The Relevance Edge (A New Science of Sports Prediction)
| Expected release date is Apr 13th 2027 |
- Availability: Confirm prior to ordering
- Branding: minimum 50 pieces (add’l costs below)
- Check Freight Rates (branded products only)
Branding Options (v), Availability & Lead Times
- 1-Color Imprint: $2.00 ea.
- Promo-Page Insert: $2.50 ea. (full-color printed, single-sided page)
- Belly-Band Wrap: $2.50 ea. (full-color printed)
- Set-Up Charge: $45 per decoration
- Availability: Product availability changes daily, so please confirm your quantity is available prior to placing an order.
- Branded Products: allow 10 business days from proof approval for production. Branding options may be limited or unavailable based on product design or cover artwork.
- Unbranded Products: allow 3-5 business days for shipping. All Unbranded items receive FREE ground shipping in the US. Inquire for international shipping.
- RETURNS/CANCELLATIONS: All orders, branded or unbranded, are NON-CANCELLABLE and NON-RETURNABLE once a purchase order has been received.
Product Details
Overview
A transparent prediction method that goes beyond regression and machine learning in sports
In The Relevance Edge: A New Science of Sports Prediction, Mark P. Kritzman, MIT Sloan Senior Lecturer and founding partner of Cambridge Sports Analytics, together with data analytics colleagues and Cambridge Sports Analytics founding partners Megan Czasonis, Cel Kulasekaran, and David Turkington, introduce relevance-based prediction to the world of sports analytics. The method is designed to overcome the limitations of both linear regression and machine learning, and has been adopted in a wide range of fields ranging from finance and investing to health and medicine. Unlike regression's static assumptions or machine learning's opacity and overfitting, relevance-based prediction delivers transparent, adaptive forecasts and flags each prediction's reliability in advance. These benefits, and more, are available because relevance-based predictions are formed as carefully weighted averages of the most relevant historical experiences.
The Relevance Edge builds from core concepts to full technical depth across seven chapters, grounding each idea in real case studies: NBA draft prospect evaluation, expected goals analysis of a Barcelona-Real Madrid match, NFL wide receiver prospect assessment with missing data, and World Series pitcher matchup confidence scoring. A dedicated final chapter provides rigorous mathematical treatment alongside profiles of intellectual forerunners who have helped pave the way to more transparent and effective predictions.
Readers will also learn:
- The fundamental building blocks that determine how relevant one athlete or experience is to another for the purpose of forming a prediction
- How to measure the task-specific importance of each predictive variable instead of relying on misleading averages
- How to handle missing data and extreme outliers that standard model-based approaches consistently struggle with
- Why linear regression and neural networks each fall short when applied to dynamic sports environments
- Real-world insights from player prediction and in-game decision case studies across basketball, baseball, soccer and American football
Written for professional sports organizations, sports technology companies, sportsbook analysts, and quantitative bettors, The Relevance Edge also serves students and faculty in sports analytics, data science, statistics, and predictive analytics. Professionals in business management, finance, risk analytics, healthcare research, and many other fields will find the methodology directly transferable to their own forecasting challenges.
