What a 6-generation pedigree actually proves

Pedigree depth isn't bragging. It's the only reliable way to surface inbreeding risk before a litter is on the ground.

Three-generation pedigrees are the lingua franca of the dog world for one reason: they're easy to fit on a piece of paper. That's it. There's nothing biologically special about three generations. It's a typesetting artifact from the 1950s that survived into the database era because nobody questioned it.

Six generations is where the math starts telling the truth.

What the math actually does

The Coefficient of Inbreeding (COI) is a percentage that estimates how much of a dog's DNA is identical-by-descent from a shared ancestor. A COI of 0% means no detected common ancestor in the pedigree window. A COI of 25% means the parents were full siblings. Anything above ~10% is the territory where veterinary genetics literature flags significantly elevated risk for the recessive disorders that affect the breed.

The catch: COI is computed over the pedigree window you give it. A 3-generation COI can't see a great-great-grandparent who appears on both sides of the tree, so it returns a falsely low number. The deeper the window, the harder it is to hide a common ancestor without it showing up in the calculation.

Why six

Six generations covers 64 ancestor positions per parent — 126 unique slots once you remove the dog itself. That's deep enough to surface the most common high-risk pattern in modern purebreds: a popular sire from twenty years ago whose descendants now make up half the breeding population, appearing five or six generations back on both sides of a contemporary litter.

A 3-gen COI on the same litter would have read 0%. The 6-gen COI reads 8.4%. One of those numbers is true.

What we do with it

Every dog registered on Armstrong K9 has its pedigree stored to whatever depth the kennel can fill in. When a kennel goes to register a planned litter — by entering the sire and dam — the platform computes the COI of the resulting puppies over the deepest window we have for both parents and surfaces a warning if it crosses the 6.25% line (the equivalent of a single first-cousin-once-removed pairing). We don't block the registration. We just put the number in front of the kennel with a date, a methodology, and a one-click way to drill into which ancestor is the bottleneck.

Most kennels we've talked to, given the warning, change the sire or dam. The math is more persuasive than any opinion.

What pedigree depth is not

Six generations is not "more proves more." A pedigree filled with placeholder names and unverified ancestors three rings deep is worse than a clean three-generation tree, because the COI computed over the unverified data will be wrong with high confidence. A pedigree is only as good as the audit trail behind every name in it.

That's why our pedigree submissions go through admin review when an unknown ancestor is added — the same approvals queue that handles transfers and breed catalog edits. A name doesn't get to count toward the COI calculation until a human has looked at it.

What buyers should ask for

Three things, in order:

1. The 6-generation pedigree of both parents — not the puppy's 3-generation summary.
2. The 6-gen COI for the litter, with the date it was computed and the registry it was computed against.
3. The breed-club genetic-screening panel results for both parents (Embark, Wisdom Panel, or breed-specific). A clean COI doesn't help if the parents both carry the same recessive disorder mutation.

A serious kennel will hand all three over without flinching. Anyone who can't, or won't, is selling you something other than an audited pedigree.