See the narrative behind the probability.
Prediction-market prices tell you what the crowd expects, but not why. NIL shows how the story around a market is changing, right next to its price, across politics, business, and geopolitics.
Illustration only, not real market data.
Three market areas, each backed by research
Politics and elections
2,434 statements coded from 32 years of U.S. presidential debates, with near-perfect agreement between independent coders (κ = 0.93). Peer-reviewed in 2025, this dataset is NIL's accuracy benchmark.
Business and corporate
Doctoral research found that specific narrative types in executive communication were statistically significant predictors of changes in company profit and revenue.
Geopolitics
In live forecasting through the RAND Forecasting Initiative, model-based assessments sometimes differed from the crowd forecast. NIL tests whether those differences carry information.
How NIL works
Gather
Collects dated news, speeches, transcripts, and social media around each market.
Read
A proprietary, research-validated framework reads how the story is told, not just how often a topic comes up.
Compare
Plots a time-stamped narrative signal beside the price and sends an alert when the two diverge.
Record
Keeps a history of how earlier narrative shifts related to later prices and final outcomes.
Built on years of research
- 2020
Research presented at RAND on whether search behavior, social signals, and other human-generated information could spot emerging events earlier than conventional analysis.
- 2024
Doctoral research developed the narrative framework at the core of NIL.
- 2025
A peer-reviewed study in the Journal of Leadership Studies applied the framework to presidential debates.
Where NIL stands
The core method works: AI now does narrative coding that once took months by hand, in minutes. An early prototype runs on Polymarket's public market data, and a builder team is in place.
Next comes a public beta, starting with U.S. political markets and expanding to business and geopolitical markets, with backtest results published openly. NIL offers evidence, not trading advice, and won't claim accuracy until testing shows it.
Get in touch
We welcome conversations with prediction-market platforms, research partners, and organizations interested in narrative intelligence. Email us at info@adviao.com.