Disclaimer

A Note on AI Analysis

Some of the analysis on this site has been produced with the help of artificial intelligence and machine learning techniques. This feels worth being upfront about, along with a few other things.

I'm not a historian or an academic — I'm someone who found this dataset fascinating and wanted to see what modern AI tools could do with it. My understanding of the data is considerably more limited than that of the original research team, who spent years in archives across Scotland carefully gathering and verifying this information. Any patterns or insights identified here should be taken in that spirit — as the observations of an interested amateur rather than authoritative historical analysis.

AI is good at finding patterns in data, but patterns aren't the same as explanations. Where this site identifies associations — between confession and execution, for example, or between certain charge types and outcomes — these are correlations in the available data, not established historical facts. A historian would rightly want to interrogate these findings much more carefully than I have.

It's also worth remembering that the data itself has real limits. It's a historical record assembled from documents that are hundreds of years old — incomplete, sometimes ambiguous, and shaped by who kept records and what they chose to write down. The gaps in the data are as historically significant as the data itself.

Finally, the people in this database were real. They lived, were accused, and in many cases suffered terribly. I've tried to approach the data with the sensitivity that deserves.