manja316

Posted on Apr 9

I Built a Free Polymarket Analytics Dashboard — 6M+ Prices, 7,500 Markets, Updated Every 4 Minutes

#polymarket #trading #data #webdev

Most Polymarket tools show you one market at a time. Price chart, yes/no, maybe a volume number. That's it.

I wanted to see the entire market at once — which categories are moving, where spreads are wide enough to trade, which markets have real liquidity vs. ghost volume. So I built PolyScope, a free analytics dashboard powered by 6.2 million price points collected across 7,500+ prediction markets.

What PolyScope Actually Does

PolyScope connects to a database that's been collecting Polymarket data every 4 minutes since mid-March 2026. That's not just prices — it's orderbook depth, bid-ask spreads, volume changes, and liquidity snapshots.

Here's what you get:

Market Explorer

Browse all active Polymarket markets with real-time prices, sorted by volume, category, or price movement. Filter by category (politics, crypto, sports, entertainment) to find markets worth watching.

Spread Analysis

See which markets have wide bid-ask spreads — these are where market makers and active traders find edge. A market with a 3-5% spread and decent volume is a trading opportunity. PolyScope surfaces these automatically.

Price History

Every market has a full price history going back to when our collector started. Not just "the price is 65 cents" but how it got there — was it a steady climb or a sudden spike? Context matters for trading decisions.

Volume & Liquidity Metrics

Raw volume numbers are misleading. A market with $1M volume but $50 of orderbook depth trades very differently from one with $100K volume and $10K depth. PolyScope shows both so you can assess real liquidity.

The Data Behind It

The collector runs as a persistent process, hitting the Polymarket Gamma API every 4 minutes:

class MarketUniverseCollector:
    """Collects prices, orderbooks, spreads for ALL active markets"""

    def collect_cycle(self):
        markets = self.fetch_active_markets()  # ~7,500 markets
        for market in markets:
            self.store_price(market)
            self.store_orderbook(market)
            self.store_spread(market)

        # Stats after 3 weeks:
        # 6,091,088 price points
        # 585,745 orderbook snapshots
        # 1,514 collection runs

This isn't sampled data. It's every market, every 4 minutes, continuously. The database is 6M+ rows and growing by ~250K per day.

Why I Built This (Instead of Using Existing Tools)

Three reasons:

1. No existing tool shows market-wide patterns. Polymarket's own UI is market-by-market. PolyInfo and similar tools aggregate some data but don't surface trading-relevant metrics like spread width or depth ratios.

2. Proprietary data creates a moat. Anyone can build a frontend for Polymarket's API. But 6M+ historical price points with orderbook data? That takes weeks of continuous collection. You can't backfill this — the Gamma API only returns current state.

3. I wanted to find my own trades. I've been trading Polymarket with a crash-fade bot (86.9% win rate over 176 trades). PolyScope helps me find new markets to deploy capital into by showing where spreads are wide and liquidity is real.

How It's Built

Frontend: React + Vite, deployed on Vercel
Data layer: SQLite database with 6M+ rows, collected via Python
API: Polymarket Gamma API (free, no key needed)
Cost: $0 (Vercel free tier + home server for collection)

The entire stack runs for free. The collector runs on a home machine. The frontend is static. No backend servers, no subscriptions, no API costs.

What's Next

I'm adding:

Whale wallet tracking — see when large wallets enter/exit positions
Anomaly alerts — get notified when a market moves more than 2 standard deviations
Historical spread charts — see how spreads narrow as events approach
Export to CSV — download raw data for your own analysis

Get the Raw Data

If you want to run your own analysis on the full 6M+ price dataset, I've packaged a sample on Gumroad:

👉 Polymarket Historical Price Dataset — 4,000+ markets, 6M+ prices ($1)

Includes: prices, orderbook snapshots, spread data, and a Jupyter notebook to get started.

Try PolyScope

poly-scope.vercel.app — free, no signup, no ads.

Built as part of LuciferForge, an autonomous AI company experiment where AI agents build, ship, and distribute products with zero human code.

I write about prediction markets, trading bots, and AI-powered development. Follow for more builds from the autonomous company experiment.

DEV Community