How far a comp can be
A sold price is not equally informative from the next block and from across town. The report multiplies each comp by a distance weight, then uses those weights in a weighted median. This page uses the same steps with two invented sales so you can see which house the median picks. It does not pull a map. The comp weight guide stacks this multiplier with recency and similarity. Here the other two multipliers are held at 1 so distance is the only reason the weights differ.
The distance steps
Start every comp at a weight of 1, then apply distance. A sale at or inside 0.25 miles is multiplied by 3. A sale outside that and at or inside 0.5 miles is multiplied by 2.5. A sale outside that and at or inside 1 mile is multiplied by 2. A sale outside that and at or inside 2 miles is multiplied by 1.5. Anything farther than 2 miles stays at 1. A missing distance, a distance that is not a number, or a distance that is not greater than zero also stays at 1. That last case is the one people get wrong, because a blank feels like “very close” and the formula treats it as unknown.
The nearer low sale wins
Comp A sold for $190,000 and sits 0.2 miles away. That is inside a quarter mile, so the distance weight is 3. Comp B sold for $230,000 and sits 1.8 miles away. That is inside 2 miles and outside 1 mile, so the distance weight is 1.5. Total weight is 4.5. Half of that is 2.25. Sort by price, low to high, and walk until the running weight reaches half. Comp A is first. Its weight of 3 is already past 2.25, so the weighted median is $190,000. The average of the two prices is $210,000. The average treats a house almost two miles away as the equal of a house a fifth of a mile away. The weight does not.
The nearer high sale can win too
Reverse only the distances. Comp A is still $190,000 but now 2.4 miles away, which is past 2 miles, so the weight is 1. Comp B is still $230,000 but now 0.2 miles away, so the weight is 3. Total weight is 4. Half is 2. Walk from the low price. Comp A contributes 1, which is not yet 2. Comp B brings the running total to 4, which passes 2, so the weighted median is $230,000. The nearer house won again, and this time the nearer house is the more expensive one. “Always use the cheaper comp” is not the distance rule. The distance rule gives the close sale more votes. If the close sale is higher, the median can be higher.
A zero in the distance field does not sneak into the quarter-mile bucket. Zero is not greater than zero, so the multiplier is the neutral 1, the same as a house past two miles. If you have been dropping far comps by hand and also treating blanks as next door, you are using two policies at once. The formula uses one: unknown is not close. Type a real distance or leave the sale at a neutral weight. Do not invent 0.1 miles because the cell was empty.
What this does not decide
Distance does not repair a house that is a poor match. A sale 0.2 miles away in a different product is still a different product, and a weight of 3 will happily promote it. Throw it out before you weight it. Recency can outweigh a mild distance difference once you stop holding that multiplier at 1. The recency rule runs the same two-sale walk with dates instead of miles. The 70 percent offer is a later step. A median of $190,000 with an invented $30,000 rehab is $133,000 minus $30,000, which is $103,000, on the MAO calculator 70 percent line and on the ARV and 70% rule calculator if you type a single sold price of $190,000. That offer is only as good as the decision that the 0.2 mile sale belonged in the set. The prices and the miles are invented.
Two sales are enough to see the walk. They are not enough to feel confident. The count portion of confidence is the number of comps times 6, capped at 30, so two comps contribute 12 of those 30 points and no distance bonus fills the rest. A close pair can still be a thin pair. Add sales that pass the same match test, then weight them. Do not add a far sale only to make the count look larger and then complain that it moved the median. If it is past two miles it only has weight 1, but weight 1 is still a vote.
Questions
Does a missing distance get treated as next door?
No. A missing, non-numeric, or non-positive distance gets a neutral 1x multiplier. A stored zero from a provider is the same gap. Zero does not earn the quarter-mile bonus.
Is the nearer sale always the ARV?
No. Distance is one multiplier. In the second pair the nearer sale is the higher price, and the weighted median lands on that higher price because its weight crosses the halfway point of the total weight.
run a full analysis free when the median you can do by hand is not enough. A full analysis weights the comps, adjusts beds and baths, and prices the repair tiers. The sample figures there are invented, the same way every dollar on this page is invented.