Machine learning has made inroads into the toolkit of both academic economists and economic consultants. Competition lawyers who grasp when and how machine learning can improve our understanding of economics stand to gain from improved empirical analyses. All others are at risk of falling victim to hype and overselling. I discuss ways in which good practice for empirical analysis is similar and different to traditional econometric methods, explain the role that machine learning plays in economics overall, and how the impact of machine learning in the broader economy can impact the effect of competition policy and the work of competition authorities.