Pre/Dicta, the litigation analytics platform notable for predicting federal court judgments, is venturing into new grounds. Previously curtailed to predicting outcomes on motions to dismiss, it can now forecast outcomes for additional motion types and even predict litigation timelines.
Since its launch in July 2022, Pre/Dicta limited its scope to motions to dismiss, achieving a remarkable 85% accuracy rate. The platform has now broadened its horizon, providing similar outcome predictions for four additional motion types: summary judgment, venue transfer, class certification, and compel discovery.
Notably, Pre/Dicta is utilizing data profiling to model outcomes for specific judges, cases, and courts, rather than predicting the outcomes directly, as it does for motions to dismiss. Pre/Dicta founder Dan Rabinowitz humorously refers to this approach as finding their “doppelgangers,” meaning the platform seeks out judges with similar biographical data to predict the likely ruling of a particular judge.
The importance of this approach comes to the fore in situations where a judge has never ruled on a certain type of case before. By utilizing a “like for like” strategy, Pre/Dicta can still show likely outcomes for cases like class certification.
This method is also applied to predict case timelines, using “doppelganger” cases to show likely outcomes for the three primary stages of litigation: prediscovery, discovery, and trial.
During a discussion, Rabinowitz mentioned that their analysis had demonstrated that these elements, namely parties, attorneys, and judges, have the highest correlation to predict outcomes—surpassing even the influence of the law and the facts of the case.
The development of these new features by Pre/Dicta came on the heels of its acquisition in January of the databank from Gavelytics, a now-defunct legal analytics firm. Though Pre/Dicta’s primary interest was the state coverage data, it also gained a wealth of insight into litigation events in federal courts.
Utilizing Pre/Dicta’s service requires merely inputting a case number. Consequently, a dashboard will display, offering a prediction as to the likelihood of a motion to dismiss being granted, an analysis of similar cases, and a case timeline showing likely outcomes at each of the three stages of litigation, as well as the probable timeline.
The dashboard also provides a Motion Model, presenting expected grant rates for the different types of motions. Lastly, there is a Judicial Benchmarking Section, which presents grant rates for similar cases for the specified judge, for the circuit, and for judges with similar biographical profiles.