Shot in the Dark Forensics has officially entered its “string-board on the wall” era — except now the string-board doesn’t fall off the wall at 2am, and every connection has to earn its place.

Here’s the problem with historical forensic genetic genealogy: it’s not just “building a family tree.” It’s reconstructing reality from messy, partial, sometimes contradictory evidence. You’re dealing with DNA comparisons between living people, shared genetic segments, pedigree collapse, endogamy, surname changes, adoptions, aliases, record errors, and the occasional ancestor who appears to have lived in three states at once. Traditional spreadsheets and family-tree software are useful until they aren’t — because they’re designed for tidy lineage, not complex, overlapping networks of evidence.

That’s why we use relationship-based analytical modeling to individualize cases.

This approach mirrors how the work actually happens. Individuals, documentary sources, DNA evidence, and hypotheses are treated as distinct elements, explicitly linked by defined relationships. A person isn’t just a name on a branch — they’re connected to sourced records, genetic evidence, shared segments, proposed parentage, and clusters of related individuals. When new evidence arrives — a DNA comparison, a newly identified descendant, a historical document — it doesn’t just get added to a notes field. It becomes part of the case structure and immediately reshapes what is most likely, what is merely possible, and what is no longer plausible.

The real power is in the questions this kind of modeling allows us to ask.

Instead of “who might be related,” we can ask:

• Show everyone connected to this individual within a defined number of generational steps, using only sourced relationships.
• Identify candidate ancestors who independently connect multiple living people through separate child lines.
• Flag contradictions in the case structure, such as impossible timelines or mutually exclusive parentage.
• Surface the shortest evidentiary paths between two individuals — the minimal chain of relationships that must be true if a hypothesis is correct.

That last one is my favorite. It turns a sprawling, emotional, century-spanning mystery into something testable: a small number of concrete relationship paths that can be verified, challenged, or falsified.

This is how we individualize historical forensic genetic genealogy cases. We don’t just say “this looks like the right family.” We build a case file as a living network of evidence — where every connection has a reason, every hypothesis is clearly labeled as such, and every new data point has a defined place to land.

In other words: fewer vibes, more structure. Still plenty of mystery. Just with better math.