Canine cognition · Research sample of 25,463 dogs
Does Breed Predict How Smart a Dog Is?
Breed is the first thing anyone asks about a dog and the first thing most people use to predict how it will behave. For coat, size, and some working drives, that prediction is reasonable. For cognition, it is remarkably weak — and this chart shows exactly how weak.
The comparison is simple. On one side, the full spread between the seven AKC breed-group averages for reasoning. On the other, the spread between individual dogs, from the 10th to the 90th percentile of the research sample of 25,463 dogs. The individual bar is more than twelve times longer. Put plainly: pick two dogs of the same breed at random and they are likely to differ far more from each other than their breed group differs from any other group.
This is why breed-based expectations so often fail their owners. The Labrador who cannot follow a point, the terrier who solves puzzles methodically, the herding dog with no interest in problem-solving are not defective examples of a type — they are ordinary evidence that cognition varies dog by dog. Your dog is not a breed average, and the only reliable way to learn how it thinks is to test the dog in front of you.
The spread in reasoning scores between breed groups, compared with the spread between individual dogs.
Breed barely predicts how a dog thinks
Individual dogs differ on reasoning more than twelve times as much as breed groups do.
Research sample of 25,463 dogs · 7 AKC group means vs the 10th–90th percentile of individual dogs
- Horizontal axis: Reasoning score
- Measured in proportion of trials correct, 0–1
- Bars: Spread between the 7 AKC breed-group means vs the 10th–90th percentile of individual dogs
- Measured in score range on the same 0–1 reasoning scale
Reasoning score — proportion correct (0–1)
Reasoning score axis (0–1). Breed-group means come from 7 AKC groups (n=856–3,600); the individual range is the 10th–90th percentile across the 25,463-dog research sample.
Source: Dognition Canine Cognition Data Center · research sample of 25,463 dogsdognition.com/data
The finding: Breed groups differ by 0.061 on reasoning. Individual dogs differ by 0.75 — more than twelve times as much. Your dog is not a breed average.
How we measured this
Frequently asked questions
- Does breed determine how intelligent a dog is?
- Barely. Across 25,463 dogs, the spread in reasoning scores between individual dogs (10th to 90th percentile) is more than twelve times the spread between the seven AKC breed-group averages.
- Why do breed intelligence rankings exist then?
- Most rankings measure obedience trainability reported by judges, which reflects biddability and working history rather than direct measures of memory, reasoning, or social cognition.
- What does predict how a dog thinks?
- Individual assessment. Measuring a specific dog across the five cognitive dimensions predicts its behavior far better than its breed label does.
Source: Dognition — a research sample of 25,463 dogs. Free to share with attribution.
Test your dog