Polymarket paper trading simulator

Polymarket Paper Trading for Strategies

Forward-test a Polymarket rule before real funds or wallet signing enter the loop. DepthFeed gives the strategy virtual cash, prices buys and sells from the displayed live CLOB book, settles positions from the market outcome, and keeps the equity curve, open positions and trade log together.

Start paper trading free See the paper-trading workflow

Polymarket paper trading simulates a strategy on current Polymarket markets without placing an on-chain order. In DepthFeed, a lab rule or webhook signal opens a virtual position at the displayed live ask or bid; the position is marked over time and settles automatically to the recorded $0 or $1 outcome.

What the Polymarket paper-trading engine actually supports

DepthFeed is built for repeatable strategy experiments rather than isolated practice clicks. These are current product parameters verified against the Paper Trading implementation.

crypto assets

BTC, ETH, SOL, XRP, DOGE, BNB and HYPE are selectable in lab-rule mode.

market windows

Run rules on supported 5-minute, 15-minute, 1-hour, 4-hour and 24-hour markets.

entry-rule families

Time-before-close, level-cross and dip-then-recovery rules run without your own server.

120/min

webhook limit

Each webhook strategy accepts bounded, idempotent signals from a custom bot.

Product configuration and worker limits verified in the DepthFeed Paper Trading implementation on July 26, 2026. Plan limits and supported markets can change.

Polymarket paper trading at a glance

Fills: Displayed live book — buys at ask, sells at bid
Settlement: Automatic at resolution ($0 / $1)
Signal webhook: POST /paper/hook/{token} · 120/min
Strategies per plan: 1 free · 5 Quant · 15 Research · 40 Desk
Starting balance: $100–$1,000,000 virtual (default $10,000)
Mark cadence: ~90 seconds
Lab assets: BTC · ETH · SOL · XRP · DOGE · BNB · HYPE
Lab windows: 5m · 15m · 1h · 4h · 24h
Entry rules: Time range · level cross · dip recovery
Sizing: Fixed dollars · fixed fraction · Kelly-style
Risk controls: Take profit · stop loss · max open · trailing drawdown
Research record: Equity curve · positions · trade ledger · CSV export

How the Polymarket simulator works

Start with a falsifiable Polymarket strategy

A useful paper test begins with a rule that can be evaluated without hindsight. Define the venue, asset, market window, outcome side, entry condition, position size and exit logic before the sample begins. For example: buy the Up outcome on Polymarket BTC 15-minute markets only after its ask crosses 0.70, cap one position at 5% of current equity, and stop the strategy if trailing drawdown reaches 20%. Writing the rule first prevents a winning trade from quietly changing the hypothesis.

Choose no-code lab rules or a webhook bot

Lab mode runs three rule families against eligible Polymarket crypto markets: enter a configured number of seconds before close when price is inside a range; enter when price crosses above or below a level; or wait for a price to dip below one threshold and recover to another. Webhook mode is for logic that lives in your notebook, model service, cron job or trading bot. It accepts the action, venue, Polymarket CLOB token ID, dollar size and an optional idempotent client ID while refusing a caller-supplied fill price.

Price virtual fills from the displayed CLOB

When a buy signal is accepted, DepthFeed reads the current displayed Polymarket order book and uses the relevant ask; a sell or close uses the displayed bid. The signal cannot improve its own execution price. This includes the spread that existed when the strategy acted and is more disciplined than marking every trade at a midpoint or last trade. It is still a single-price simulation: the paper fill does not walk multiple depth levels or remove resting size from the real book.

Keep backtest and forward-test assumptions aligned

A Backtest Lab result can be handed to Paper Trading with its venue, coin, market window, side, rule, stake, sizing method, take-profit and stop-loss prefilled for review. Nothing starts until you confirm the settings. The important discipline is to preserve the original hypothesis while moving from historical replay into an unseen forward sample. If the paper configuration differs materially from the backtest, record the change as a new version rather than blending both results.

Test position sizing and risk controls, not only direction

The same entry signal can produce very different portfolio paths under fixed-dollar, fixed-fractional and Kelly-style sizing. DepthFeed can cap the percentage of bankroll committed per trade, limit concurrent positions, take profit, stop loss and enforce a trailing drawdown circuit breaker. The circuit breaker compares current equity with the strategy's high-water mark; when the configured loss is breached, the paper account is flattened and halted. This makes survival, concentration and drawdown part of the test instead of an afterthought.

Read the equity curve together with the trade ledger

Each strategy maintains virtual cash, current equity, realized and unrealized P&L, open positions, marks and a paginated trade history. A green ending balance is not enough. Review trade count, average entry spread, concentration by market window, maximum drawdown, losing streaks, time in position and whether performance depends on one resolved market. The dashboard retains the newest 50,000 trades for up to 180 days, and the full trade log can be exported to CSV for a durable research record.

Let the forward sample settle before judging it

Open positions are marked on the paper worker cadence and are settled from the recorded Polymarket outcome. A strategy that looks profitable before resolution can still be carrying concentrated binary risk. Compare results only after enough independent markets have completed, and keep unresolved positions separate from realized outcomes. For short-window crypto strategies, collect multiple market regimes rather than stopping after one favorable day.

Use paper performance as a gate, not a promise

A strong paper record supports the narrower claim that the signal behaved acceptably under the simulator's stated assumptions. It does not prove that the same order will receive the same fill live. Before risking funds, replay the full ladder at the intended size, estimate fees and slippage, check venue access and operational requirements, and begin with a deliberately smaller live canary. DepthFeed never converts a paper strategy into an authenticated Polymarket order automatically.

Worked example: forward-test a Polymarket momentum rule

This example shows the research decisions a paper record can support. The numbers are illustrative settings, not a performance claim or a recommendation to trade.

1Write the hypothesis before opening the dashboard
Suppose the hypothesis is that late momentum persists in BTC 15-minute Up markets: after the displayed ask crosses 0.70, the outcome is more likely to finish at $1 than the price implies. Define the eligible window, the direction of the cross, how long the rule may hold, and the minimum number of resolved markets required before inspection. Do not add a dip filter after seeing early losses unless you create a separately named strategy version.
2Match the historical test
Backtest the same venue, token side, market window and level-cross rule using historical order-book observations rather than last-trade prices. Record the spread and any size assumptions. Use Deploy to paper only as a prefill: review every field, especially the outcome side and entry level, before creating the strategy. The backtest answers how the rule would have behaved on past samples; the paper strategy begins a new, unseen sample.
3Choose bankroll rules before the first signal
Give the test a fixed virtual starting balance, use a five-percent fixed-fraction stake capped at a chosen maximum per trade, allow only a small number of simultaneous positions, and set a trailing drawdown threshold. Fractional sizing changes stake with live bankroll, so a losing sequence naturally reduces exposure. A fixed-dollar version can be run as a separate strategy when the research question is specifically about constant notional.
4Inspect the modeled entry
When the ask crosses the configured level on an eligible market, the worker records the virtual buy from the displayed ask available to the simulator. Check the trade timestamp, token ID, price, shares and notional. If the observed spread was wide, that cost belongs in the record. Do not replace the recorded price with a cleaner midpoint because the strategy would not have been able to buy at the midpoint without resting an order.
5Wait for complete outcomes and review distributions
After markets resolve, compare realized P&L with the backtest and examine more than win rate. Review calibration by entry-price band, maximum drawdown, average gain versus average loss, losing streaks, concentration by time of day and the share of profit contributed by the best market. Export the CSV and preserve the strategy configuration beside it so the result can be reproduced.
6Apply a precommitted decision gate
Continue, revise or reject the rule using thresholds chosen before the test. A reasonable gate might require a minimum number of independent resolved markets, positive expectancy after a conservative fee/slippage allowance, drawdown below the risk budget and no dependence on one outlier. If it passes, run a longer paper window or a small live canary; if it fails, diagnose the hypothesis rather than hiding the losing strategy.

Where Polymarket paper trading differs from live execution

The simulator is intentionally explicit about what it measures and what remains unknown. These gaps should shape the decision gate for any live deployment.

Displayed price is not guaranteed executable size

The simulator prices a virtual trade from the displayed top of book, but it does not reserve or consume that liquidity. A live order larger than the resting size may walk the ladder, partially fill or cancel. Compare intended notional with full depth before treating the paper price as a live estimate.

Queue priority and cancellation races are absent

A resting live limit order competes for time priority and can be exposed when other traders cancel or reprice. Paper Trading models an immediate virtual action from observed quotes; it does not model maker queue position or the probability that a displayed quote disappears before an order reaches the matching engine.

The worker is a research tracker, not a low-latency executor

Lab rules and position marks run on an approximately 90-second cadence. That is appropriate for forward-testing slower strategy hypotheses, but it cannot validate sub-second reactions, latency arbitrage or exact intraminute path dependence. Those require event-level replay and purpose-built execution measurements.

Fees, market impact and operational failures need a separate allowance

Paper P&L does not guarantee the net P&L of an authenticated Polymarket order. A live-readiness calculation should include applicable fees, spread, size-dependent slippage, rejected orders, connectivity failures, wallet signing and asynchronous on-chain settlement.

Resolution and market selection can create hidden dependence

Binary outcomes create lumpy P&L, and related markets may not be independent. Ten trades on overlapping BTC windows can represent far fewer than ten independent bets. Review the market question, token ID, resolution status and overlap before interpreting sample size.

Paper success is not financial advice

DepthFeed provides an independent research simulator and does not recommend a strategy, guarantee profitability or place a Polymarket order. Venue eligibility, legal access, account security and the decision to risk capital remain outside the paper workflow.

Start a paper strategy with no real funds

Explorer includes one concurrent paper strategy at no cost. Create a rule in the dashboard or connect your own bot by webhook; no card or trading-wallet connection is required.