIAL STANDARDS

Our written research follows the same honesty rules as our models. Specifically:

  • Every statistic is linked to a primary or reputable secondary source. If we can't source it, we don't state it.
  • Articles carry a visible author, publish date, and last-updated date. Material corrections are noted.
  • We separate fact from interpretation, and we say plainly when something is uncertain or still developing.
  • Financial topics carry a risk disclaimer. We don't give individualized financial advice.

Found an error? Email research@mispricehq.com and we'll correct it.

AFFILIATE DISCLOSURE

As of today, MispriceHQ has no affiliate or referral relationships and earns nothing from any platform. Links to platforms like Kalshi, Polymarket, and Manifold are plain, untagged links to their homepages. In the future we may join prediction-market referral programs and earn a commission when you sign up through our links. If and when that happens, we will mark those links clearly (rel="sponsored" plus a visible notice) and update this disclosure. Affiliate relationships will never influence our model output or which markets we flag — the numbers are computed independently of any commercial relationship.

GET LAUNCH UPDATES

One email when live predictions launch. Nothing else.

No spam. One email when live predictions launch. We never share your address.

NOT FINANCIAL
ADVICE.

MispriceHQ publishes independent probability estimates for educational and research purposes. Models can be wrong; market microstructure can shift overnight; tails do tail. We don't take positions, we don't route orders, and we are not your fiduciary. Risk of total loss is real and material.

Event contracts on Kalshi are regulated derivatives traded under CFTC oversight; other venues operate under their own jurisdictions. By using this site you acknowledge our methodology, our calibration, and that the numbers above describe the past — not the future.