Gerrymandering, which occurs in district-based legislative systems, is defined as the drawing of boundaries to protect or target an individual or a group. In the United States, partisan gerrymandering has distorted representation to all-time modern highs. Data analysis and computing, which powered the original offenses, can also be harnessed for diagnosis and prevention. Diagnosis, which can help state and federal courts, can be as simple as a t-test, or as complex as the automated exploration of billions of plans. Prevention requires the ability to draw plans and predict their consequences. With the right open data and software, reformers and interested citizens can get involved in the redistricting process. This talk will describe the road to reform from a data-analytic perspective, and ask how data science can help terminate gerrymandering.