Project 4β Impact: How Bellingcat Tracked $14 Billion in World Cup Bets Across Prediction Markets

Narmin Mammadova

2026-09-07

2 min read

AI Summary:

Supported by Project 4β, Bellingcat analyzed over $14 billion in prediction market bets during the 2026 FIFA World Cup. Using Oxylabs' residential proxies to collect public trade data from Polymarket and Kalshi in a single day, the investigation found that 86% of all winnings were concentrated in the top 1% of accounts.

Overview

Our long-standing Project 4β partner Bellingcat analyzed money flow through prediction markets during the 2026 FIFA World Cup. Their investigation tracked over $14 billion in wagers across nearly 60,000 outcomes on Polymarket and Kalshi, utilizing publicly available data to identify market activity and winners.

The challenge

Prediction markets operate like stock exchanges, making the massive volume of 104 matches difficult to analyze. Comparing platforms is also complex: Polymarket reports actual traded volume, while Kalshi reports notional volume (counting every contract at $1 regardless of price). This requires reconstructing data from the ground up to achieve a fair comparison.

The only feasible way to scrape the complete historical trade data for nearly 40.000 tradeable prediction market outcomes we needed for our analyses was with heavy parallelization and that could only be achieved effectively with fast and reliable proxies. Using Oxylabs turned a multi-week scraping task into a single-day task.

Miguel Ramalho, Investigative Technologist at Bellingcat

The web intelligence behind the investigation

Bellingcat’s Miguel Ramalho and his team conducted this analysis by collecting public data across 104 matches and over 21,000 individual markets. Supported by Project 4β, the team used residential proxies to ensure reliable data collection. They built a heuristic to standardize volume data by calculating the real volume based on daily average prices, enabling an accurate comparison between Polymarket and Kalshi.

Key findings

Bellingcat's analysis revealed both the staggering scale of the betting and a striking concentration of the winnings among a very small group of accounts.

The underlying dataset combines public transaction histories from tens of thousands of markets to give a complete, side-by-side view of tournament betting behavior.

Key investigation metrics

Metric	Value
Total Wagered	$14 Billion+
Total Outcomes Traded	Almost 60,000
Polymarket Volume (Total)	$10 Billion
Kalshi Volume (Total)	$4.3 Billion+
Spain vs Argentina Final (Polymarket)	$212 Million
Winners Concentration (Top 1%)	86% of all winnings
Median Winner Profit	$21
Median Loser Loss	$32
Biggest Winner Profit	$13 Million+
Biggest Loser Loss	$11.6 Million

The dataset brings Polymarket and Kalshi together on equal footing by recalculating contract values against daily average prices across all 104 matches.

Why this matters

The prediction market industry faces growing criticism regarding insider trading, market manipulation, and unregulated gambling. Bellingcat’s findings quantify these risks, showing that winnings are heavily concentrated among a tiny fraction of participants while most users lose money.

This mirrors recent reporting on algorithmic traders dominating these platforms. As the lines between financial trading and betting blur, such open-source scrutiny is essential for transparency. Through Project 4β, Oxylabs provides the web intelligence tools necessary for these investigations. To learn more, contact the team at 4beta@oxylabs.io.

Forget about complex web scraping processes

Choose Oxylabs' advanced web intelligence collection solutions to gather real-time public data hassle-free.

About the author

Narmin Mammadova

PR Content Manager

Narmin is the PR Content Manager for Project 4β at Oxylabs. She enjoys the challenge of getting people to care, and pro bono work gives her good stories to tell. In her spare time, she travels whenever possible or indulges her love of poetry and reciting.

Learn more about the author Narmin Mammadova Learn more about the author Narmin Mammadova

All information on Oxylabs Blog is provided on an "as is" basis and for informational purposes only. We make no representation and disclaim all liability with respect to your use of any information contained on Oxylabs Blog or any third-party websites that may be linked therein. Before engaging in scraping activities of any kind you should consult your legal advisors and carefully read the particular website's terms of service or receive a scraping license.