Patent Examiner Behavior Analytics
Search over 45,000 USPTO examiners by name. View grant rates, claim scope reduction, and prosecution burden — three behavioral dimensions derived from k-means clustering of 15,014 examiners with complete prosecution data.
Based on the Novelty Hill IP examiner clustering research
Three Orthogonal Behavioral Dimensions
PCA and correlation analysis of 21 examiner metrics identified three independent dimensions that together account for 74% of behavioral variance.
Grant Rate
Percentage of decided applications resulting in grant. This is the foundational outcome metric and is largely independent of the other two dimensions.
Claim Narrowing
Percentage change in ClaimScore 2 from the published application to the granted patent. Measures the examiner's effect on claim scope during prosecution, normalized to enable comparison across technology areas.
Prosecution Burden
Weighted composite of prosecution events: (Non-Final × 1) + (Final × 2) + (RCE × 3) + (Appeal × 4). Captures the cumulative procedural cost of reaching grant.
Why Claim Narrowing Changes the Analysis
Grant rate alone cannot distinguish between examiners who reject applications outright and those who allow applications only after extracting significant claim scope. Consider two examiners from the research dataset with nearly identical grant rates:
| Dimension | Examiner Pryor The Skeptic | Examiner Springer The Adversary |
|---|---|---|
| Grant Rate | 49.1% | 42.2% |
| Claim Narrowing | −26.3% | −43.3% |
| Scope Retained | ~74% | ~57% |
| Prosecution Burden | 3.17 | 4.81 |
A grant-rate-only analysis would classify these as nearly identical prosecution environments. The claim scope and burden dimensions reveal that patents examined by Examiner Pryor retain approximately 74% of originally-filed scope, while those examined by Examiner Springer retain approximately 57% — a difference that can determine whether a granted patent covers a competitor's product or can be designed around.
Four Examiner Archetypes
K-means clustering (k=4) of 15,014 examiners identifies four behavioral groups. Silhouette score: 0.328.
The Ally
37.6% of examiners · n = 5,647
High grant rates with minimal prosecution burden. Applicants retain most of their originally-filed claim scope. The largest cluster; individual examiners may differ meaningfully from cluster averages.
The Collaborator
29.1% of examiners · n = 4,367
Above-average grant rates with moderate prosecution burden. Claims are reduced meaningfully but not severely. Standard prosecution approaches typically yield allowance after reasonable engagement.
The Skeptic
18.5% of examiners · n = 2,785
Low grant rates but near-average prosecution burden. Applications tend to resolve without extended prosecution. The moderate claim narrowing among granted cases suggests rejections focus on patentability rather than scope reduction.
The Adversary
14.8% of examiners · n = 2,215
Low grant rates combined with prosecution burden nearly double the population average and severe claim narrowing. The −45.2% narrowing reflects only granted applications; abandoned applications may have faced steeper requirements.
ClaimScore 2 Methodology
Claim scope reduction is measured using ClaimScore 2, a computational linguistics system that quantifies patent claim breadth by analyzing the structure and element count of claim text. Scores range from 1 to 100, where higher values indicate broader claims. Claim narrowing is computed as the percentage change from the published application score to the granted patent score.
ClaimScore 2 has been validated against 8,197 U.S. district court infringement decisions. Broader claims (higher scores) show a 13.4 percentage point higher infringement rate than narrower claims (chi-squared = 69.2, p < 0.000001).
Read the full methodology paper