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 17 million+ price points collected across 17,000+ prediction markets.
Update 2026-05-30: dataset has grown from 6M prices / 7,500 markets at first publish to 17M+ prices / 17,000+ markets today (verified 2026-06-05 from source:
api.protodex.io/stats). Numbers below reflect current state.
What PolyScope Actually Does
PolyScope connects to a database that's been collecting Polymarket data every 15 minutes since late 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 on a fixed cycle:
class MarketUniverseCollector:
"""Collects prices, orderbooks, spreads for ALL active markets"""
def collect_cycle(self):
markets = self.fetch_active_markets() # ~17,000 markets
for market in markets:
self.store_price(market)
self.store_orderbook(market)
self.store_spread(market)
# Stats as of 2026-05-30:
# 17,413,418 price points
# 17,358 markets tracked
# ~5.8 GB SQLite database
This isn't sampled data. It's every market, every cycle, continuously. The database has crossed 17M rows and is still growing every 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 17M+ historical price points with orderbook data? That takes months of continuous collection. You can't backfill this — the Gamma API only returns current state.
3. I wanted to find my own trades. I trade Polymarket with a crash-fade bot. PolyScope helps me find new markets to deploy capital into by showing where spreads are wide and liquidity is real. Verified P&L over claimed P&L — see the next section.
Don't Trust Win-Rate Claims (Including Mine)
Every Polymarket bot screenshot you see online claims a high win rate. Most are computed wrong — they ignore fees, redemption haircuts, and unresolved positions.
I built pnl-truthteller — an open-source tool that pulls your actual on-chain trades from Polygon and reconciles them against your DB-claimed P&L. If your bot says +$50 and the chain says +$12, you find out before you scale.
pip install pnl-truthteller then pnl-truthteller --wallet 0x.... MIT licensed. Run it on my wallet, run it on yours.
How It's Built
- Frontend: React + Vite, deployed on Vercel
- Data layer: SQLite database with 17M+ 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
- 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 17M+ price dataset, I've packaged it on Gumroad:
👉 Polymarket Historical Price Dataset — 17,000+ markets, 17M+ prices ($9)
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.
Top comments (0)