With our Machine Bias series, we are investigating the algorithms that are increasingly making decisions about our lives, from what news or ads we see online to what sentences are meted out for crimes. Algorithms are often proprietary "black boxes," raising important questions about transparency and due process. By collecting and analyzing the output of these systems, we set out to reverse-engineer and make accountable some of the algorithms that were having the biggest impact on people's lives. Our investigative methods included linear regression, statistical analysis, and the creation of our own software. Among the series' findings were evidence of racial bias in risk assessment systems, and the preferential treatment of Amazon's own products in its so-called open market.