Pick a trader. Polly catches their fills onchain within a second, checks the order book, and repeats the trade on your account under your rules.
Polymarket trades probabilities, not shares of a company. Every market is a question with two outcomes, and the price answers it with a number.
A share costs between 0 and 1. A price of 0.62 means the market puts the event at 62%.
If the outcome happens, every share redeems for $1. If it does not, it redeems for $0.
Buy YES at 0.62 and be right — you collect $1. That is 38¢ of profit per share.
Two consequences shape everything else: order books are thin, and prices near the edges (0.05 or 0.95) behave nothing like prices in the middle.
The catalogue is built from live onchain fills. You see volume, capacity and style — not just a flattering ROI.
Size, caps, tolerated slippage, price range. A dry run over their history shows what would have been copied.
Leader fills → book check → execution, or a refusal with a plain-language reason. All of it lands in the feed.
Not a showcase of made-up numbers, but what actually happened on the platform over the last few days. The boards fill themselves; appearing on the named ones is opt-in.
Real trades taken all the way to resolution: the market played out, shares redeemed at $1 or $0. The feed is anonymised — amounts bucketed, timestamps rounded to the minute, no copier names.
Ranked by closed result: only markets that already resolved, where the payout is a fact. Open positions do not count — that is hope, not a result.
Who gathered the biggest flock. Counted by active members — those who traded in the last 30 days: the programme rate depends on exactly them, so the ranking must measure the same thing.
There is no password anywhere: Polly recognises you through Telegram. Come in whichever way suits you — in the messenger on your phone, or in a browser on a big screen. Same account, same subscriptions, same history.
Copy-trading bots usually read Polymarket’s public trade feed. We measured it, and it turns out to serve a frozen snapshot.
| Source | Lag |
|---|---|
data-api /trades?limit=500 | 136–400 s |
data-api /trades?limit=100 | 214–227 s |
data-api /trades?limit=499 | 5–40 s |
Onchain OrderFilled on Polygon | 0.1–1.5 s |
Across 70 seconds of continuous polling with cache busting, the feed did not refresh once. Copying from that source means entering at yesterday’s price. Polly listens to the chain directly.
Summing cashPnl across a wallet’s positions produces a catastrophe. Losing legs of resolved markets sit there forever while winning ones vanish on redemption — so the sum counts every loss and not a single gain.
Cash flow: sells + redemptions − buys + open positions. Same wallet, same data. And when the history does not fit the API window, Polly shows a dash instead of a number.
Copying one-for-one is almost always wrong: the leader may be running a bankroll a hundred times yours. There are four modes.
| Mode | What it does | Who it suits |
|---|---|---|
| Fixed | The same amount on every trade | Beginners: predictable and easy to reason about |
| Ratio | A percentage of the leader’s trade size | When you want to follow their swings |
| Bankroll match | The same share of capital they risked | The only mathematically sound one |
| Bankroll fraction | A fixed percentage of your own capital | The cautious: size grows with the account |
Caps apply on top of any mode: per trade, per market, per event and per day. When a rule hits a cap, Polly shrinks the order instead of refusing outright.
A skip is a guard doing its job, not a failure. Each one lands in the feed with a plain-language reason and the exact number behind it.
The book shifted while the trade was in flight. Entering at the new price is a different trade.
The gap between bid and ask would eat the edge long before the market resolves.
Your size would not fit without you moving the price against yourself.
Near the edges risk and reward are asymmetric, and slippage bites hardest.
Entering five minutes before the outcome is not copy trading, it is roulette.
Every market has a minimum order. Smaller than that and the order is simply rejected.
Per market, per day, or on open positions. Your rule, your boundary.
Their order rested in the book and paid no spread. Repeating it as a taker means paying it for them.
Equity, daily result, the account curve, and open positions with live price and P&L on each.
A catalogue with cash flows, capacity and style. Polly says outright when a number cannot be trusted.
Sliders for size, slippage and filters. “Check against history” replays them over the leader’s recent fills.
Everything that happened: copied, skipped, and why. At the top, median slippage across all your copies.
The fee is charged by the Polymarket protocol itself, and only to the taker — whoever consumes an order from the book.
The p(1 − p) term peaks at a price of 0.50 and falls to zero at the edges. The non-obvious consequence: coin-flip trades are the most expensive ones to copy. Polly warns you when a leader’s share of them runs high.
Polly’s own fee applies to live accounts only, and it is charged on notional rather than on profit. A demo account carries none of it — sizing up a leader costs you nothing.
“Price moved 3.2¢ against a 2¢ limit.” Every skipped trade comes with a reason and a number.
How many copiers a leader can carry before they start slipping against each other.
Median across all your copies. The metric that decides quality stays in plain sight.
The leader risked 2% of their capital — Polly risks 2% of yours. Not a blind one-for-one.
Spread, depth, the real average price for your size. Worse than your limit means no trade.
When a wallet’s history is too short for an honest number, you get a dash instead of a pretty one.
Onchain fill detection, the rules engine, execution priced against the live book, the leader catalogue, the feed with reasons, the bot and the app. The numbers on this page come from Polymarket, not from a mockup.
While you are sizing a leader up, your account can be a demo one: the money is virtual, everything around it is real. It is how you test someone else’s strategy under your own rules without paying for the test.
Nowhere — there is none. A demo account is virtual: nothing to fund, nothing to withdraw. What is real are the leaders’ fills and the order book used to price execution.
Because your filters are working. Coverage of 30–60% is normal for careful settings. If you want more, relax the rules — the feed shows which reason fires most often.
You lose alongside them. Copy trading does not protect you from bad decisions, it repeats them. That is why there is a drawdown stop, a daily cap and a one-tap pause.
Because the leader’s bankroll may be a hundred times yours. Their $5,000 bet is 1% of their capital and 100% of yours. Bankroll matching exists for exactly this.
No. Polymarket is closed to the United States, Russia and roughly forty other jurisdictions, and working around that breaks their terms. Reading data is open everywhere; trading is not.
No. Polly is an execution tool, not an adviser. It does not pick the leader for you and does not judge whether you should be doing this at all.
A claim on an outcome. Redeems for $1 if the outcome happens, $0 if it does not.
The moment a market settles and the outcome becomes known.
Exchanging winning shares for cash after resolution.
The gap between the leader’s price and the price your copy actually filled at.
The distance between the best bid and the best ask in the book.
A taker consumes someone else’s order and pays the fee. A maker posts one and waits.
How many copiers a leader can absorb before they start getting in each other’s way.
Pick a leader and watch it work — on a demo account it costs nothing. Come in whichever way suits you.
