FForebenchJoin early access →
BUILT FOR POLYMARKET RESEARCH

Polymarket historical data & backtesting.

Explore historical markets, test strategies, and understand how they would have performed under realistic execution conditions.

Historical markets·Realistic execution·Research-ready results
EARLY ACCESS01

Tell us what you need.

Join the private beta and help shape the first datasets and backtests.

No spam. We’ll only email you about Forebench research and early access.
FOREBENCH / PHASE 01

One clean workflow from historical market discovery to a defensible backtest.

POLYMARKET FIRST
WHAT YOU GET

Stop rebuilding the data layer.
Start testing the idea.

01

Historical data

Find resolved markets and work with standardized price, trade, order-book, liquidity, and settlement data.

MARKETS · PRICES · TRADES · BOOKS
02

Strategy backtesting

Test late-market entries, probability thresholds, and hold-to-resolution strategies with explicit execution assumptions.

FEES · SLIPPAGE · LATENCY · FILLS
03
+18.4%

Research-ready results

Inspect return, drawdown, capital usage, reversal markets, and every simulated trade behind the result.

PNL · DRAWDOWN · POSITIONS · EXPORTS
RESEARCH PREVIEW / 001

Does buying “near-certain” markets actually work?

Our first public study will test entries above 95¢ within 24 hours of resolution—and show what happens after fees, spread, slippage, reversals, and capital lock-up.

  1. 01How often does the favorite reverse?
  2. 02Does the edge survive realistic fills?
  3. 03What is the return on locked capital?
Get the report when it ships →
WHY FOREBENCH
“We are building the research infrastructure we wished existed when studying prediction-market strategies.”
Join the early access →
FREQUENTLY ASKED QUESTIONS

Forebench, briefly.

What is Forebench?

Forebench is an early-access research platform for exploring historical Polymarket data and backtesting prediction-market strategies.

What historical Polymarket data will Forebench provide?

We are building standardized access to historical markets, prices, trades, order books, liquidity, and settlement data.

How will backtests model realistic execution?

Each study is designed to state its assumptions for fees, spread, slippage, latency, fills, and capital lock-up instead of assuming every observed price was tradable.

Who is Forebench for?

Prediction-market researchers, quantitative traders, developers, and teams building or evaluating AI prediction agents.