Sports by the Numbers

Hey Everybody,

The second episode of the Harvard Data Science Review Podcast was "Tracking the (Money) Balls: How Data Science is Becoming a Game Changer," focusing on the growth of data science and statistics being used in sports. Kirk Goldsberry, an NBA data analyst at ESPN, and Brian Macdonald, faculty in Sports Analytics at Carnegie Mellon University, said that there are two main ways sports organizations use data analysis: 

1. management decisions: who to trade for, draft, pick up in free agency and other player personnel decisions, and 

2. coaching decisions: effectiveness of plays or player combinations. 

For example, in the NBA, three point shooting has become a much more dominant part of the sport than it was ten years ago. Players have trained for long distance shots in order to go for that extra point. This changes the decisions management makes when bringing shooters versus other players to the team, and changes the coaches’ decisions on drilling three point shots more often. Another example is how the MLB uses statistics - baseball has become the sport that uses statistics the most, and led people to see an increase in both strikeouts and home runs and therefore less balls in play. 

Another topic they covered was sports betting. There’s been a large increase in betting recently with countless sportsbook apps available to use. Through these apps, the use of data analysis has made it easier and more profitable to predict outcomes, and it has become an arms race of the best data and intel in each sport. The future of sports betting will be filled with not just new types of statistical analysis but new technological advancements and codes to assist betters with their predictions.

Also, I’ve always known about the vast amounts of statistics in baseball (from On Base Percentage to Earned Run Average to much more), but I never knew that these numbers have divided the sport into more strikeouts and home runs, which are two very different aspects of the game. This also makes me worry about baseball’s future because I care so much about it. The fact that there are more strikeouts and home runs than before means that less balls will be put in play and there won’t be as much action on the field. This is bad because fans want to see action on the field rather than just strikeouts. Brian Macdonald asked "How can we govern our sports analytically with data science?" This stands out to me because we can use statistics to change rules and regulations for a more safe and exciting game.

 I was also surprised to see how serious people can take sports betting and use statistical analysis in order to win very large amounts of cash rather than just bet for fun.

I wonder how NFL regulators use stats to determine safety precautions with protective gear, grass vs turf, or play designs. And, how do statistics advance soccer (or futból) - the world’s favorite sport?

Your MVP,

Noah