Understanding Hocking Identification Guide: A Practical Walkthrough
I've spent years sorting through DNA matches and building out family groups, and the Hocking Identification Guide is one of those things that sounds more complicated than it actually is. It's basically a structured method for grouping unknown DNA matches by shared ancestry and then narrowing down which branch of a family tree those matches belong to. If you're doing adoption search work or trying to break through a brick wall on a specific surname line, this is how people actually use chromosome data to figure it out. The core idea is that when you have multiple DNA matches who all share the same surname or family line, you can systematically organize them to see where their common ancestors are. The Hocking Identification Guide is really just a framework for that process. Here's how you actually do it. First, you pull your full match list from whatever testing platform you use — AncestryDNA, 23andMe, MyHeritage, FamilyTreeDNA. You export the data if you can. What you're looking for are clusters of matches that share significant segments of DNA with each other, not just with you. That's the key detail most beginners miss. Matching you individually doesn't tell you much. Matching you and matching each other does.
You use a tool like GEDmatch's One-to-Many, their Cluster tool, or a third-party app like DNA Painter to find those shared matches. When you see three or more people who all match you in the same chromosomal region and also match each other, that's a triangulated group. You've found a common ancestor. The Hocking Identification Guide takes this a step further by organizing those triangulated groups into a sortable, visual format. You assign each group a surname tag, a likely ancestor, and a chromosomal range. Over time you build out a map of your genome that shows which segments belong to which branch of your family tree. When a new match comes in, you look at where they fall and immediately know which side of your family they're from.
Setting It Up: The Actual Steps
Start by exporting your match data from every testing site you have results from. Ancestry gives you a CSV of top matches with centimorgan values. GEDmatch has their own export tools. Pull them all into a spreadsheet. Columns you'll want: Kit number, surname, email, centimorgans shared, number of SNPs, and shared match information if available. Next, you run the shared match analysis. On Ancestry you click into individual matches and check the shared matches tab. On GEDmatch you use the One-to-One comparison tool across multiple kits. On DNA Painter you use their Shared CM Project data combined with your own matches. Group the matches that share significant segments together. Anything over about 15 cM is worth investigating. Below that, you're likely looking at pseudo-segments or coincidental sharing, especially in endogamous populations. This is where the Hocking Identification Guide framework becomes useful. Instead of letting the groups stay messy, you assign them identifiers. Hocking-1, Hocking-2, whatever makes sense. You note the probable common ancestor based on any tree information you can pull. You record the chromosomal location. You do this for all the surnames in your background, not just Hocking. The guide is adaptable.
Get the Full Details
A Real Problem I Ran Into
My first serious attempt at using this approach hit a wall with a cluster of matches on chromosome 12 that all appeared to share a Hocking ancestor, but the segments were all between 7 and 12 cM. At that range, especially for matches on the same chromosome, I was getting a lot of false positives — people who looked like they shared a recent common ancestor but actually matched me through a different segment or through endogamy. My initial grouping was completely wrong. About 40% of the matches I'd assigned to that Hocking cluster were actually from a different surname line entirely. The workaround was to stop relying on the shared match tool alone and instead use chromosome mapping with documented cousins. I found three second cousins whose trees I knew were solid, ran the One-to-One comparisons with each of them, and used their known segments as anchors. Any Hocking match that aligned with their confirmed segments stayed in the group. Anything that didn't overlap got kicked out. This cut my cluster from about sixty people down to roughly twenty who were actually related through the Hocking line. It took me about six hours of careful segment-by-segment checking, but after that the rest of the work was straightforward.
Common Pitfalls
Endogamy is the biggest problem. If your ancestors come from a population with lots of intermarriage — Jewish communities, Acadian families, isolated rural populations — the shared segment thresholds need to be much higher. I use 20 cM as a minimum for endogamous lines instead of the standard 10 or 12. The cluster will be smaller but far more accurate. Another pitfall is confirmation bias. You'll find a Hocking match and start forcing other matches into that group because you want them to fit. The data will tell you when they don't belong. A match showing 45 cM with you but zero shared segments with the rest of the Hocking cluster is not part of that group. It's its own thing. Segment size matters more than raw centimorgan count. Two people can share 30 cM total but across five tiny segments, or share 15 cM in one large continuous block. The large block is far more reliable as evidence of a recent common ancestor. I prioritize single-segment matches over scattered multi-segment ones every time.
What This Method Doesn't Do Well
The Hocking Identification Guide, like any cluster-based approach, is nearly useless for adoptees with fewer than about fifty total matches. You need enough data points to form clusters. With twenty matches or fewer, you're just looking at individuals, not groups, and the method collapses. In that case, focusing on shared match trees and traditional genealogical research is more productive. It also doesn't help when the common ancestor is too far back. Once you get past maybe 5 to 7 generations, the segments get too small and too fragmented to reliably triangulate. The guide works best for brick walls in the 3rd to 5th great-grandparent range, not for deep ancestral lineage work. For people who need something simpler — or who have very small match lists — the approach is still useful if you scale it down. Start with the ten strongest matches, find their shared connections, and build one cluster at a time. The Hocking Identification Guide doesn't require a complete system. It works incrementally.
