Why postcode averages mislead house hunters

"The average house price in this postcode is X" is one of the most quoted and least useful sentences in property. It is not that the number is wrong — it is usually calculated correctly from real sold prices. It is that the number is answering a different question from the one you are asking. You want to know what a three-bed terrace on a particular street should cost. The average tells you what a random assortment of flats, semis and the odd farmhouse happened to sell for over some window of time. Those are not the same thing, and the gap between them regularly runs to tens of thousands of pounds.

The mix problem: averages compare different houses, not prices

A postcode district contains everything from studio flats to detached houses with paddocks. When the average moves, the first question should never be "have prices changed?" but "have the houses that sold changed?"

Suppose a new block of one-bed flats completes and thirty of them register in the same quarter. The area's average sold price drops sharply — while every existing house on every street is worth exactly what it was before. The reverse happens when a developer finishes a run of executive detacheds. Statisticians call this a mix effect, and it is the single biggest reason small-area averages jump around. National indices spend serious modelling effort adjusting for it, which is one reason they disagree with each other — a story we tell in house price indices explained. A raw postcode average has no such adjustment. It is all mix, all the way down.

The small sample problem: one mansion moves the whole street

Averages need volume, and streets do not have it. A typical street might see a handful of sales a year; a full postcode sector may only see a few dozen. With samples that small:

• One large house selling makes the "street average" leap.
• One probate sale in poor condition makes it slump.
• A quarter with two sales tells you almost nothing at all.

The mean is especially fragile because house prices are skewed: there is a long tail of expensive properties and a hard floor near zero. One seven-figure sale in a street of modest semis drags the mean far above what any typical house would fetch. The median resists this better — half of sales above, half below — which is why serious analyses prefer it. But even a median from five sales is a coin toss dressed as a statistic.

Photo by Mingming Ouyang on Unsplash

Boundaries, time windows and other quiet traps

Three more ways the number misleads, worth a quick scan:

None of these are exotic. Most postcode averages you see online step on at least two of them at once.

What to look at instead

The fix is not a better average. It is to stop compressing an area into one number and look at the individual sales, filtered until they resemble the house you care about.

1. Start with actual registered transactions. HM Land Registry publishes every completed sale; rielty makes the full record — over 15 million transactions — browsable at /sold-prices. Real addresses, real dates, real prices.
2. Filter to comparables. Same street or immediately adjacent ones, same property type, similar size. Three genuinely comparable sales beat three hundred mixed ones.
3. Prefer recent, but read the trend. A comparable from two years ago plus a sense of the direction since is better than nothing; just do not treat an old price as a current one.
4. Check price per square metre where floor area is known. It is the quickest way to compare a 70m² terrace against a 95m² one honestly. We cover what the sold record reveals in price per square metre from sold records.
5. Look at the distribution, not the centre. The range between the cheapest and dearest comparable sales tells you how much the street rewards condition, extensions and position — information the average destroys.

Photo by Mingming Ouyang on Unsplash

Averages also hide what actually sets prices apart

Even a perfect street-level average would miss the point, because the things that make two nearby houses differ in value are specific: which side of the railway line, inside or outside a school's likely intake, on the flood plain or above it, under a flight path or not. Two streets a few hundred metres apart can sit in different worlds while sharing a postcode average. That is the whole premise of searching by how you want to live rather than by area label — commute times, schools, flood risk, noise and green space are searchable directly on rielty, and the underlying sources are documented at /data. Why identical houses a mile apart carry different prices is a big enough question that we gave it its own article: what drives local house prices.

The practical close

Next time a listing or a news story quotes an area average, do this: ask how many sales it is built on, over what period, and whether the houses in it look anything like the one you want. Then go and find five real sold prices for genuinely comparable homes and reason from those. It takes twenty minutes, it uses the same official data the average was built from, and it will tell you more than the average ever could. The average is a summary of the past mix of an arbitrary boundary. Your offer is about one specific house. Price the house.