Methodology

Every Polymarket trade settles on-chain as an OrderFilled event emitted by Polymarket's exchange contracts on Polygon. SharkTag reads these events directly from the chain — the primary source — across both exchange generations: the original CTF Exchange and NegRisk Exchange (2022 → April 2026) and the current v2 contracts that replaced them. Our historical coverage begins with Polymarket's first orderbook trade in November 2022 and extends to the present, refreshed throughout the day, staying within hours of the chain tip.

Because the dataset is built from public blockchain data, every number on SharkTag is independently verifiable: any fill we report can be checked against the transaction that produced it, byte for byte.

Each fill carries the maker, the taker, the outcome token, the USDC amount, the share amount, and the fee. From these raw fields we derive, ourselves, every metric on the site: trade price (USDC ÷ shares), trade side (buy or sell of an outcome), per-wallet volume, fill counts, market reach, maker/taker balance, and activity timing. Exchange router addresses are excluded from wallet statistics — they are plumbing, not traders. Multi-leg matches are attributed to the real counterparties behind them.

Everything on SharkTag is labeled as one of two things, and we never blur them:

Measured facts are deterministic aggregations of on-chain events — trading volume, number of fills, markets touched, days active, maker share, fees paid, realized flows. Two people recomputing them from the chain would get identical numbers.

Projected attributes are our estimates layered on top: trader archetypes such as market maker, systematic algo, whale, or retail, each with an explicit confidence score. They are behavioral reads, not facts, and the platform always displays them as such. When we do not know, the confidence says so.

We describe behavior; we do not sell prophecy. A wallet's history of sharp entries is a measured fact about the past — it is not a guarantee of future performance, and we do not present "smart money" signals as predictions. SharkTag exists so you can see, with evidence, who is active in prediction markets and how they behave.

• Polymarket orderbook history from November 2022 → present
• 1.3B+ on-chain fill events indexed from Polygon
• Hundreds of thousands of distinct trading wallets analyzed
• Both exchange generations (v1 CTF/NegRisk and current v2), one unified dataset
• Continuous ingestion — the dataset tracks the chain within hours

The SharkTag app gives you the full time-series of Polymarket activity since inception, a wallet-by-wallet explorer, and per-wallet profiles. Exports and a commercial data API are on the way — see pricing.