A friend tosses out "a unit across the block just sold for 1.5B won," and from that day your sense of what your own home is worth quietly shifts. Strange, isn't it. You never checked whether that number was right, yet the line in your head slides over anyway. Few markets leave a single number stuck this long. Price tags lodge fast and leave slow. The two speeds are nowhere near symmetric.

How 1.4B becomes the line in your head

1) Single-number learning. "A 31-pyeong unit nearby just transacted at 1.4B" — one line, and it's your reference.
2) Range contraction. After that, only listings between 1.3B and 1.5B feel reasonable. Anything outside just looks wrong.
3) Confirmation bias. You keep the sales that fit the range and wave off the rest as "special cases."

The whole cycle wraps up in weeks. And for a good while after, that narrow band is your entire market.

Why real estate, of all things, is this fragile

Four structural reasons pile up.

• Trades are rare. Stocks change hands thousands of times a day. A specific apartment trades a handful of times a year. Small sample, so each single transaction swings outsized weight.
• Information is uneven. Not every sale is public, and media and platforms spotlight only what they choose to show.
• Emotion creeps in. A home is where you live, not an asset on a screen. The price gets tangled with daily life and memory.
• Leverage muscles in. Most buyers borrow, so it's not just the price — mortgage rates and loan-to-value tug at the decision at the same time.

Questions to throw at a suspicious anchor

1. Is this a closed sale price, a listing ask, or a media edit of one high-profile deal?
2. Once you control for floor, orientation, size, and renovation — how many genuinely comparable sales are left?
3. How far have interest rates and taxes moved since that reference transaction?
4. How much of the price is just this building's own story — school zone, redevelopment hopes?

How to read PriceGuess's real estate mode

The answer here is not a specific transaction. It's a representative value inspired by public listings — hold onto that. We don't recommend listings and we don't forecast prices. There's exactly one reason to keep playing: to grow a range sense, the gut feel of "for these conditions, prices usually sit in this band."

When you actually buy

Real purchase or rental decisions start from official transaction registries — Korea's MOLIT registry, the U.S. MLS, Zillow, Redfin — plus a sit-down with a licensed broker. The surest way not to get dragged around by one number is plain: lay several independent sources side by side and compare them at once.

Same home, different starting point — in numbers

A simple example shows best how an anchor steers you. For the same 31-pyeong apartment, friend A says "a unit next door transacted at 1.4B," and the same day friend B says "the complex across the road slipped to 0.9B." Both are hypothetical figures as of writing. Even so, whichever you hear first sets the starting point in your head.

• Hear 1.4B first — a 1.2B listing suddenly looks cheap. Even though you never once examined that unit's actual condition.
• Hear 0.9B first — that same 1.2B listing now looks expensive. The home didn't change; only your reference point did.

Here's the part that genuinely unsettles me: the first number becomes the zero on your ruler. Set side by side, 0.9B and 1.4B are most likely separate records differing in floor, orientation, renovation, and timing. A one-line price strips all that context out — and just remembering that one fact is enough to reset where you start from.

Two common myths, and one question

Myth 1: "It's the most recent sale, so it's the most accurate benchmark." Recency does narrow the timing gap, true. But whether that single sale is representative is a wholly separate question. It could be a distressed sale, a related-party deal, or a home with an unusually good floor and orientation.

Myth 2: "Average a few asking prices and that's market value." An ask is what a seller hopes to get, not what closed. This is where everyone gets fooled. An average of asks tends to sit above actual transactions.

Q. So is it better to ignore anchors entirely? No. The reference point itself isn't the problem — clinging to just one is. Lay two or three numbers from different sources side by side, and the pull of any single figure visibly weakens.