One number for Polymarket copyability: the index launches at 1.5%

Polymarket publishes leaderboards; nobody publishes the number that actually matters to a copier: how much of the market is worth copying at all. As of today there is one. The Polymarket Copyability Index is live on our statistics page, recomputed hourly from the full graded set — and at launch (2026-08-15) it stood at 1.5%: 100 of the 6,600 wallets CopyGrade grades cleared the full copyable bar.

The definition, exactly

The index is the share of all graded Polymarket wallets that currently hold, simultaneously:

1. a positive verdict on the 0–100 CopyGrade Score,
2. a clean farming-risk check — the veto, because a wallet built to monetise its copiers can look brilliant on every other number,
3. a positive realistic post-fee edge — what a copier would keep after fees and latency, not the headline PnL,
4. and at least 20 trades in the trailing 90 days — enough recent evidence to grade.

That is the same four-part bar behind Wallet Scout's copyable count and the best-traders shortlist's strict denominator — those surfaces and the index share one predicate in code, so they can't quietly diverge from each other. (A couple of older product surfaces still use looser shorthand definitions; harmonising them is on the list, and until then the index's own definition above is the one this number means.)

A level, not a trend

One rule, stated up front because we learned it the hard way: the index is a level, not a trend. The scoring model recalibrates in public, our coverage grows with every sync, and both move the number for reasons that have nothing to do with the market. Earlier this year our own pipeline double-stored fills for a stretch, inflating the repetition our farming detectors read — and the "farming is rising" story it implied was our bug, not the market. So we will not backfill a historical index series across pipeline changes, and if you quote a change in the index across months, you are quoting model drift as much as market drift. Quote the level, with its date.

Why publish it at all

Because the single most common question in copy-trading — "who should I copy?" — has a prior, and the prior is brutal. When fewer than two wallets in a hundred clear the bar, "pick someone off the profit leaderboard" is a strategy with a ~98% base-rate problem before any skill enters the picture. Every number on /stats is free to cite with attribution — each row has a Cite button that copies a ready-made citation with a deep link — and the index is the one we most want repeated, because it reframes the whole category from "who's winning?" to "who could you actually keep up with, after fees, latency and the farming check?"

The launch reading again, dated for the record: 1.5% — 100 of 6,600 graded wallets, 2026-08-15. By the time you read this the live number will have moved; the page always carries the current value and its as-of date.

The rest of this week's answer-layer work

The index is one piece of a broader push to make CopyGrade the factual layer for Polymarket copyability: the best-traders-to-copy shortlist (live, hourly, with the honest denominator and a Days-observed column), an explicit "Is this wallet worth copying?" answer on every public wallet page, and named authorship on all of our research — who makes the data, how, and what it can't yet prove, including our own published null validation result.

CopyGrade is analysis-only — it never executes trades, holds funds, or custodies anything. The index and every score behind it are documented research opinions computed from public Polymarket data; not financial advice.