How We Detect Whale Activity in Polymarket's Order Book

Prediction markets are fascinating because prices reflect collective belief about future outcomes. But not all participants are equal — when a large player ("whale") moves significant capital into one side of a market, it can reveal information that smaller participants might not have. At PolyEdge AI, we detect these moves in real-time and surface them alongside our AI forecasts.

What Is Polymarket's CLOB?

Polymarket uses a Central Limit Order Book (CLOB) — the same type of order book used by stock exchanges and cryptocurrency markets. Instead of a simple "buy at market price" mechanism, users place limit orders at specific prices, creating a visible book of bids and asks.

This order book is a treasure trove of information. You can see not just the current price, but also how much capital is sitting at each price level, where the depth of the market is, and — crucially — when large orders appear or disappear.

How Whale Detection Works

Our whale detection system continuously monitors Polymarket's order book for unusual activity patterns:

• •Large-block orders: We flag individual orders that exceed a configurable threshold relative to the market's average order size. A $50,000 buy in a market where typical orders are $200-$500 is a strong signal.
• •Rapid accumulation: Sometimes whales break their activity into many smaller orders to avoid detection. We track cumulative buying or selling pressure over sliding time windows to catch these patterns.
• •Order book imbalances: When the depth on one side of the book dramatically outweighs the other (e.g., 10x more bids than asks), it suggests directional conviction by large players.
• •Price impact analysis: We measure how much the price moves after detected whale activity. Sustained price shifts after large buys confirm informed activity rather than market-making.

Why This Matters for Research

Whale activity is not a prediction in itself, but it's a valuable signal that adds context to our AI forecasts. Consider this scenario:

Our 6-model ensemble estimates a market at 72%. The current Polymarket price is 65%. Our AI sees a +7 percentage point analytical gap. Then our whale detector fires: a large player just bought $80,000 of YES shares at 66-68¢. This corroborates our AI's directional view — a sophisticated market participant with capital at risk agrees that YES is underpriced.

Conversely, if our AI says 72% but we see whale selling on the YES side, that's a reason for caution. Maybe the whale knows something our models don't. We flag this divergence explicitly so researchers can weigh both signals.

It's important to emphasize: whale activity is informational context, not a recommendation. Large players can be wrong. They can also be hedging, market-making, or responding to private information that may or may not pan out. We surface this data so you can incorporate it into your own research — we never tell you what to do with it.

How We Surface This Alongside AI Forecasts

On the PolyEdge AI dashboard, whale activity appears as a contextual signal alongside each market analysis. You'll see:

• •A whale activity indicator when significant order book imbalances are detected
• •The direction of whale activity (buying YES or NO)
• •Whether whale direction agrees or disagrees with our AI forecast
• •Approximate magnitude of the activity

Researcher-tier subscribers get access to the full whale activity feed with historical data, while Starter plan users see a summary indicator on their daily analyses.