What makes a comparable actually comparable
Start with the buyer, not the map. A usable comp attracted the same buyer yours will. That usually means the same pocket, similar lot grade and view, a similar era of construction, and similar system age behind the walls.
- Same pocket — bench, historic core, or west side. Distance is a poor proxy for similarity here.
- Similar lot grade and usable yard. A flat half-acre and a steep half-acre are not the same asset.
- Comparable view exposure. View is priced, and it does not average.
- Similar mechanical age — roof, furnace, water heater, electrical panel. Two identical floor plans can be a full renovation apart.
- Similar finished-basement status and garage count, since both are commonly mispriced by automated adjustments.
- Closed within a recent window, and closed — not pending, not expired, not withdrawn.
Reject the comp if…
A short reject list saves more money than a long adjustment spreadsheet:
- It sold between family members, in an estate rush, or in any transaction that was not openly marketed.
- It carried heavy seller concessions that are not visible in the recorded price.
- It is new construction with builder incentives, being used to price a resale — the incentives are part of that price.
- It is on the other side of the freeway from you, in either direction.
- It has a view your home does not have, or vice versa, and the adjustment is being guessed at rather than supported.
- It has been on the market so long that its final price reflects fatigue rather than value.
Why radius searches fail in a small, mixed market
Automated valuation models draw a circle and average what is inside it. In Farmington that circle routinely contains a bench view home, a 1940s cottage off Main, and a three-story townhome near the station. Averaging those three produces a number that describes none of them and misleads all three owners. The fix is not a bigger radius — it is a hand-picked set of five or six sales chosen for similarity, and an honest note about what each one is missing.
How to sanity-check a comp set someone hands you
Ask three questions of every comp set, whether it came from a portal, an appraiser, or an agent: Would a buyer touring my home have also toured this one? What is the single biggest difference, and how was it adjusted? And what did the seller actually net after concessions? If the answers are vague, the price built on them is vague too.
The comps you never see
The homes that failed to sell are evidence as well. A pocket with several expired listings at a given price band is telling you where buyer resistance starts, and that ceiling is often more useful than the highest closed sale — which may have been a one-off buyer with a specific reason to overpay.