What Are Triangulated Groups?

A triangulated group is three or more people who all match each other on the same segment of the same chromosome. That's a stricter standard than shared matching, and the strictness is the point: shared matches tell you that people are connected somehow, while triangulation tells you they're connected through the same piece of inherited DNA, which must have come from a common ancestor. It's the closest thing genetic genealogy has to proof, and it's the step most people skip. This page explains how triangulation works, how to do it, and where its limits are. For working out what relationships the amounts imply, the DNA match calculator is the companion tool.

What Is Triangulation, Exactly?

Three conditions must hold:

• You match Person A on a specific segment — say chromosome 7, positions 25–48 million
• You match Person B on that same segment
• Person A and Person B match each other on that same segment

If all three are true, the segment is triangulated. It came from a single ancestral source, and all three of you inherited it from the same common ancestor or ancestral couple.

Why the third condition matters so much. Without it, you could be matching A on your father's copy of chromosome 7 and B on your mother's copy of the same region. The positions look identical on a chromosome browser — you have two copies of every chromosome, and most tools don't tell you which one a match sits on. A and B matching each other is what rules that out.

That confusion has a name: a false triangulation. It's the reason "we all match in the same place" isn't sufficient, and it catches people constantly.

A triangulated group is simply three or more people meeting these conditions on the same segment. Larger groups are stronger evidence, and once you can identify the common ancestor of any two members, the segment is assigned to that ancestral line for everybody.

How Is This Different From Shared Matches?

The distinction that matters most in this whole area, and the one most often collapsed.

Shared matching is a sorting tool. It divides a list of thousands into branches quickly, and it's where everyone should start — see what are shared DNA matches.

Triangulation is a proving tool. It's slower, needs more data, and answers a narrower question: is this particular connection real, and through which line?

The trap: if you and A both match B, that's a shared match. It does not mean the three of you share a segment. You might connect to B through your father and A might connect to B through B's mother's side entirely — two different relationships, one shared match, no common ancestor between the three of you.

The practical workflow uses both: cluster with shared matches to find the branch, then triangulate to confirm the specific ancestral line before building a conclusion on it.

How Do You Actually Triangulate?

You need a chromosome browser. 23andMe, MyHeritage, FamilyTreeDNA and GEDmatch all have one. AncestryDNA does not, so matches there can't be triangulated without uploading raw data elsewhere — which is the single biggest practical obstacle, given Ancestry's database size.

The process:

• Pick a match with a decent-sized segment, ideally 20 cM or more
• Note the exact segment — chromosome number, start and end positions
• Find others who match you on the same region, using the browser or a segment-match tool
• Check that those matches match each other on that same segment. MyHeritage and GEDmatch will do this step for you; elsewhere you may need to ask the matches to compare
• Record the group — names, segment boundaries, cM
• Identify the common ancestor using trees, shared matches and the amount shared
• Label the segment with that ancestral couple

Tools that help: MyHeritage's built-in triangulation view, GEDmatch's triangulation and matching-segment utilities, FamilyTreeDNA's matrix, and DNA Painter for recording and visualising the result.

The mapping payoff. Once you've assigned a segment to an ancestral couple, every future match on that segment is immediately placed. Over time you build a chromosome map where large stretches are labeled with ancestral lines — and new matches sort themselves.

Realistic expectations: this is hours of work per conclusion, not minutes. It's worth it for a specific question — an unknown parent, a contested line, a brick wall — and overkill for general match browsing.

How Do You Record a Triangulated Group?

Doing the analysis is half the job; recording it so it's still usable in a year is the other half.

Record the segment precisely. Chromosome number, start position, end position, and size in cM. "Chromosome 7, 25,400,000–48,100,000, 34 cM" is a record. "That bit on chromosome 7" is not.

Record every member of the group — name or username, testing site, shared cM with you, and the segment boundaries for each. Members' segments overlap rather than matching exactly, and the overlap is the shared region.

Record who you've confirmed and how. Which members have known relationships to you, which are still unidentified, and what evidence supports each identification.

Assign the ancestral couple once you can, and note the reasoning. "Assigned to John Barlow and Mary Hutchings because members A and C both descend from them and C's line is documented" is what makes the conclusion reviewable.

Keep a chromosome map. Whether that's a spreadsheet or a purpose-built tool, the point is a single place where each labeled region is recorded. New matches then sort themselves — a match on chromosome 7 at that position is immediately known to be on the Barlow line.

Note what's uncertain. Which groups are provisional, which segments are borderline in size, and where endogamy might be affecting things.

Date your entries. Match lists change as more people test, and knowing when you built a group tells you when it's worth revisiting.

Where Does Triangulation Fail?

Real limitations, and knowing them prevents false confidence.

Small segments. Below roughly 15 cM, false triangulations become common, and below 10 cM the exercise is close to meaningless. Segments that small are frequently population-level artefacts rather than recent inheritance. Use 20 cM as a working minimum where you can.

Endogamy. In populations that intermarried over generations, groups triangulate through multiple ancestors simultaneously. A triangulated group in an endogamous population may be genuine and still not point to a single identifiable couple. Standard interpretation needs heavy adjustment.

Pile-up regions. Certain areas of the genome — around centromeres and in regions with unusual structure — produce excess matching for reasons unrelated to genealogy. Segments there triangulate readily and mean nothing. Most tools flag or filter these.

AncestryDNA's absence. The largest testing database can't be triangulated directly. For many people that means the matches most likely to solve their question are the ones they can't confirm this way.

It requires cooperation. Triangulation needs three people's data, and on some platforms that means contacting matches and asking them to compare. Response rates are what they are, and a group you can't complete because one member never replies is a common frustration.

Privacy settings can block it. Some testers opt out of segment sharing or set their profiles to limit comparison, which is entirely their right and does remove them from the analysis.

Identical by state versus identical by descent. Not every matching segment is inherited from a recent common ancestor — some matching reflects the general population sharing common haplotypes. Larger segments are far more likely to be genuinely inherited, which is the whole reason for the size threshold.

It confirms a segment, not a relationship. A triangulated group proves those people share ancestral DNA. It doesn't tell you the relationship or how far back the ancestor sits — that still comes from shared amounts and trees, as in how much DNA do cousins share.

When Is It Worth the Effort?

Worth triangulating:

• Unknown parentage or adoption, where certainty matters and the stakes are high
• A contested or uncertain line in your tree, where DNA can settle it
• Brick walls where you have candidate ancestors and need to test which is right
• Confirming a paper trail you're not sure about
• Building a chromosome map, which pays off over years of research

Not worth it:

• Ordinary match sorting — clustering is faster and covers more ground
• Small segments, where the result won't be reliable anyway
• Endogamous lines, unless you're prepared for the extra complexity
• Any case where a straightforward document would settle the question more cheaply

A worked example of when it earns its keep. Suppose you have four candidate couples who could be your great-great-grandparents on one line, and documentary evidence is ambiguous. You find a 32 cM segment shared with two matches who both descend from one of those couples through documented lines, and confirm all three of you triangulate on it. That segment now has a source, and three of the four candidates are eliminated. No amount of shared-match clustering would have separated them, because all four families lived in the same parish and everyone matches everyone.

A sensible order of work for most people: cluster your matches with shared matching, identify what you can from trees and shared amounts, then triangulate the specific conclusions you want to be certain about. X-DNA fits in here too, since it excludes lines outright and can shortcut the search considerably — see X-DNA inheritance explained, and centimorgans explained for the underlying unit.

Frequently Asked Questions

What does triangulation mean in DNA genealogy?

Three or more people all matching each other on the same segment of the same chromosome. Because that segment must have a single ancestral source, triangulation demonstrates that all three inherited it from a common ancestor — which shared matching alone doesn't establish.

Is a shared match the same as triangulation?

No. A shared match means you and another person both match a third person, possibly on completely different segments through different ancestors. Triangulation requires all three to match one another on the same chromosome positions.

Which DNA sites support triangulation?

23andMe, MyHeritage, FamilyTreeDNA and GEDmatch all provide chromosome browsers, and MyHeritage and GEDmatch have built-in triangulation tools. AncestryDNA does not offer a chromosome browser, so its matches can't be triangulated without transferring raw data elsewhere.

What is the minimum segment size for triangulation?

Around 15 cM as an absolute minimum, and 20 cM or more preferably. Below 10 cM, false triangulations are common because small matching segments frequently reflect population-level sharing rather than recent inheritance.

Why does triangulation fail with endogamy?

In populations that intermarried over many generations, people share DNA through numerous distant paths at once. Groups triangulate genuinely but through multiple ancestors simultaneously, so the result doesn't point to a single identifiable ancestral couple.

Do I need to triangulate every match?

No. Clustering with shared matches sorts a match list far faster and covers everyone. Triangulate specific conclusions you want certainty about — unknown parentage, contested lines, brick walls — rather than as a routine step.

One thing worth being clear about: triangulation proves that a segment descends from a common ancestor. It does not prove which ancestor, and it does not prove a documented relationship is correct. What it does is make a hypothesis testable — and in a field where a great deal gets asserted on the strength of a name matching in an online tree, that's a meaningful step up.

Triangulation is the difference between "these people are connected to me somehow" and "these people and I inherited this segment from the same ancestor". Use clustering to find the branch and triangulation to prove the line, keep to segments of 20 cM or more, and record the groups as you build them. Check what the shared amounts suggest alongside it with the DNA match calculator.