Alex McCullough: Polymarket Data Analytics and Dune Dashboards

Quick Overview

Data analyst Alex McCullough creates Dune Analytics dashboards that provide crypto traders with actionable Polymarket insights by analyzing on-chain data to reveal the historical accuracy of prediction markets and uncover pricing flaws like Acquiescence Bias and Degen Theory.

About alexmccullough

What is alexmccullough?

Alex McCullough is a dedicated data analyst, blockchain enthusiast, and a self-described "humble degen." He is best known for creating deep-dive Dune Analytics dashboards that explore Polymarket insights and prediction market trends. Outside of his data research, he brings his expertise to the crypto space by working in customer success at Awaken Tax.

His analytical work focuses on testing the historical accuracy and hidden biases of decentralized betting platforms. By cracking the code on expected versus actual market outcomes, Alex provides a transparent look into how web3 prediction markets truly operate.

Who is it for?

These data insights are built for crypto traders, sports bettors, and everyday prediction market participants who want a competitive edge. If you buy shares on platforms like Polymarket, this information helps you stop guessing and start understanding the actual math behind the odds.

It is also highly valuable for data researchers, market makers, and web3 enthusiasts studying crowd psychology. Whether you want to avoid overpriced long-shot bets or understand how liquidity affects a market, these reports give you the exact numbers you need to navigate the crypto betting space smartly.

How it works?

The research works by extracting on-chain data and comparing the predicted platform odds to the actual real-world results. Alex groups these markets into probability buckets to measure the exact historical accuracy of Polymarket at different time intervals, such as four hours or twelve hours before an event officially resolves.

His dashboards analyze key market behaviors to expose common pricing flaws. For example, he tracks Acquiescence Bias, where unsure users mentally default to betting "yes," and Degen Theory, where risky long-shot bets are heavily overpriced by bettors chasing big returns. By filtering out extreme outliers and watching the timeline of head-to-head sports markets, the data reveals exactly when the crowd gets it right and when the market is fully biased.

Key Features

Historical Accuracy Tracking

Analyzes prediction calibration by comparing implied market probabilities against actual real-world outcomes across various timeframes.

Probability Bucket Analysis

Groups markets into 5% price increments to assess how well individual odds align with the frequency of 'Yes' resolutions.

Bias Identification

Detects systemic overestimation in markets, tracking phenomena like acquiescence bias, herd mentality, and the mispricing of risk.

Sports Market Calibration

Evaluates head-to-head sporting event accuracy, revealing how prediction precision improves significantly once games are in progress.

Time-Series Performance

Monitors changes in market accuracy from one month before resolution down to the final four hours of trading.

Outlier Refinement

Features the ability to exclude extreme long-shot markets to provide a more genuine representation of core prediction performance.

Frequently Asked Questions

Polymarket predictions are accurate the vast majority of the time, with historical accuracy reaching approximately 92% in the final four hours before a market closes.

Only about 28% of markets resolve to Yes because the platform has an abundance of long-odds markets where many candidates or outcomes are unlikely to occur.

Yes, there is a demonstrated overestimation bias where almost every probability bucket resolves to Yes at a lower rate than expected, suggesting most markets are slightly overpriced.

Potential causes include acquiescence bias where people prefer answering Yes, herd mentality, low liquidity in older markets, and a preference for gambling on long odds for higher returns.

Accuracy generally improves as the resolution time approaches, showing a significant jump in the final 4 hours. In sports markets, accuracy increases even more pronouncedly once the game has started.

Head-to-head sports markets provide a more genuine representation of performance due to a normal distribution of predictions, unlike political or general markets which are often skewed by extremely long odds.

Yes, excluding markets priced above $0.90 or below $0.10 aligns Polymarket's accuracy more closely with traditional head-to-head sports markets.

Data shows noticeable accuracy spikes every 30 minutes after a game begins, which may be linked to market makers' limit order strategies or specific timing of high-volume trades.

alexmccullough