Paper-Trade Polymarket Without Risking Capital (and Without Lying to Yourself)

Most "paper trading" tools cheat. They fill your orders at the mid, they ignore fees, they don't model queue position, and they show you the equity curve you wanted to see. When you flip the same strategy to live execution, you find out the simulator was lying.

The PolyQuantLab paper trader runs on live Polymarket orderbook snapshots, fills at the walk-the-book price the live engine would have paid, and subtracts the 2% Polymarket taker fee plus modelled slippage. The equity curve you see in your dashboard is the equity curve a live bot running the same spec would have produced — with one honest caveat we'll cover at the end.

How it works under the hood

1. You compose a strategy in the Strategy Builder — condition tree, position sizing, stop/take rules.
2. Click "Run as paper trade". The spec is shipped to the paper-trader worker on the VPS.
3. The worker subscribes to the same Redis snapshot stream the backtest engine consumes. Every new snapshot per matching market is evaluated against your entry/exit conditions.
4. When a condition fires, the worker simulates a walk-book order at the displayed ask depth, deducts Polymarket's taker fee, and persists the position in Postgres.
5. At resolution (or stop/take), the position closes and the equity curve updates. Same logic the live system would use.

Paper vs backtest, side by side

When you launch a paper run from the Builder, we simultaneously kick off a backtest of the same spec on the most recent N markets. The paper detail page shows both equity curves overlaid — dashed (backtest) over solid (paper). The tracking number ("Paper is +12% vs backtest baseline" or "tracking 8% below") flags when reality drifts from simulation.

That tracking number is the entire reason paper trading exists. If backtest says 65% win rate and paper says 41% after 200 trades, your model is overfit, your fee assumptions were wrong, or the regime changed. You find out before deploying capital, not after.

What paper trading WON'T catch

One honest caveat: paper trading is a quote-followingsimulator. The walk-book fill assumes the depth that was displayed at the moment the signal fired was actually available, in full, for your size. In practice:

• Adverse selection. If your model is right, the market-maker quoting the ask probably has incoming cancellations queued behind your order. Real fills may be partial.
• Network jitter. Polymarket's 8 snapshots per second per market is a noisy upper bound; production latency includes your network, the WS handler, the order placement, the order confirmation. Paper assumes instant.
• Maker queue priority. If your strategy uses limit orders, paper doesn't know your queue position. We model an upper bound; live could be worse.

In our experience the gap is typically 5–15% of backtest PnL for taker-only strategies and materially larger for maker-heavy ones. We log this every time we move a strategy from paper to live so users see the gap empirically rather than trust an opinion.

Tier limits

• Free: 1 active paper strategy at a time.
• Pro ($9/mo): 3 active.
• Plus ($19/mo): 10 active.
• Premium ($39/mo): 100 active — enough to grid-search a strategy family in parallel.

You can pause and resume strategies freely from the dashboard; pausing doesn't count against the active cap.