To auto copy Polymarket leaders end-to-end you connect a non-custodial wallet, choose a plan ($299 Pro or $499 Elite) on BNB Chain or Ethereum USDC, set a per-trade USDC range (commonly $25–$200 for new users), pick three to seven wallets from the leaderboard, configure a category whitelist and a daily loss cap, and start the engine. The whole flow takes about twelve minutes the first time. This walkthrough goes step by step with concrete dollar examples, the gotchas we see most often, and the settings that matter most for your equity curve.
Before you start: what you need
You need three things in hand before you open the dashboard:
- A non-custodial EVM wallet (MetaMask, Rabby, or a hardware wallet via WalletConnect). Poly Syncer never custodies funds; every signature comes from your wallet.
- USDC on either BNB Chain or Ethereum mainnet. The amount is up to you, but the engine's risk gates expect at least $250 to be useful, and the math behind subscription break-even (covered in our cost analysis) starts working in your favour above roughly $2,000 of working capital.
- A clear answer to "what am I trying to do?" — the conservative-core, balanced, or specialist filter from the leaderboard guide. The setup flow is the same for all three; the difference is which wallets you pick.
Step 1: Connect your wallet
Open the Poly Syncer dashboard and click Connect wallet. The flow asks for two signatures: one to prove ownership of the address, and one to delegate a session key to the engine. The session key is scoped — it can sign EIP-712 orders against the Polymarket exchange contract from your wallet, and it cannot move USDC anywhere except into orders on that contract. The full delegation surface is described in the architecture post.
If you are using a hardware wallet, the second signature will prompt for confirmation on the device. The session key has a 30-day rotation by default; you can shorten or lengthen it under /dashboard/settings.
Common gotcha: wrong chain
If your wallet is on Ethereum mainnet but you intended BNB Chain, the dashboard will prompt you to switch. The choice is permanent for that wallet's session and is tied to where your USDC actually lives. Switching later requires bridging, which we cover in the USDC payment network guide.
Step 2: Choose a plan
The plan picker shows three tiers. Free is leaderboard-only, no automation. Pro is $299 and includes the full filter set, saved presets, and the ~1.5 s p99 execution lane. Elite is $499 and adds the 600 ms p99 lane plus the realtime watchlist push and per-wallet alerts.
| Feature | Free | Pro ($299) | Elite ($499) |
|---|---|---|---|
| Browse leaderboard | Yes | Yes | Yes |
| Auto-copy execution | No | Yes | Yes |
| Execution lane p99 | — | ~1.5 s | ~600 ms |
| Number of followed wallets | — | 10 | 50 |
| Saved filter presets | — | 5 | Unlimited |
| Realtime watchlist push | No | No | Yes |
| API access | Read-only | Read + webhook | Full |
For most new users Pro is the right starting point. The latency difference matters in volatile categories (politics during a debate, sports during a live game) but rarely makes or breaks the strategy at small position sizes. Upgrade is one click later.
Step 3: Set your USDC range
The USDC range is the (min, max) per-trade size the engine will sign. The engine sizes pro-rata to the leader's bankroll, then clips into your range. A few concrete examples to make this concrete:
| User bankroll | Suggested min | Suggested max | Reasoning |
|---|---|---|---|
| $500 | $10 | $40 | Min preserves trade count; max keeps any single bet ≤ 8% of bankroll |
| $2,000 | $25 | $150 | Sweet spot for first-time users |
| $10,000 | $50 | $600 | Max around 6% of bankroll; min avoids dust trades |
| $50,000 | $100 | $2,500 | Mind slippage on the upper end — some markets cannot absorb a $2,500 taker without moving |
The single most common error here is setting the max too high. If your max is $1,500 and a leader puts 4% of their (much larger) bankroll into a single trade, the engine will clip to $1,500 and you will have far more concentration than you intended. Setting max around 5–7% of your effective bankroll is the safe rule.
Step 4: Pick wallets to follow
This is where the leaderboard browsing pays off. Open /leaderboard in another tab and apply a filter recipe.
For a balanced first set, pick 3–5 wallets meeting:
- 30-day Sharpe ≥ 1.6 and 90-day Sharpe ≥ 1.4
- Trades ≥ 100
- Max drawdown ≤ 15%
- HHI ≤ 0.3 (so no single market dominates)
- Last active ≤ 72 hours
- Across the set, at least 2 distinct dominant categories
The category-diversity rule is the one most new users skip. A follow-set of three "elite" wallets that are all 90% politics is not a diversified follow-set; it is a politics book with three execution paths. The evaluation checklist details how to vet each candidate.
Worked example: a $5,000 follow-set
Suppose you have $5,000 and pick four wallets: A (politics specialist), B (sports specialist), C (crypto/macro generalist), D (earnings specialist). You set USDC range $30–$300. Per-leader exposure cap is at the default 20% of bankroll = $1,000 maximum simultaneous open USDC mirroring any one wallet. With four wallets and the engine sizing pro-rata, on a typical day you might see eight to twenty fills total spread across the four leaders, with peak open exposure roughly $1,800–$3,000.
Step 5: Configure the category whitelist
The category whitelist is a coarse-grained filter that says "do not copy this leader's trades into category X." This is useful for two reasons:
- You may simply not want exposure to a category (e.g., politics if you have a mandate against it).
- A leader might be elite in their primary category and merely average outside it — you only want their best work.
The default is "all 25 categories enabled." A common refinement: enable only the categories where the leader's edge-adjusted hit rate is > +0.04 in the rolling window. The leaderboard shows per-category breakdowns on the wallet profile page.
Step 6: Set risk gates
Risk gates are the safety nets. The defaults are sane; tightening them is conservative, loosening them is aggressive. Concrete settings:
| Gate | Default | Conservative | Aggressive |
|---|---|---|---|
| Daily loss cap (% of bankroll) | 5% | 2% | 10% |
| Open position cap (count) | 15 | 8 | 40 |
| Per-leader exposure cap | 20% | 12% | 30% |
| Resolution proximity skip | 10 min | 30 min | 5 min |
| Min sample for leader | 30 trades | 100 trades | 30 trades |
The daily loss cap deserves special attention. When the engine breaches it, all new fills are paused until the rolling 24-hour window clears below the threshold. This is the single most-important gate for staying in the game over a year; the philosophy behind it is in our risk management piece.
Step 7: Dry-run mode (recommended for first-timers)
Before flipping the engine into live mode, enable dry-run. The engine will run the full pipeline — listener, scoring, gates, sizing, would-be signing — and write a complete simulated trade ledger, without actually signing anything. After 24–72 hours of dry-run you can review what the engine would have done. Two things to check:
- The number of skipped fills due to risk gates (look for "skipped: max position cap" or "skipped: daily loss cap" in the ledger).
- The realised slippage estimate on each would-be fill, which the engine reconstructs from the leader's fill price and the market's price 600 ms later.
If dry-run shows excessive gate skips, your gates are too tight. If it shows zero gate skips ever, they are too loose.
Step 8: Go live
Toggle the engine to live. The first thing you will see is the ledger filling in real time as leader fills come through. Most users see their first mirrored fill within 15–90 minutes; high-frequency leaders are sometimes immediate. The dashboard shows: leader fill time, your fill time, latency delta, fill price delta, and slippage in basis points.
For the first 48 hours we strongly recommend leaving the bankroll small (say, 25% of what you eventually intend to deploy) so you experience the engine's behaviour in your real risk environment without significant exposure. After that, top up to your target bankroll.
Step 9: Tune over the first week
The settings you started with are a starting point, not a finish line. Two adjustments are common after the first week:
- Drop the leader who underperforms. If one of your four wallets has been negative across the week and the underlying composite has slipped, replace them. The leaderboard will have produced new candidates.
- Tighten the USDC max. Most new users discover their max was set too high relative to the actual fill distribution. Halving the max often improves equity-curve smoothness without materially changing total returns.
What "started auto-copy" looks like in numbers
For a $2,500 user on Pro with USDC range $25–$150 and four leaders:
- Typical day: 6–15 mirrored fills
- Median fill latency vs leader: ~1.4 s
- Open position count peak: 4–9
- Open USDC exposure peak: $300–$900
- Subscription cost amortized: $299 / 12 months = ~$25 / month
The break-even math — how much your follow-set needs to earn for the subscription to pay for itself — is in the cost analysis.
Frequently asked questions
Do I need to bridge USDC before connecting?
Only if your USDC sits on a chain we do not support at the engine layer. We support BNB Chain and Ethereum mainnet directly; USDC on other chains needs to be bridged first. Our USDC payment network guide covers the routes.
Can I follow more than the wallet count my plan allows?
You can have any number of wallets watched (the leaderboard tracks them all). The plan limit applies to wallets that the engine is actively mirroring trades from. Pro is 10, Elite is 50.
What happens if a leader makes a trade while I'm offline?
Nothing changes. The engine runs server-side; your client being closed has no effect. You will see the fill in your ledger when you next open the dashboard.
Can I pause the engine without losing my settings?
Yes. The pause toggle in /dashboard/settings stops new fills immediately and preserves all configuration. Open positions continue to resolve normally; no new ones open while paused.
What if I want to manually override a copy decision?
You cannot intercept a single in-flight order — the latency budget does not allow human-in-the-loop — but you can pause the engine, close any open position from the standard Polymarket UI, and resume. The engine will not re-open the position because it scopes copy to new leader fills, not to existing leader positions.
What happens at subscription renewal if I'm in the middle of an open position?
Renewal is independent of position state. If renewal fails (card expired etc.), the engine pauses for new fills but does not interfere with open positions. We surface a banner well in advance; the billing flow is at /dashboard/billing.
Where to go from here
If you have completed setup, the next reading is the risk-management framework and the patterns analysis — the two together explain why your follow-set behaves the way it does. The strategies page is a more advanced reference. For deeper engineering context, the architecture post covers what is happening under the hood every time the engine fires.