The published record.
Model validation papers, outcome studies, and methodology documentation from the Splitifi research team.
Data science for legal decision-making.
Calibrated Outcome Prediction in Legal Contexts
How deterministic models outperform LLM inference for legal prediction. Foundational methodology paper documenting training standards, discrimination validation, and calibration requirements across the Splitifi production registry.
Judicial Behavioral Analytics
What judges' decision patterns reveal about outcome prediction and jurisdiction-specific strategy. Behavioral clustering methodology and its predictive utility, with documented ethical framework.
Settlement Zone Modeling
Finding where contested cases actually resolve. The mathematics of negotiated outcomes derived from real court records across custody, asset division, and support contexts.
Six domains. One methodology.
Outcome Prediction
Custody, support, division, fee awards. Calibrated probability models trained on real final orders across all 50 states and four international jurisdictions.
Judicial Analytics
Behavioral patterns, ruling tendencies, decision factors. Derived entirely from public court records across verified federal and state judges.
Settlement Intelligence
Zone calculation, timing, anchor points. Where litigants actually resolve contested issues, modeled from outcomes not surveys.
Litigation Finance
Win probability, award ranges, and duration modeling for litigation finance. The duration research series is published openly by Criterica Intelligence.
Model Validation
Calibration, time-aware splits, leakage detection. Actuarial-grade validation standards applied to every model in the production registry.
Multi-Jurisdictional Patterns
How outcomes differ by state, circuit, and judge. The same case filed in different jurisdictions produces measurably different, predictable outcomes.
Published findings.
Downloadable research papers across outcome prediction, judicial analytics, settlement modeling, and litigation finance.
Collaborate with the research team.
Academic institutions, courts, and research partners can apply for structured data access and collaborative research arrangements.