How many comparables you actually need
Three to five tight ones. More is not stronger, and it is often weaker.
Four to six good ones. The number matters far less than the quality, and past a point more comparables actively weaken a filing.
Why four to six
Fewer than three and a single unusual property distorts the median. Two comparables give you an average, not a level of assessment.
More than about eight and you are almost certainly including properties that do not really match — because you have run out of good ones and started widening.
Four to six is enough for a median to mean something, and few enough that you can defend each one in a sentence.
Why more is worse
This is counter-intuitive and it is the point most people get wrong.
An analyst reviewing your filing is looking for a reason to reject it — not out of hostility, but because that is what reviewing means. A set of twelve comparables containing two obviously unsuitable ones gives them that reason, and rejecting two is commonly treated as grounds to doubt the other ten.
Your filing is only as strong as its weakest comparable. Leaving out the marginal ones is not hiding evidence; it is not offering a target.
The test that actually decides
Look at how much your comparables disagree with each other, and compare that to the gap between you and their median.
Loose group: comparables at $11.20, $13.50, $15.80, $17.10. Spread about 44%. You are at $15.96, near their median. There is nothing here, and adding four more scattered properties will not create anything.
If the spread among your comparables is wider than your gap above them, you do not have a case yet — you have a noisy sample. Tighten before you file.
What to do when you cannot find four
Widen deliberately, one attribute at a time, and say what you did:
- Extend the age range.
- Extend the floor area range, remembering that larger comparables show a lower rate per square foot for reasons unrelated to fairness — so this cuts against you and you should say so.
- Extend to an adjacent valuation neighbourhood, and name it.
Stating the widening in the filing is far better than hoping it goes unnoticed. "The three nearest matches are slightly larger, which if anything understates the disparity" is a sentence that builds credibility. Being caught quietly including a property twice the size does the opposite.
The summary
Four comparables you can each justify beat twenty you cannot. Check the spread before the gap. And if there is genuinely nothing to compare against, that is a signal a uniformity argument is the wrong ground — not a reason to assemble a weak one.