Powering SIG’s Sports Prediction Market Models with SportsDataIO

Case Studies
July 14, 2026

Case Study Overview

Susquehanna International Group (SIG) is a global quantitative trading and technology firm founded in 1987 and headquartered in Bala Cynwyd, Pennsylvania. One of the largest market makers in U.S.-listed equity options, SIG has extended that same market making discipline into prediction markets through its dedicated Susquehanna Predictions desk, one of the first quant trading operations to build a business unit focused specifically on regulated prediction markets, including sports.

As a market maker, SIG provides liquidity and structures tradeable contracts for prediction market exchanges. To do that credibly for sports markets, SIG needed a quantitative foundation built on deep, reliable historical sports data, not just for backtesting, but as a live input to the models that price contracts every day.

The Challenge

SIG's trading desk builds proprietary models to price and trade sports contracts on prediction market exchanges. Those models depend on historical pattern recognition, understanding how teams, players, and markets have behaved over many seasons to generate a statistical edge rather than a guess.

SIG needed a data partner to:

  • Provide Deep Historical Coverage: A single source with enough historical depth spanning 2012 through the current season to support robust, multi-season modeling.
    • Carry That Depth Into Live Play: In-season data delivered through the same schema and definitions as the historical archive.
      • Support Betting-Market Modeling: Historical and current betting line data to model how markets have moved historically against real outcomes.
        • Meet Quant-Grade Standards: A provider built for quantitative trading use cases, not just fan-facing stats.

          The Solution

          SportsDataIO delivered a unified NFL data package built around exactly this continuity requirement.

          • NFL Competition Feeds: Structured league, season, and pre-game data forming the backbone of SIG's historical and in-season models.
            • Event Feeds: Granular statistical and play-by-play game data supporting deep statistical and situational modeling.
              • Sports Betting Data: Historical and current line movement, allowing SIG to model how pricing has shifted in response to real-world outcomes.
                • Projections: Forward-looking statistical projections feeding directly into SIG's pricing and trading models.

                  All four data products were delivered under a consistent integration pattern, documented in SportsDataIO's NFL Workflow Guide, allowing SIG's engineering team to build a single ingestion pipeline that works identically whether pulling 2012 historical data or the current week's live slate.

                  Products Leveraged

                  SportsDataIO delivered the following tools to power SIG's prediction market models:

                  • NFL Competition Feeds: League, season, and pre-game structure spanning 2012–2026.
                    • Event Feeds: Play-by-play and statistical game data for deep modeling.
                      • Sports Betting Data: Historical and current line movement across markets.
                        • Projections: Forward-looking statistical projections for pricing inputs.

                          The Results

                          With a consistent data spine running from 2012 through the current 2026 season, SIG's prediction market desk trains its market making models on a deep historical base and deploys those same models against live, in-season NFL data without re-engineering its pipeline season over season.

                          • Continuity Across Seasons: A single schema spanning 15 seasons of historical data through live in-season play eliminates rework between backtesting and production.
                            • Faster Model Deployment: Consistent feed architecture lets SIG apply historical models to live games with minimal engineering overhead.
                              • Confidence in Pricing: Historical betting line data combined with real outcomes gives SIG's models a statistical edge grounded in verified data.
                                • A Single Source of Truth: SIG creates and trades NFL contract markets knowing its models work from a dependable data source, from historical trend to live game.

                                  Conclusion

                                  SIG's partnership with SportsDataIO shows how a consistent, quant-grade data pipeline spanning over a decade of historical coverage through live in-season play can underpin sophisticated market making in sports prediction markets. By eliminating the gap between historical modeling and live deployment, SportsDataIO gives SIG's Predictions desk the foundation to price and trade sports contracts with confidence.

                                  As market makers like SIG continue to drive trading volume across prediction markets, SportsDataIO's historical and in-season coverage is positioned to support and strengthen their ability to model and trade contracts across all major sports leagues.

                                  Want More?

                                  If you're ready to leverage data for similar success, explore the links below to learn more, or contact sales@sportsdata.io 

                                  Available Products

                                  Here's a list of the powerful tools that made it all possible:

                                  Learn More:

                                  Latest SportsDataIO Articles
                                  Copyright © 2026 SportsDataIOTM
                                  Contact Us