Methodology
Why the shortlist is useful, and why it is not a verdict.
The matcher compares eight household preferences with published breed-level tendencies. It ranks research starting points. It cannot predict the temperament, health, training needs, or fit of an individual dog.
1. Source snapshot
The structured store contains 206 profiles from the official American Kennel Club breed directory and individual breed pages, retrieved 2026-07-16. 205 profiles have all fourteen required 1-to-5 scores and are eligible for quiz ranking. A profile with an unpublished field remains browseable but is excluded from ranking rather than having a value inferred.
- Stored identity fields
- Breed name, group, origin, recognition year, life expectancy, size, height, weight, coat types, coat lengths, and popularity rank.
- Stored trait scores
- Adaptability, affection, barking, grooming, children, drooling, energy, other dogs, mental stimulation, strangers, playfulness, shedding, trainability, and protective tendency.
- Not copied
- AKC article text, breed-standard prose, and photography are not republished. Each profile links to its official source page.
2. How ranking works
Each selected answer earns a closeness score against the related AKC trait. The weighted points are divided by the points available for that household, then rounded to a score out of 100. The number is relative within this database and is not a probability of success.
| Input | Weight | Treatment |
|---|---|---|
| Preferred size | 24 | Exact requested size receives full weight. Any size leaves this neutral. |
| Activity level | 22 | Distance from the selected daily-activity level. |
| Grooming frequency | 11 | Penalizes needs above the selected maintenance ceiling. |
| Shedding level | 10 | Penalizes shedding above the selected household tolerance. |
| Barking level | 10 | Penalizes vocal tendency above the selected noise tolerance. |
| Young children | 11 when relevant | Higher published child-tolerance receives more weight only when selected. |
| Other dogs | 7 when relevant | Higher published other-dog tolerance receives more weight only when selected. |
| First-time owner | 10 when relevant | Combines published trainability and adaptability. |
A specific size preference acts as a strong constraint. Very low published scores for young-child or other-dog tolerance also cap the final score when those household needs were selected. Popularity is used only to break an exact scoring tie, never as a quality signal.
3. Result labels and explanations
- Strong fit
- 82 or higher within the selected preference model.
- Worth a look
- 70 to 81 within the selected preference model.
- A wider stretch
- Below 70, shown only if needed to complete a five-breed research list.
The interface shows up to three strongest contributing reasons plus the first material mismatch. It does not hide tradeoffs behind a single score.
4. Limits
- Breed scores describe broad tendencies, not an individual dog.
- The matcher does not evaluate health, behavior problems, bite risk, training progress, local restrictions, housing rules, or breeder quality.
- No dog is guaranteed to be quiet, child-safe, dog-friendly, easy to train, low-allergen, or suited to an apartment.
- Young children and dogs always require active adult supervision, regardless of a published breed score.
- Mixed-breed dogs are not ranked because the source dataset is breed-profile based. Shelter and rescue staff may offer more useful individual observations.
5. Update policy
The checked-in source script retrieves the current AKC directory, retains only ranked breed profiles, reads factual fields and trait scores from each official profile, and fails on broken or structurally incomplete source pages. The snapshot date changes only when the source data is regenerated and reviewed.
Use the shortlist to decide which dogs to meet and which questions to ask. A responsible shelter, rescue, breed club, or breeder should help you evaluate the individual dog in front of you.
Build a shortlist