PolyScope covers Polymarket, Kalshi, and the broader prediction-market space — guides for people getting started, reporting on regulation and platform disputes, and original data research into how these markets actually trade.
Who’s Behind It
PolyScope is written and run by Stepan Kosarev. Full background on the author page.
Data and Methods
Our market and wallet research is built on Polymarket’s own public infrastructure: the Gamma API and Data API for market and position data, the CLOB API for order-book and pricing data, and on-chain transaction history for wallet-level analysis. Where a piece compares Polymarket to a traditional sportsbook, we use that book’s own historical odds, matched by timestamp — we don’t name the specific vendor or method used to source that comparison data, since publishing it would make the pattern trivially easy to arbitrage away.
The Smart Money Dossier tracks wallet performance directly from Polymarket’s Data API — position history, realized P&L, and trade timing — and rebuilds every profile from fresh on-chain data overnight. The “data as of” date on that page reflects when that rebuild last ran.
Original Research
Beyond guides and news coverage, we publish research built on our own analysis of on-chain data, not just reporting on what Polymarket or other outlets say:
- I Filtered 512 “Profitable” Polymarket Wallets Down to 314 — the methodology behind the Smart Money Dossier, including a wallet that tested as genuinely skilled and was still uncopyable at any realistic position size.
- How a Profitable Polymarket Esports Wallet Really Trades a Sharp Book’s Odds — a measured decay curve showing how long a pricing edge against a sharp sportsbook actually survives, plus a full-population study finding that 2% of esports wallets capture 27% of all esports profit on the platform.
- Inside a Polymarket Esports Wallet That Turned $580K Trading LCK and CS2 — a single-wallet trade-history breakdown.
Editorial Standards
A few rules we hold ourselves to:
- Wallets we profile are anonymized. We don’t publish handles or exact profit/loss figures for individual traders we didn’t interview directly — ranges and patterns instead, so the research doesn’t double as a way to dox someone’s trading account.
- Limitations get their own section, not a footnote. Every data piece states what the numbers can’t tell you — sample size, timing precision, correlation versus causation — next to what they can.
- We correct mistakes in place rather than leaving a wrong number live because the post already shipped.
- Affiliate and referral links are marked as such (
rel="sponsored"), including our own links to Overdog.
None of this is financial advice. See our Disclaimer for the full version.
Contact
Corrections, tips, or data questions: [email protected].