The sensor model
A prediction market price is not a bet. It is a sensor reading. That distinction is the foundation of the S10 series, and it is the reason the Kingdom of Truth treats Polymarket data as first-class intelligence rather than gambling noise. When a Polymarket contract trading at $0.45 moves to $0.72 three days before a geopolitical event resolves, that 27-cent move is not a lucky bet — it is the market’s information sensor detecting a signal that traditional reporting has not yet published. The price is the signal. The question is how to read it.
Traditional forecasting relies on polls, expert panels, and analyst reports. These are slow sensors: they sample periodically, process through editorial filters, and publish with delays measured in days or weeks. A prediction market operates on a fundamentally different principle. Every trade is a micro-measurement. When a participant commits real capital to a YES contract at $0.52, they are expressing a probability estimate backed by skin in the game. The market aggregates thousands of these micro-measurements into a single continuously updated price.
This is the sensor metaphor made concrete. A thermometer does not predict temperature — it measures the current state of a thermal system. A prediction market does not predict the future in the mystical sense — it measures the current state of collective belief about an event, weighted by the capital participants are willing to risk. The difference between a poll and a prediction market is the difference between asking someone what they think and watching what they do with their money.
The scale of these sensors has become significant. Monthly trading volume across prediction markets grew from $1.2 billion in early 2025 to over $20 billion by January 2026, with more than 800,000 unique wallets participating each month. At that volume, the market is not a niche curiosity — it is a distributed sensing infrastructure with liquidity deep enough to produce meaningful price signals across geopolitics, macroeconomics, and political outcomes.
What moves before the news
The most compelling evidence that prediction markets function as early-warning sensors comes from research measuring lead times. A 2025 study introducing the Decentralized Prediction Market Voter Framework found that Polymarket price trends preceded polling shifts by up to 14 days in contested swing states during the 2024 U.S. presidential election. The researchers used cross-correlation analysis and dynamic time warping to establish that the lead-lag relationship was statistically significant and non-spurious.
This is not hindsight bias. The finding is structural: prediction markets aggregate private information from participants who may have direct knowledge, local observation, or analytical edges that have not yet reached public discourse. When a defense contractor employee notices increased procurement activity, when a logistics manager sees unusual shipping patterns, when a political operative observes grassroots shifts — these signals enter the market through trades before they appear in headlines. The price absorbs the information. The headline follows.
A 2025 U.S. Army War College paper explicitly frames prediction markets as intelligence sensors, arguing that contract trading data can serve as a new source of information for analysts assessing national security threats. The paper notes that Polymarket’s self-correcting nature — where mispriced contracts attract trades that move the price toward true probability — creates a continuous calibration loop that traditional intelligence sources lack.
The gradual incorporation problem
The sensor model has a limitation: prediction markets are fast sensors, but they are not instantaneous ones. A June 2026 study using real-time NBA event contracts on Kalshi found that when public information arrives rapidly, prices respond directionally but do not incorporate the full magnitude of the signal on impact. The one-minute benchmark probability change was associated with only a 0.64-for-one contemporaneous price change — meaning the market moved in the right direction but underreacted by roughly 36%.
This underreaction is predictable. It correlates with liquidity: in liquid markets, salient public signals are incorporated relatively quickly. In thin markets, the same signals generate substantially greater underreaction. The missing adjustment predicts subsequent price drift over the following minutes. For the sensor model, this means prediction markets are most reliable as early-warning systems when liquidity is deep enough to support rapid price discovery. A market with $33 million in volume on a single geopolitical contract has the depth to function as a genuine sensor. A market with $50,000 in volume is a noisy thermometer.
Information leakage and the forensics layer
The sensor metaphor cuts both ways. If prices absorb information before news, then informed traders can profit from that information advantage. The ForesightFlow framework, published in May 2026, introduces an Information Leakage Score that quantifies how much of a market’s terminal price move was priced in before the corresponding public news event. The framework analyzed a corpus of 911,237 Polymarket markets and documented hundreds of millions of dollars in anomalous profits.
This is the dark side of the sensor: the same information channel that makes prediction markets useful as early-warning systems also creates opportunities for informed trading on material non-public information. The structural finding is that documented insider-trading cases on Polymarket are systematically deadline-resolved contracts — “Will event XX occur by date YY?” — where the trader had advance knowledge of the timeline. Both Polymarket and Kalshi announced insider-trading restrictions in March 2026, partly enabled by the blockchain transparency that makes such forensics possible.
What this means for the S10 series
The S10 series is about market sensing and research — the tools and methods that convert raw market data into actionable intelligence. Prediction markets as sensors are the first building block because they establish the core principle: price is information, not speculation. When you scan Polymarket through the polymarket skill, you are reading a distributed sensor network. When you ground your analysis in academic research through grounded-citations, you are calibrating the sensor against peer-reviewed evidence. When you build a knowledge base through llm-wiki, you are creating the persistent memory that lets you track how sensor readings evolve over time.
The Kingdom of Truth’s architecture treats prediction market data as one input stream among several — alongside blogwatcher RSS feeds, arxiv paper releases, and internal research. No single sensor is sufficient. But a prediction market that moves 14 days before a polling shift, that incorporates public information at 0.64-for-one in liquid markets, and that processes $20 billion in monthly volume is a sensor worth reading. The price is the signal. The series is about learning to decode it.
