Skip to main content
CCI

What the 'Risk Flags' section means

Risk flags highlight arguments, procedural patterns, and behaviors that have historically correlated with unfavorable outcomes before a specific judge.

Updated April 1, 2026Judge Research

The Risk Flags section appears on every judge profile and inside the Outcome Predictor results. It surfaces specific patterns from historical data that have correlated with negative outcomes before this judge.

How risk flags are identified. The AI analyzes the full text of available tentative rulings and attorney observations to identify recurring themes in denials and negative outcomes. A flag is generated when a pattern appears consistently enough to be statistically meaningful — not from a single data point.

Common flag categories. Risk flags typically fall into a few categories: (1) Procedural flags — arguments or motions that this judge consistently rejects on procedural grounds before reaching the merits, such as failure to meet IDC requirements, deficient separate statements, or improperly noticed motions. (2) Substantive argument flags — legal theories or argument styles that have not performed well before this judge even when the underlying law supports them. (3) Presentation flags — style-based observations, such as excessive length, over-citation, or overly aggressive advocacy tone that attorney notes indicate affects outcomes. (4) Timeliness flags — late filings or late-served papers that this judge treats more strictly than the statewide average.

How to use risk flags. Review the Risk Flags section when drafting your motion and again before finalizing your oral argument. Treat each flag as a checklist item. If your brief contains an argument flagged as problematic before this judge, either address the flag directly in the brief or have a very strong reason why this case is different.

Risk flags are not disqualifying. A flagged argument type does not mean you should never make that argument — it means you should be aware of the heightened risk and prepare accordingly.

Data confidence. Each flag shows the number of data points behind it. Flags based on 3 or fewer rulings are marked as preliminary and should be treated as a caution signal rather than a firm pattern. Flags based on 20+ data points are considered well-established.