Developing an Elo-rating System for Criminal Justice Practitioners: A Superior Method for Resource Allocation?

Record numbers of inmates are being released to the community for supervision each year. This poses a challenge to the agencies and officers in charge of providing supervision—there are far more offenders in need of supervision than officers can reasonably attend to. This has forced agencies to be innovative in how they allocate their resources. Empirical evidence suggests resources should be allocated to high-risk offenders, but how can officers determine which offender is at a higher risk of misconduct at any given time? Traditionally, this has been achieved by relying on standardized risk assessment. But risk assessment has been criticized for making only marginally better-than-chance predictions of future misconduct. We offer a novel solution by integrating the Elo-rating system into the community supervision decision-making process. We show, by drawing on test data from the Pathways to Desistance Study, that combining the Elo-rating system with traditional risk assessment may lead to increases in predictive power. We then discuss how Elo-rating systems could be used by community supervision agencies to more effectively prioritize their caseloads.