Docs

Tradar is a research desk for public Polymarket flow. The usual loop is: find wallets that have been winning recently, read how they actually trade, follow a small set, then watch new fills and copy ideas as they happen.

• Start on Leaderboards. Open Smart Money and look at recent skill, then check overall profit and win rate beside it.
• Open a wallet. Confirm they win in categories and price ranges you would actually copy. Then follow.
• Use Watchlist and the bell for new fills the minute they land.
• Optional: Live Copy ranks new trades from wallets that are strong in that probability and category.
• Use Markets to see which books tracked whales are in, and Insider Alerts as rare leads to research.

Home is a 24-hour snapshot of that loop: copy ideas, fills from wallets you follow, and unusual bets. The in-app guide is the short version of this page.

A WSJ analysis showed 67% of Polymarket profits go to 0.1% of accounts. On your own, you are statistically likely to lose. Tradar is built so you can follow the 0.1%: find who consistently wins, and spot unusual bets.

The product does four jobs. Leaderboards rank recent skill. Wallet profiles show how a trader actually behaves. Live Copy and Watchlist surface new fills while they are still open. Markets and alerts point at books and bets that look crowded or unusual.

Scores are research tools to help you judge wallets and trades. You still size the bet, pick the price, and decide whether the market is one you understand.

Open Leaderboards

Polymarket prices are shares that pay $1 if that outcome wins and $0 if it loses. A Yes share at 30¢ costs $0.30. If Yes wins you receive $1, so you keep $0.70. If Yes loses, the $0.30 is gone.

Price is implied probability

A 30¢ Yes is the market saying "about 30% chance." A 75¢ Yes is "about 75%." The other side is 1 minus that price, before fees and spreads.

The payout if you are right is 1 ÷ price. At 25¢ a winning share returns 4× your stake. At 80¢ it returns 1.25×. Longshots pay more when they hit and lose more often.

The breakeven win rate is the price

To break even over many bets at the same price, you need to win about as often as the price says. At 25¢ you only need to be right about one time in four. At 75¢ you need to be right about three times in four.

Why win rate is not skill on its own

Two wallets can both show a 50% win rate. One buys 25¢ underdogs and is printing money. The other buys 75¢ favourites and is slowly bleeding. Same win rate, opposite skill.

• 50% at 25¢ is excellent. Breakeven was 25%.
• 50% at 75¢ is terrible. Breakeven was 75%.

Always read win rate next to the prices they take, plus PnL, ROI, and Smart Score.

PnL and ROI

PnL is dollars made or lost. A whale can have huge PnL from size even with a thin edge. A smaller book can have modest PnL and a stronger process.

ROI is return versus what they put at risk: profit divided by cost. +40% ROI on $2,000 risked is a different story from +40% on $50. Tradar shows both dollars and rate so you can see scale and quality together.

A single settled trade

Buy $100 of Yes at 40¢. That is 250 shares. If Yes wins, those shares pay $250. PnL is +$150. ROI is +150%. If Yes loses, PnL is -$100 and ROI is -100% on that ticket.

Edge, in one line

You have edge when your true chance is better than the price you pay. If you think Yes is 50% and you can still buy it at 38¢, the leftover 12¢ is the edge. Copying someone only works if some of that leftover is still there at your fill.

That is why Tradar cares about recent return, how well winning trades paid, and whether a wallet is actually good in the same category and price band as the trade you are looking at.

Open vs resolved

An open position has not settled. Mark-to-market PnL can move with the price. Resolved means the market closed and the shares paid $1 or $0. Smart Money ranks on recent resolved outcomes. Live Copy is about fills that are still open.

Open Smart Money to see wallets ranked by consistent gains, targeting those with a winning strategy. The number on the board is Smart Score, our algorithm for recent skill over about 30 to 90 days.

You can also sort by overall profit or win rate. Those are raw Polymarket stats. Smart Score is the skill ranking. Use them together: a high score with almost no settled trades is weaker than a high score with a long, stable book.

Category experts

Category boards start from Smart Money, then keep wallets whose recent book is mostly that category. Weather #1 is a weather specialist with a strong Smart Score whose recent book is mostly weather.

Rank on a category board follows overall Smart Score among those specialists. Read category win rate and category ROI next to the score to see the edge in that sleeve.

Categories include Politics, Sports, Crypto, Weather, Macro, Tech, Culture, and Other. A wallet can be elite in one and average in another. That is why the profile exists.

Open Leaderboards

Use a profile to judge if and when to follow. A high Smart Score is the start. Confirm categories and price bands before you follow.

• Overview: official lifetime PnL and volume from Polymarket, recent-window performance, Smart Score, and rank.
• Behaviour: how they size, how often they trade, and whether the book looks steady or bursty.
• By probability: which price bands they take. A 20 to 40¢ specialist is a different copy target from a 70 to 90¢ favourite buyer.
• Categories: where they win and lose. Follow someone in the sleeve you would actually copy.
• Copy trade: how their recent fills would have looked if you had copied them in their strong cells.
• Recent trades: the live tape. Click a fill to open it on the trade feed.

Official PnL is lifetime profit and loss from Polymarket. Smart Score uses the recent resolved window. A wallet can be down lifetime and still rank on Smart Money if the last 30 to 90 days are strong. The reverse is also true.

Toggle 90d and 30d on the profile when both exist. Prefer the longer window when it is complete. A short window can look hotter and fall apart.

Pick a wallet

Follow a wallet and the quick feed shows new trades the minute they get placed. The same fills land in the bell. Suggested wallets are high Smart Money names.

Follow a small set you have actually read. A long list of names you never open is noise. Pin the ones you want on the live position feed if you only want a tight tape.

The live feed is the full tape for wallets you follow. Each row is a position. Open means it has not settled. Resolved means the market closed. Newest activity is first. Prices and profit update while the tab stays open. Click a row for trade detail: entry, current mark, size, and the wallet behind it.

Home's Watchlist preview is that same tape, shortened. Clicking a row opens the watchlist live feed with that fill already selected.

Open Watchlist

Live Copy watches public fills and ranks trades from wallets that are experts in similar probabilities and categories. Copy Score is for this trade. A great weather longshot from a weather specialist can score well even if that wallet is average on politics favourites.

Standard and Moonshot

Every live candidate is assigned to one mode.

• Standard: higher-confidence copies. Higher win rate, more modest payouts. Built for trades where the wallet's fair chance still looks comfortably above the price.
• Moonshot: lower-confidence, higher-ROI copies. Lower win rate. When they hit, they pay more. Split stake across many of them.

A trade appears in Standard or Moonshot, not both. Best now is the tight list of open trades that still look worth following. Research shows a wider book you can explore.

Track record

Track record is the resolved history of those suggestions, scored as they looked at signal time. Use it to see how Standard and Moonshot have done over a window. It audits the feed. The next ticket still has to clear on its own.

Copying still needs a fill. If the price has already moved, the leftover edge may be gone. Treat the listed price as a guide for sizing, then check the book yourself.

Open Live Copy

Markets shows which books top whales are in, and what they are betting on. The universe is open positions from tracked featured wallets.

• Popular ranks books by how many tracked whales hold them.
• Suspicious ranks books where those whales lean against the current price.

Market Whale score is a suspicion guide that a book may be worth investigating. A high score means several strong wallets are on one side, often cheaper than the mid. Open the market, read the question, and decide if the lean makes sense.

Insider alerts on the same market can raise the score further. Whale columns and insider rows are separate so you can see crowd lean and unusual flow side by side.

Open Markets

Alerts flag unusual bets: new wallets putting serious size on low-chance outcomes, and similar patterns that look insider-like. The score measures how unusual the bet looks. Treat each alert as a lead to research.

Sports and crypto books are left out. Those markets move on public games and tape, so the same pattern is much noisier.

Read the market, see who else is in it, and decide if the story is even tradable for you.

Open Insider Alerts

This is the qualitative breakdown. Exact weights, shrinks, and cutoffs stay internal so the boards cannot be gamed from the docs. What follows is what a trader should actually use.

Smart Score

Recent skill over about 30 to 90 days. Lifetime profit is shown separately on the profile.

• Recent return: the main ingredient. How much they made versus what they risked in the window. Thin samples are pulled toward average so a handful of wins cannot look elite.
• Payoff quality: whether winning trades actually paid. High win rate with negative return does not rank.
• Consistency: breadth of resolved outcomes, whether one ticket dominates the PnL, and how the book holds up after a peak.
• Recent dollar PnL: a lighter term for the sign and size of the window. It does not replace ROI.

Boards only rank wallets with a complete enough recent window and a positive recent edge. Very small samples stay off the skill board or show as weak / provisional. Some huge books cannot finish a full 90-day scan and may use 30 days or sit out.

You always see raw ROI on the page. The score uses a conservative internal version of return so one spike cannot carry the rank.

Category expert score

First a filter: the category has to be a large share of their recent trades, with enough settled outcomes in that sleeve. Then those wallets are ordered by overall Smart Score. The sleeve also has its own quality check: category return, win rate, and consistency.

Copy Score

Built for one fill. It asks: is this wallet good in this category and price band, and is there still something left if you copied near their entry?

• Specialist edge: win rate versus the prices they actually paid, in a nested cell: overall, then category, then probability band, then both together. Thin cells borrow strength from the parent cell.
• Fair chance: their entry blended with that specialist edge, kept in a sane range.
• Expected copy return: leftover edge if you copied at their average entry. If the quote has already moved, the market ended, or the price is glued to 0¢ / 99¢, the score collapses.
• Evidence: how much history sits in the cell that is doing the work. Strong / moderate / weak.
• Conviction: this fill versus that wallet's own usual size.
• Consensus: other qualified wallets on the same side help a little. Opposed flow hurts.
• Stability: whether the last month still looks positive versus the two months before it.
• Concentration: one lucky outcome carrying the profits downgrades the signal.

Standard ranks the higher-confidence side of that split. Moonshot ranks the higher-ROI side, usually in cheaper price bands. Track record freezes the score at the first actionable moment.

Suspicion Score

How insider-like a bet looks. It does not use the later result or the payout multiple.

• Absolute stake: dollars at risk. The main lever.
• Wallet novelty: new or quiet wallets. Unknown history gets no freshness bonus.
• Relative conviction: stake versus that wallet's normal size, when we have a real baseline.
• Entry odds: cheaper, longer-shot entries add a little context.
• Thesis: the same wallet stacking related markets over recent weeks.
• Coordination: several wallets hitting the same market or thesis.
• Market context: stake versus the book's volume.
• Episode: several fills that read as one order.

A few strong dimensions together can unlock the top of the scale. There are also named patterns (fresh wallet plus a longshot, a cluster piling in, a huge mid-price ticket) that can carry a floor on their own. Hedges, spray across many longshots, and chasing a price already moving against them all get cut down. Public macro books are dampened. Sports and crypto stay off the alert feed.

Market Whale score

Market-level suspicion from tracked whales' open books.

• Headcount: how many tracked whales are in the book. The main lever. Many wallets on one side beats one tiny longshot.
• Contrarian lean: whether the dominant side is cheaper than the mid. Extreme 2¢ tickets are flattened so they cannot outrank a real 30¢ lean.
• Side concentration: how one-sided the whale weight is.
• Smart-weighted consensus: a light tie-break using those wallets' recent Smart Scores.
• Open notional: dollars on each side, used more for display and close ties.

Active insider alerts on the same market can lift the displayed score. Popular is "who is in the book." Suspicious is "who is leaning against the price."

Tradar is built on public Polymarket data: fills, open positions, closed positions, and official headlines.

• Home is the last 24 hours.
• Smart Money prefers a verified 90-day window and falls back to 30 days when that is the complete one.
• Copy profiles are built from about 90 days of resolved history, with the last month checked against the two before it.
• Watchlist / live feed default to recent days and widen if you open an older fill.

Some wallets are too large to scan completely in a long window. Those show as capped or drop off a skill board until a shorter complete window exists. Empty or tiny windows are treated as "not enough data" rather than a skill score of zero.

Official lifetime PnL, volume, and markets traded come from Polymarket's own leaderboard. Local reconstructed stats are used for recent windows, categories, and calibration. If those two views disagree on a huge book, trust official dollars for the headline and the recent window for skill.

Sign in with Whop to keep a watchlist across devices. A paid seat unlocks the product.

Paid features are gated on the server. That typically includes Live Copy, insider alerts, whale-overlap markets, and unlimited wallet analysis. There is no public API for third-party apps. The bell is in-app only. There is no email or push yet.

Billing and cancellation run through Whop. Theme (light / dark) is saved on your account when you are signed in.

Tradar is an analytics product. It does not place trades, hold funds, or give financial advice. You are responsible for your own risk limits. See Terms and Privacy.

Terms of Service

Privacy Policy