Identical by Descent vs Identical by State

When a DNA service tells you two people “match,” it sounds definitive. But a match can mean two completely different things, and only one of them signals a real family connection. Genetic genealogy hinges on telling them apart: identical by descent versus identical by state.

Two reasons DNA can match

At any given position, there are only four possible letters, so with hundreds of thousands of positions in a raw file, some matching between any two people is inevitable. The question is why a stretch of DNA matches.

• Identical by descent (IBD): two people carry the same segment because they both inherited it, intact, from a shared ancestor. The match reflects genuine relatedness.
• Identical by state (IBS): the letters simply happen to line up, with no recent common ancestor behind them. The match is coincidence.

Both look like “matching DNA” on the surface. The difference is entirely in the history that produced them, and that history is what genealogy cares about.

Identical by descent: a real inheritance

An IBD segment is a piece of DNA passed down through a family tree. Somewhere back, an ancestor had that stretch, and it was handed down to two of their descendants, who now share it. Because recombination breaks DNA into smaller pieces each generation, IBD segments from a recent ancestor are long, while those from a distant ancestor are shorter.

That length is the key clue. A long shared segment is hard to explain by chance - the odds of a lengthy stretch matching letter for letter by accident are vanishingly small. So long shared segments are strong evidence of true, traceable relatedness.

Identical by state: coincidence in disguise

An IBS match is different. Across a short run of positions, two unrelated people can match purely because each position has limited options and common variants are, well, common. There is no shared ancestor doing the work; the alignment is statistical luck.

Short matches are where IBS lives. The shorter the shared stretch, the more likely it is to be noise rather than signal. A handful of matching positions between two strangers tells you almost nothing about their family connection, even though a naive comparison would flag it as a “match.”

Why length is the deciding line

This is why real relative-matching depends on finding long IBD segments and treating short ones with suspicion. Serious tools apply a minimum segment length before they will call a match meaningful, precisely to filter out IBS coincidence. Below that threshold, a shared region is more likely to be background chance than genuine inheritance.

It also explains a common source of confusion: two people can share a small amount of DNA and yet not be relatives in any useful sense. The DNA overlaps by state, not by descent. Distinguishing the two is not a technicality - it is the difference between a real cousin and a statistical mirage.

What this means for your matches

When you read a match list, the strong entries are those built from long segments that are almost certainly IBD. The tiny ones near the bottom deserve genuine skepticism, because a fair share of them are IBS artifacts rather than family. This is closely tied to how much total DNA is shared, measured in centimorgans - our explainer on centimorgans covers why segment length and total sharing matter so much.

The broader mechanics of turning your raw genotypes into these shared segments are covered in how DNA relative matching works. And if you want to compare two of your own files to see shared regions for yourself, you can do it privately with on-device comparison that keeps both files in your browser.

The single most useful habit in DNA genealogy is asking, of any match, whether it is by descent or merely by state.

This article is educational and is not medical advice.