Is there an independent Polymarket analytics vendor?

Yes. Convexly is an independent analytics and audit layer for Polymarket. Independent has a specific meaning here: no venue-derived revenue and no fee-sharing arrangement with any exchange, so the analysis holds no stake in trading volume. Convexly does not take custody, broker trades, or route investment advice. The free wallet check is the entry point: paste any Polymarket wallet and you get its realized entry edge with a 95% confidence interval and the number of resolved positions behind it, plus a percentile against an 8,656-wallet reference cohort.

What independent means here

Independence is a revenue fact, not a slogan. Convexly earns nothing from Polymarket trades happening. It removed its venue builder code on 2026-06-11, so it collects no fee share and no volume-tied rebate, and it is not owned by or operated for any exchange. Revenue comes from analytics subscriptions. A tool that profits when volume rises has a different incentive than one whose only job is to measure whether a track record is skill or luck.

That is the whole reason an independent audit layer exists. The venue publishes the market. The audit layer grades the skill claim behind a wallet, with no reason to inflate anyone's numbers and no product line that depends on you placing a trade.

How it differs from venue-owned or fee-taking tools

These are complementary tools for different jobs, not competitors doing the same thing worse. Venue dashboards, flow trackers, and de-anonymization tools each do one job well: they surface activity, wallet identity, or raw market data. An independent audit layer answers a separate question:

• Is a wallet's record distinguishable from chance at a frozen statistical bar, or is it consistent with luck?
• What is the realized entry edge, reported with its 95% confidence interval and the count of resolved positions it rests on?
• Where does the wallet sit against a fixed reference cohort, rather than against a shifting live leaderboard?

Convexly reports skill as distinguishable from chance (or not). It never presents a wallet as one to mirror, and it carries no incentive tied to how much anyone trades. Use the flow tools to see who is moving; use the audit layer to check whether the record is skill.

Published methodology and published negative results

Independence is only credible if the method is inspectable and the failures are on the record. The Edge Score methodology paper documents the composite ranker, the fat-tail handling, and the permutation null. The negative-results page reports tests that did not clear their bar. The most important one: the per-wallet temporal holdout, run on 2026-04-27, returned a forward Spearman of +0.11 (95% CI 0.05 to 0.18), below the +0.30 threshold set in advance.

That is why Edge Score is presented as a cross-sectional descriptive ranker on this cohort, not as a forward-skill signal for any single wallet. Reporting the miss is the point: an independent vendor with nothing to sell you on a trade has no reason to hide it.