top of page
Wood and Tiles


Relevance-based prediction is a flexible system for generating effective predictions across many domains. 

NBA draft paper.png

Predicting NBA Draft

Pick Performance

Working Paper

Megan Czasonis, Mark Kritzman, Cel Kulasekaran and David Turkington show how relevance-based prediction identifies the optimal blend of previously drafted players and predictive variables to predict the performance of an NBA draft prospect. They also show how fit, which measures a specific prediction's reliability, offers guidance about how committed a team should be to a draft prospect. 

Harry M. Markowitz


The paper "Relevance" by Megan Czasonis, Mark Kritzman, and David Turkington won the 2022 Harry M. Markowitz Award for best paper in the Journal of Investment Management. The final selection of the award was determined by a group of Nobel Laureates. 

Prediction Revisited


Megan Czasonis, Mark Kritzman, and David Turkington released their book "Prediction Revisited: The Importance of Observation" in 2021. It takes the reader on a tour of the foundations of relevance-based prediction, with each topic presented three ways: by intuition, by mathematical rigor, and by empirical illustration. Learn a new perspective on how data informs the future.

Prediction Method

Journal Article

In this 2023 article, Megan Czasonis, Mark Kritzman, and David Turkington introduce extensions to the foundations of relevance-based prediction, including the application of codependence to the simultaneous selection of predictive variables and observations. 



Prasanta Mahalanobis was a Cambridge-trained statistican who became one of the most celebrated intellectuals in India. While working with an anthropologist, he derived an elegant formula for summarizing a set of measurements of human skulls. His formula, called the Mahalanobis distance, has since been used to measure financial turbulence, to diagnose diseases, and to develop autonomous vehicles. The Mahalanobis distance is a critical component of relevance-based prediction.



As a student at MIT, Claude Shannon invented digitization. He then went on to Bell Labs where he developed a comprehensive theory of communication, which rests on the notion that information is an inverse logarithmic function of probability. This insight enabled the Digital Age as well as relevance-based prediction.

bottom of page