The inherent nerdiness of baseball
On Wednesday, Major League Baseball implemented its latest attempt to better measure the sport with the debut of a rule change: teams now have two chances to challenge balls and strikes per game, forcing a review. The innovation was a middle-ground between full “robot umps” and no review at all, with fans and players preferring some human element involved (explore the new ABS rule).
The sport’s obsession with data collection goes back to the 1850s. English migrant Henry Chadwick adapted cricket’s scoring conventions to the nascent American game, inventing the box score in 1859 and key statistics like batting average and the “K” abbreviation for a strikeout, which are still used today.
In time, some viewed the conventional metrics as insufficient. That skepticism culminated in Bill James, whose sabermetrics brought empirical analysis to baseball statistics in 1977 (jumpstarting a revolution narrated in the book and film “Moneyball”). The league went all-in on the approach in 2015 when it deployed Statcast in its stadiums, which tracks every pitch, hit, and play in the field using 12 ballpark cameras, generating more than 7 terabytes of data per game for real-time machine learning.
