Guide

How Polymarket Copy Trading Works — 2026 Guide

Everything a new user needs to start mirroring profitable Polymarket wallets — covered end to end, from wallet connection to first fill.

Last reviewed · Eli Marsh, Poly Syncer

Polymarket copy trading is the practice of automatically mirroring the trades of a profitable Polymarket wallet from your own non-custodial wallet. Poly Syncer is a Polymarket bot that lets you auto copy Polymarket leaders: it watches 1,891–2,492 indexed wallets, refreshes their scores every 15 minutes, and fires a matching trade in roughly 600 ms (p99) on a co-located RPC after a leader fills. You keep custody, you set the risk gates, and the engine handles the keystrokes.

What is Polymarket?

Polymarket is the largest on-chain prediction market, with cumulative trading volume above $9 billion across politics, sports, crypto, earnings, and more. Every market is a binary contract that resolves to $1 (YES) or $0 (NO), settled in USDC on an EVM-compatible chain. Order books are public, every fill is on-chain, and the wallet that placed the trade is observable in real time. That public-by-default architecture is what makes copy trading possible at all — the signal is already broadcast; Poly Syncer just acts on it faster than a human can.

What "copy trading" means in prediction markets

If you have used a copy-trading product on a centralized spot-crypto exchange, prediction-market copy trading will feel familiar but is mechanically distinct. On a spot exchange you mirror an open position with continuous price action. On Polymarket each contract has a discrete resolution event — an election result, an earnings beat, a final score — and the market closes against $0 or $1 at that moment.

That changes three things:

The mental model is not "I'm following a guru." It is "the price-discovery mechanism is doing its job, and I am letting it reach my account without lag."

How the Poly Syncer engine mirrors leader trades

The full architecture is documented on the whitepaper and the methodology page, but the short version has four moving parts:

  1. Indexer. A node colocated with a top-tier Polymarket-aware RPC reads every fill from the order-book contract. New trades from indexed wallets are detected within ~200 ms of inclusion.
  2. Filter. Each detected trade is checked against the user's risk gates: position size cap, daily-loss cap, allowed categories, allowed markets, minimum leader Sharpe, etc.
  3. Sizer. The user's allocation rule converts the leader's USDC size into the user's USDC size. Default is proportional-to-bankroll with a hard ceiling.
  4. Executor. A pre-signed, narrowly-scoped EIP-712 authorization lets the engine submit a matching order on the user's behalf. End-to-end p99 latency is roughly 600 ms; median is closer to 320 ms.

Step-by-step: your first 24 hours on Poly Syncer

1. Connect a wallet

Use any EIP-1193 wallet — MetaMask, Rabby, Coinbase Wallet, or a hardware-backed signer. Poly Syncer never holds your keys and never asks for an email or phone number. Your wallet address is your account.

2. Pick a plan and pay

Pricing is flat: Pro at $299/month (250 mirrored wallets, premium RPC) or Elite at $499/month (unlimited wallets, AI alpha discovery, mempool sniping). Payment settles in USDC on either BNB Smart Chain (BSC) or Ethereum mainnet. A free view-only tier lets you browse the leaderboard before paying; you can cancel a paid plan from /dashboard/billing at any time.

3. Browse the leaderboard

Open /leaderboard. Sort by 30-day Sharpe (we recommend ≥ 2.0 as a starting filter), inspect the wallet's full trade history, and check the win-rate-vs-implied-odds chart. We cover the full evaluation checklist in our scoring methodology post.

4. Set your risk gates

From the strategies page, configure:

5. Follow 1–3 leaders and watch the first fills

Don't follow ten wallets on day one. Pick two or three with low return correlation, set a small position-size ceiling ($50–$100 max), and let the system run for at least 72 hours before scaling. The dashboard logs every detection, every filter decision, and every fill with its leader-vs-mirror price slippage.

Risk gates in detail

Defaults matter more than features. Poly Syncer ships with conservative defaults because most blowups in retail trading come from sizing, not stock-picking. The full risk framework is detailed in our risk-management guide; the highlights:

GateDefaultWhy
Max position size$250 USDCCaps tail risk per trade.
Daily loss cap5% of bankrollForced 24h cool-off when hit.
Stop-loss−30% from entryAvoids riding zeros to resolution.
Min leader Sharpe (30d)2.0Filters small-sample lucky wallets.
Category whitelistAll 25 enabledTighten as you learn your edge.

Why automate instead of trading manually?

The honest answer is latency and attention. By the time a top wallet's trade is visible on the public Polymarket front-end, the order book has often moved 1–3 cents. Across a year of mirroring, those 1–3 cents per fill compound into the difference between matching the leader and underperforming them by 6–9% of return. Manual mirroring also requires being awake when the leader trades, which they often do at 03:00 UTC on a non-US calendar day. We compare the costs in detail on Poly Syncer vs Polymarket native and vs Kalshi, and the line-by-line break-even math lives in copy trading vs manual.

Pricing and payment in plain English

Poly Syncer charges in USDC. There are exactly two paid tiers plus a free view-only tier:

You pay in USDC on BNB Smart Chain or Ethereum mainnet. We deliberately do not accept credit cards: it keeps the product non-custodial and the data footprint minimal. Manage your subscription at /dashboard/billing.

What "non-custodial" actually means here

The word "non-custodial" gets used loosely. On Poly Syncer it has a precise technical meaning: at no point in the lifecycle of any trade does Poly Syncer software hold your private key, hold your USDC, or have unconstrained authority over your wallet. The mechanism is a narrow EIP-712 typed-data authorization scoped to (a) the Polymarket order-book contract, (b) a maximum cumulative USDC notional, and (c) an expiry timestamp. The authorization can be revoked from your wallet's standard "approvals" panel at any time, which terminates new mirror activity within one block. Your existing positions are unaffected because they live in your wallet, not ours. There is no email, no "account password," and no off-chain ledger of your funds — the only state Poly Syncer holds about you is the wallet address that owns your subscription.

The practical consequences are worth stating plainly:

This is the structural reason we did not build a credit-card checkout. A custodial billing rail would create exactly the attack surface the rest of the architecture refuses to have. The full posture is documented in the whitepaper.

What the engine looks like in operation

A typical mirror cycle, end-to-end:

  1. t = 0 ms. Leader wallet 0xA31F…9C24 fills a $4,200 YES position on a politics market at $0.61.
  2. t ≈ 180 ms. The Poly Syncer indexer sees the fill on the co-located RPC subscription, matches the wallet to a follower's strategy, runs the risk filter (size cap, daily loss check, category whitelist, market liquidity floor).
  3. t ≈ 240 ms. The sizer converts $4,200 leader notional into a follower notional — suppose $210 (5% of a $4,200 follower bankroll), capped at the user's max of $250.
  4. t ≈ 320 ms. The executor signs an order using the user's pre-authorized EIP-712 envelope and broadcasts via a private mempool route to defend against MEV (covered in MEV protection).
  5. t ≈ 600 ms (p99). The mirror fills at $0.611 — one tick above the leader's price, the typical microstructure cost on Polymarket.

That single-tick slippage is the dominant operational cost of copy trading and is the reason latency matters. A 4-second delay would routinely cost 2–3 ticks instead of one, multiplying your tracking error by 3×. We publish slippage statistics in the changelog.

Common pitfalls in your first month

  1. Following too many wallets. Correlated leaders amplify drawdowns. Start with 2–3.
  2. Skipping the size cap. A leader with a $50,000 bankroll betting 8% has a different risk profile than you copying 8% of yours into a thinner book.
  3. Chasing yesterday's leader. 7-day Sharpe is noise. We surface 30-day and 90-day metrics for a reason.
  4. Ignoring categories you don't understand. If you can't explain what a market resolves on, whitelist it off.

Frequently asked questions

Do I need to keep my computer on?

No. The mirror engine runs server-side on Poly Syncer infrastructure. Your browser can be closed; your wallet only needs to be connected long enough to grant the time-bounded EIP-712 authorization. The engine then signs orders within the limits of that authorization until you revoke it.

Is Poly Syncer custodial?

No. Your USDC stays in your wallet at all times. Poly Syncer holds a narrow, revocable trading authorization — not your keys and not your funds. You can revoke from your wallet's "approvals" panel at any moment.

Can I cancel my subscription?

Yes, any time, from /dashboard/billing. There is no minimum term. The current billing period stays active until its end; we do not pro-rate refunds, but you keep access for what you've paid for.

What happens if the leader I follow has a losing streak?

You lose money proportionally to your sizing rule, capped by your stop-loss and daily-loss limits. The engine will pause your mirror after the daily cap is hit and you'll receive a dashboard alert. We strongly recommend the diversification rules in our risk-management guide.

How fast does the engine fire after a leader fills?

End-to-end p99 latency is roughly 600 ms on the co-located RPC; median is around 320 ms. Detection alone is sub-200 ms. In practical terms: by the time a human refreshes a page and sees the leader's trade, your mirror has already filled.

How copy trading on Polymarket compares to manual trading

The honest accounting compares total cost, not just subscription. For a typical $10,000 bankroll trading 30 positions per month at $200 each, the all-in monthly cost stack looks roughly like this:

The break-even depends on how aggressively you trade and how much your time is worth. The full line-by-line breakdown lives in copy trading vs manual. The short version: at $5,000+ bankroll and 20+ trades a month, automation pays for itself on slippage savings alone before you count the time you reclaim.

What the dashboard actually shows you

One of the questions we get most often from new users is what they will see after connecting. The dashboard surfaces five panels:

  1. Followed wallets — per-wallet ROI, Sharpe, last fill timestamp, current open positions, and a one-click pause toggle.
  2. Live mirror feed — chronological log of every detection, every filter decision (including filtered-out trades and the rule that filtered them), and every fill with its leader-vs-mirror price slippage.
  3. Risk gates — the current values of position cap, daily loss cap, stop-loss, and category whitelist, all editable in place.
  4. Equity curve — your account's USDC balance plotted against followed-wallet aggregate equity, so you can see tracking error and your own drawdown.
  5. Audit trail — signed receipts of every authorization, every revocation, and every executed trade, exportable as CSV for your records or your accountant.

The audit trail is deliberately verbose. Non-custody only works if the user can verify that the executor stayed within the authorization, and the only credible way to demonstrate that is to log everything and let the user inspect it. Every receipt links back to the on-chain transaction so the public ledger is the final word.

Ready to start?

Open the leaderboard, find a wallet whose 30-day Sharpe and trade history you understand, and follow it with a $50 cap for the first week. The free view-only tier lets you study the leaderboard before paying; when you’re ready to mirror trades, head to /dashboard/billing and upgrade to Pro. The glossary is there if any of the terms above need expanding, and the API is open for anyone who wants to build on the same data.