Publications & Working Papers

Research Output

2025–2026 Academic Year
Working Paper Q3 2026

Favourite-Longshot Bias in Prediction Markets

Analysing systematic mispricing of high- and low-probability events across 2,400+ resolved Polymarket markets to identify category-specific trading opportunities.

Shubhaankar Gupta · Oracle Research Upcoming
Market Analysis Q4 2026

Polling Aggregation as a Signal in Electoral Prediction Markets

A framework for integrating polling aggregates into prediction market positioning, backtested against 2022 midterm and 2024 presidential data.

Shubhaankar Gupta · Oracle Research Upcoming
Quantitative Model Q4 2026

Kelly Sizing Under Parameter Uncertainty in Binary Markets

Deriving practical bet-sizing rules that account for estimation error in prediction markets, with simulation and historical backtests.

Shubhaankar Gupta · Oracle Research Upcoming

How We Work

Research Methodology

01

Empirical First

All research begins with data. We pull historical Polymarket data via API, clean and structure it, and let empirical patterns guide our hypotheses.

02

Actionable Output

Research that cannot inform a trading decision is deprioritised. Every paper we produce should have a clear section on trading implications.

03

Academic Rigour

We engage with the academic literature on prediction markets and forecasting (Tetlock, Hanson, Wolfers, Arrow et al.) and hold our own work to the standard of peer-reviewed publication where possible.

Collaborate

Interested in our research?

We welcome collaboration with UChicago faculty, researchers at other institutions, and industry practitioners. Reach out on X or through our sponsor contacts.