Algorithms permeate decision making at each intercept of the criminal punishment system. In turn, the algorithm used at one intercept produces a data point that informs an algorithm at the next.This talk will examine the role of algorithms and predictive models in the criminal punishment system. I will use findings from a team research project that combines oral histories with an otherwise unavailable data set from the PA Department of Corrections to better understand the Pennsylvania Additive Classification Tool (PACT) algorithm. The PACT algorithm determines the likelihood a person would endure additional disciplinary actions, can complete required programming, and gain experiences that, among other things, are distilled into variables feeding into the parole algorithm. Exploring the impacts of this algorithm's enactment tendencies will serve as an entry point to discuss the role of predictive algorithms at each node in the criminal punishment system. Given such power, examining algorithms used on people currently incarcerated offers a unique analytic view to think about how algorithms create our futures.