S
Thanks, both points are right, and I checked your numbers: "Ann Lee"/"Ann Lea" scores 0.857 and "Jon Smith"/"Jan Smith" 0.889 with NAMESCORE, so a flat 0.85 lets short different names through. Scoring the raw "Smith, John" against "John Smith" gives just 0.09, while the sorted key gives 1. I've updated the guide to recommend a stricter cut-off (or a "check by eye" column) for short names, and to score the cleaned key rather than the raw cell. Appreciate the careful read.