Skip to main content

Trader data and intelligence

Professional-grade trader data, built from the blockchain up

Polymarket publishes every trade and almost none of the meaning. We rebuilt the whole venue from the chain up, so you can tell at a glance what kind of trader you are looking at, where they actually win, and whether any of it is worth following.

1,637,015,470

trades rebuilt

2,598,463

markets classified

3,052,840

wallets seen on-chain

2,179

days, no gaps

126

indicators per trader

We can see 3,052,840 wallets. We compute a full record for the 4,373 that trade enough to measure, and publish the 3,687 whose numbers reconcile against Polymarket's own.

What we hold

Six things we know about every trader

Enough to tell them apart at a glance, and enough to dig into when you want to.

Who they are

Trader type

Directional trader, market maker, arbitrage, or machine speed. Named, not guessed at.

Why it matters. Some edges cannot transfer to anyone following them. This is the first thing worth knowing.

Where they win

Category record

Ranked in 47 slices: category, kind of bet, entry price, stake size.

Why it matters. Almost nobody is good at everything. Overall rank hides which part they are good at.

What they made

Track record

Profit split by how it was earned, return on money actually deposited, win rate, profit factor.

Why it matters. Return on deposits is the honest version. Return on volume flatters anyone who trades a lot.

How rough the ride is

Risk profile

Worst drawdown, daily swing, share of weeks in profit, current streak, best and worst day.

Why it matters. Two traders can finish level and be nothing alike on the way there.

How they trade

Behaviour

Hold time, how positions end, sizing consistency, hedging, and pre-match against in-play.

Why it matters. Tells you what following them would actually feel like, day to day.

Is it worth it

Copy Score

Whether following them has historically left anything after what copying costs.

Why it matters. Free on every trader. Built to rule people out rather than to rank them.

The full set

126 indicators on every tracked trader

Grouped so you can find what you need. Every one of them appears on a trader page or in the glossary.

Trader classification

14

What kind of trader this is, before you look at a single number.

Market makerArbitrageVery high frequencyTransfer-fundedDirectional traderTrading tempo bandHold styleExit routeSpecialist or generalistTenureForm: rising, steady or coolingSizing consistencyActive, quiet or dormantMarket-neutral flow share

Performance and P&L

20

What they made, split by how they made it.

Realised P&LSettled while heldClaimed payoutsUnrealised P&LTotal P&LReturn on deposited capitalReturn on volumeWin rateProfit factorAverage winAverage lossWin to loss size ratioBest dayWorst dayOpen position valueOpen markets7-day P&L30-day P&LLifetime P&LP&L rank

Risk and consistency

12

How rough the ride was, set against the rest of the field.

Sharpe ratioSortino ratioCalmar ratioMaximum drawdownDrawdown ratioDaily volatilityProfitable weeksCurrent win streakActive daysUp days and down daysLargest single-market exposureUnresolved exposure still open

Fees, cash and platform income

13

What trading cost them, and how much of the profit came from the venue rather than the market.

Total fees paidRealised fee rateMaker fill shareTaker fill shareMaker share by valueMaker rebates earnedLiquidity rewards earnedTaker rebates earnedDays earning rebatesCash balanceNet depositedTotal depositsTotal withdrawals

How they trade

16

The habits that never show up in a profit column.

Median hold timeFastest quarterSlowest quarterShare of positions soldPositions closed by mergeBets placed before the eventBets placed in playCashed out before closeMarkets hedgedValue hedgedAverage trade sizeSizing spreadAverage entry priceInventory arriving off-marketOrders per active dayFills per order

Time of day

7

Every hour of every weekday, as a 168-slot grid.

Trades per hour slotBuys per hour slotSells per hour slotValue traded per hour slotOvernight shareActivity spread across the weekBusiness hours or round the clock

Category and market standing

12

Where they win, ranked against everyone who traded the same thing.

31 market categories8 kinds of bet4 entry price bands4 stake size bandsRank in eachSize of the field they beatP&L in eachWin rate in eachReturn in eachValue traded in eachLifetime standing30-day standing

Position by position

12

Every position they have ever held, opened and closed.

Average entry priceAverage exit priceEntry against current priceOpen or closedWon or lostFirst buyFirst sellLast tradeTime heldBuys and sells in the positionPeak capital committedP&L on the position

Live trade feed

10

Every trade as it happens, enriched before it reaches you.

Side and outcomePrice and sizeTimestampMarket and categoryKind of betThe event behind itWhether the event has startedLarge-trade flagCurrent price against their entryMarket status

Market and event data

10

The context every trade is measured against.

Category and sub-categoryBet structureParent eventEvent start timeLive outcome pricesResolution and winning sideVolume over 1 weekVolume over 1 monthLifetime volumeClose and settlement times

Why it matters

Copy trading is the obvious use. It is not the only one.

Copy the right people, not the loudest ones

Good trader data is what separates a real edge from a big bankroll, and shows you which markets to follow someone into.

Learn from traders who are provably good

Watch how the strongest traders size, exit and hedge. Most people read these records without copying anyone.

Also used to

  • Check who genuinely knows a category before you enter it
  • Research any wallet, including ones outside the ranked set
  • Benchmark your own record against the field
  • See what the strongest traders are doing, as a daily market read
  • Know who is on the other side of your trade in a thin market
  • Study how good traders size, exit and manage risk

Glance to detail

Read a trader in three seconds, or in three hundred numbers

The same record, at whatever depth you need it.

1

At a glance

Three seconds

Type, category record and current form, on the row before you click anything.

  • What kind of trader this is
  • Which markets they have a record in
  • Whether they are rising, steady or cooling
  • Whether following them has paid
2

Drill down

The full record

Open a trader and every number is there, over any window you choose.

  • Performance over 7 days, 30 days or lifetime
  • Category by category, with the cohort they beat
  • Risk set against the rest of the field
  • Behaviour, from hold times to hedging
3

Act on it

Straight into the product

The same stored record drives every screen, so nothing contradicts anything else.

  • Boards for who performs in one specific market
  • A live feed of what tracked traders are doing
  • Copy settings that respect what the data says
  • A daily index across the strongest traders

Copy Score

Is any of it worth following?

Copying isn't free — you buy after they do, and there are fees. Copy Score is what was left on their finished trades after that. The cost is much the same whoever you copy; what changes is whether their edge clears it.

Great traders aren't always great to copy. Copy Score tells you which ones are.

You seeWhat it means
ProvenCopying them has worked on their finished trades.
AheadThey finish ahead of what copying costs — but only just.
CautionThey make money. It doesn't survive being copied.
LosingThey've been losing on their own trades, before copying costs even start.
Not comparableThey close out early, so we can warn about them but never rank them alongside hold-to-the-end traders.
Thin historyToo few finished trades to stand behind. A big return either way can still be one bet that came in.
Not scoredThe classifier has not graded this wallet. That is not a middling score; it is the absence of one.

Free on every trader, and not one of the things we rank people by. It works to rule traders out, not as a shortlist to work down. Read the Copy Score explainer or see how it has performed.

Where it shows up

One record, every screen

The same stored answer drives all of it, so nothing contradicts anything else.

Under the hood

Why this is hard, and why ours is different

Everything on Polymarket is public, so people assume the hard part is done. It is not.

The chain records fills, not positions

Every buy has to be matched to whatever closed it before a wallet has a cost basis or a real result.

Money moves without a trade

Transfers, merges, late claims and venue rewards all shift a balance without anyone predicting anything.

A trade means nothing without its market

Polymarket does not publish a consistent category or bet type, so we work it out for every market ourselves.

What most sources do not do

Every market sorted

99.99% of 2,598,463 markets carry a category and a kind of bet.

Tied to real events

1,639,007 markets matched to the fixture behind them, so pre-match and in-play are separable.

Venue rewards split out

Rebates and liquidity rewards are held apart from money won by trading.

Ranked within a market

Not one leaderboard position, but a standing in each category a trader has traded.

And it never stops

Live

Prices and the order book

Minutes

Market results, as they settle

Hourly

Trader results and category records

Daily

Behaviour, rankings and classification

How we know it is right

Checked against Polymarket's own numbers

If we come out higher than the venue does, the trader is not published at any margin. Understating is the safe direction to be wrong in.

One answer, read everywhere

Classification is worked out once and stored, so no two screens can show different verdicts about the same trader.

Thin evidence shows as thin evidence

Where there is not enough finished history, the page shows nothing rather than a number that looks like discipline.

No gaps in the history

Every day since September 2020 is indexed, 2,179 of them without a break.

Three counts, three different things

Seen

3,052,840

Wallets with on-chain activity we can read. Most have traded a handful of times and there is nothing meaningful to measure.

Tracked

4,373

Wallets that trade enough to compute a full record for, across every indicator on this page.

Published

3,687

Tracked wallets whose numbers reconcile against Polymarket's own. The rest are held back until they do.

Seeing a wallet and having a record worth publishing about it are three steps apart. Warehouse counts as of 20 August 2026, and the last two move as new wallets are computed and reconciled.

Limits

What this data does not do

  • A record of what has already happened, not a forecast.
  • A wallet is not always one person, and one person is not always one wallet.
  • No score is not a bad score. It means we will not claim something we cannot stand behind.
  • We do not rank traders by Copy Score, and we do not publish a list of traders to copy.
  • Your fills will differ from theirs. You buy after they do, at whatever price is there then.

FAQ

Questions about our trader data

Where does Polycopy's Polymarket trader data come from?

Polycopy's trader data comes from the Polygon blockchain itself, rebuilt into 1,637,015,470 individual trades. Polymarket runs on a public chain, so every trade, transfer and settlement is recorded permanently. What the chain does not provide is meaning: it records that an address bought shares at a price, not what that address now holds, what it paid on average, or whether the market was an NBA game or an election. We read those raw records, match every purchase to whatever closed it, attach the market and the event each trade was about, and roll the result up per trader. We also use Polymarket's own public market data and live sports schedules to know what each market was and when it started. Nothing here comes from a leaderboard someone else computed, which is why we can measure things a leaderboard cannot show. Figures as of 20 August 2026.

Why is measuring a Polymarket trader harder than it looks?

Measuring a Polymarket trader is hard because a wallet's balance changing is not the same thing as that trader doing well. Value moves in and out of a Polymarket wallet for reasons that have nothing to do with predicting anything: shares arrive by transfer, positions get split and merged, winnings are claimed weeks after a market settles, and the venue pays some traders rewards simply for providing liquidity. Any of those can make a wallet look profitable without a single good call. On top of that, a trade means very little without knowing what the market was about and when it started, and Polymarket does not publish that consistently. Separating genuine trading results from everything else, then attaching each trade to the market and event behind it, is most of the work. It is also why two sites can quote different profit figures for the same wallet.

How is this different from the Polymarket leaderboard?

A leaderboard ranks traders by profit, which answers who won rather than who is good. The largest profit on any board usually belongs to whoever moved the most money, and it says nothing about whether that result came from judgement, from scale, or from a strategy that only works for the person running it. It also treats a trader as a single number, when most traders are strong in one or two market types and unremarkable everywhere else. Because we attach a category and a bet type to every market, we can rank a trader inside each one: 47 separate slices per trader, over their full history and over the last 30 days. That turns a leaderboard position into something more useful, which is what a trader actually has a record in, and where they have none.

What is the trader classification system?

Polycopy's trader classification has two separate layers, and we keep them apart on purpose. The first is descriptive and carries no judgement: how a trader exits positions, how consistently they size, how many categories they trade, how long they have been active, and whether their results are rising, steady or cooling. None of that makes a trader good or bad, it just tells you what following them would be like. The second layer is a caution. Some traders earn in ways that do not survive being mirrored, such as market makers who collect the spread you would be paying. We flag those types so a low score has an explanation attached. A flag is not a verdict, and the absence of one is not an all clear, it only means we have not identified that wallet as one of those types.

How current is the trader data?

Prices update continuously, and the rest of the trader data refreshes on a schedule through the day. Market results are checked and applied within minutes of a market settling, so a page does not sit showing an open position in something that already finished. Trader results and category records refresh hourly. The slower-moving parts, meaning behavioural patterns, rankings and classification, are rebuilt daily, because those describe habits rather than events and recomputing them more often would make them flicker without telling you anything new. Every figure carries the date it was computed, and where a number is stale we would rather show its age than quietly present it as current. The underlying history runs continuously from September 2020, which is 2,179 days without a gap, so a trader who was active three years ago is measured on the same basis as one who started last month. New markets are sorted into a category as soon as they appear rather than waiting for the next daily pass.

See it on a real trader

Open any wallet and read the whole record, from what kind of trader they are to where they actually win.

Browse traders

Keep reading

Disclaimer: Polycopy is an independent third-party tool and is not affiliated with or endorsed by Polymarket. Nothing on this page constitutes financial, investment, or trading advice. Past performance does not guarantee future results. All trading involves risk of loss — only trade with funds you can afford to lose. Trading fees apply to all trades executed through Polycopy (a flat 0.5%). Full terms.

Polymarket trader data: what we measure and how | Polycopy