Director Reasoning Analysis
Pattern analysis of discretionary denial factors and predicted outcomes
Tutorial Mode
Loading analysis data...
Statistics Mode
Showing statistics across all analyzed proceedings.
Argument Rankings
Using AI analysis of petitioner and patent owner briefs, this tab ranks the most common arguments by frequency. "Winning" arguments appeared in the prevailing party's brief.
Winning Arguments
Petitioner
| # | Argument | Frequency | Avg Weight | Factor |
|---|
Patent Owner
| # | Argument | Frequency | Avg Weight | Factor |
|---|
Losing Arguments
Petitioner
| # | Argument | Frequency | Avg Weight | Factor |
|---|
Patent Owner
| # | Argument | Frequency | Avg Weight | Factor |
|---|
Factor Breakdown
Legal categories the Director considers, split by party. A Patent Owner "prevails" when the Director denies the petition. A Petitioner "prevails" when the Director refers the case back to the merits panel.
Patent Owner Prevailing Rate by Factor
Factors where Patent Owner arguments led to Deny outcomes
No data available
Patent Owner Factor Cards
Petitioner Prevailing Rate by Factor
Factors where Petitioner arguments led to Refer outcomes
No data available
Petitioner Factor Cards
Predictions
The following represents AI-generated predictions in proceedings where briefs have been filed but the Director hasn't decided yet. The AI model predicts the likely outcome (Deny or Refer) with a confidence score based on arguments extracted from the briefs and presents a summary of its thinking. As the "Predicted Outcome" and "Summary" columns contain AI-generated content, the accuracy of the prediction carries no guarantee or warranty. Your individual assessment of a predicted outcome for a case can only be reached by your own study of the briefs.
| Proceeding | Petitioner | Patent Owner | Predicted Outcome | Confidence | Summary |
|---|
Model Accuracy
Evaluates the AI model's predictions against actual Director decisions. This tab only shows cases where we made a prediction BEFORE the decision was issued, then later compared our prediction to the actual outcome. Incorrect predictions are highlighted in red. As the "Accuracy" assessment depends on AI-generated content, the reliability of the accuracy assessment carries no guarantee or warranty. Your individual assessment of a predicted outcome for a case can only be reached by your own study of the briefs.
| Proceeding | Actual | Predicted | Confidence | Match |
|---|