Why sold prices beat asking prices

Asking prices anchor wrong. Here’s why eBay’s Sold filter is the only honest signal — and how to read median sold like a pro.

← All articles · April 27, 2026 · 3 min read

What an asking price actually is

An asking price is a wish. Someone listed their thing, named a number they hope a buyer will pay, and now the listing exists. Most listings expire unsold. The asking price never tested itself against a real buyer with a real wallet, so it carries no information about what the item is worth. It only carries information about what the seller is willing to dream about.

If you anchor on asking prices — mentally averaging the three top listings on the search page — you’re anchoring on three sellers’ optimism. The real market sits somewhere lower, and you don’t know how much lower until you check what actually sold.

Why median sold is the truth

Every entry in eBay’s Sold filter is a real exchange. Money moved. The buyer agreed the price was fair and the seller agreed too. That’s the only definition of value that matters when you’re deciding what to bid.

Use the median, not the mean. The mean is one $400 outlier away from being misleading; the median is the price the middle transaction landed at, and the middle transaction is the most representative one. A used Roomba with median sold $80 is worth $80, full stop. If you can buy it all-in for $50 it’s a winner. If you have to bid past $65 it isn’t.

Reading the spread

The High and Low values bracket the median. Pay attention to how far apart they are.

• Wide spread (High is more than 2x Low): the SKU has condition variance. Same model number, very different real-world condition. Trust the median, not the high. Bid as if you’ll receive the worse version, because that’s the modal outcome.
• Tight spread (High is within 30% of Low): the market is liquid and confident. Your bid range should be tight too — there’s no upside in stretching toward the high.

The trap of recent listings

eBay’s default sort surfaces recent. That’s usually wrong for research. You want representative, not recent. Two filters fix this:

• Filter to Sold (last 30 days). Anything older is a different market — supply has moved, demand has moved, and you’ll mis-price.
• Sort by Date sold, descending. You want the latest examples within the 30-day window, not the highest or lowest of all time.

If the SKU is thin and 30 days isn’t enough comps to draw a median, expand to 60 days but discount each older transaction mentally. If you have to go past 90 days to find anything, you’re in territory where the comps don’t support a confident bid — treat it as exploratory and bid low.

ThriftEdge — our free Chrome extension — pulls real eBay sold comps onto every mac.bid lot page automatically. See how it works.