Measuring Trade Direction in a Prediction Market: Settlement Ground Truth and Trading-Cost Measurement on Polymarket

By Philipp D. Dubach

Rating

1776
Battle Count: 50

Relevance

8/10
Highly relevant for quantitative traders operating in prediction markets or using microstructure data. It highlights significant biases in standard trade classification algorithms (like Lee-Ready) when applied to prediction markets with specific settlement mechanisms, directly impacting cost estimation and strategy backtesting.

Implementation Complexity

7/10
Requires access to and parsing of on-chain blockchain data (Polygon) and high-frequency order book feeds. Implementing the matching logic and handling mint/merge settlement types requires careful engineering.

Reproducibility

5/5
The paper provides a comprehensive replication package on Zenodo (DOI: 10.5281/zenodo.23164515) including analysis scripts, result JSONs, and derived inputs. On-chain data is public on Polygon, and feed data sources are identified.

About this paper

Methodology: On-Chain Validation and Trade Classification Benchmarking. Problem types: Classification, Measurement Error Analysis, Market Microstructure Analysis.

The interactive Everscope explorer (charts, battles, favorites) loads below.