How to Read an Auction Result
By Antonio Lopes · · 11 min read
Most people who look up an old book find a number and stop there. That number is usually the wrong one, because the first thing you see online is almost always someone's asking price.
The results on this site are different, and the difference is the whole point. Every lot in the database is a completed sale — a price somebody actually paid, on a specific day, for a specific copy. There are 592,936 priced book, map and print lots in there once non-book categories are stripped out, and none of them is a hopeful listing.
That distinction is worth learning properly, because once you can read a page of comparable sales you can value almost anything without asking anyone.
Sold price beats asking price, every time
An asking price tells you what one seller wants. It survives forever whether or not anyone pays it. A marketplace can show you a copy at €600 for three years running, and the €600 has never once been tested.
A sold price tells you what the market did. It is the point where a real buyer stopped bidding, in a room or on a screen, against other people who also wanted the book.
When you search a title, you are reading the second kind of number. Treat any asking price you find elsewhere as a ceiling that has never been proved, and treat these results as the floor of reality.
One high sale is almost never the answer
Here is the trap. You search a title, you see eleven results, ten are between €40 and €90 and one is €700 — and the €700 is the one your brain keeps.
It should be the one you discard first. The book market is enormously bottom-heavy. Across all 592,936 priced lots, the typical sale is €60 and the average is €118 — the average is nearly double the typical because a few thousand very expensive lots drag it upwards. Almost seven in ten sales land under €100.
So when you look at a page of comparables, ignore the top result and ask a different question: what did the middle copy do? Line the prices up smallest to largest and take the one in the middle. That is your working figure. The single highest sale is a fact about one copy on one day, not about your book.
A worked example: Anne Frank, Het Achterhuis
The database holds 62 completed sales of Anne Frank's Het Achterhuis across its early Dutch printings. It is a perfect teaching case, because the spread looks terrifying until you sort it.
The lowest sale was €20. The highest was €7,500. If you had found only the top one you would conclude you were sitting on a small fortune.
Sort all 62 and the picture changes completely. The middle copy sold for €160. Sixteen of the 62 sold under €100. Thirty-eight — well over half — sold under €200. The average across the set is €680, and that average describes almost none of the copies; it exists because of the handful at the top.
Now the useful part. Those eleven top results are not random. Eight of them are lots whose description explicitly says eerste druk, 1e druk or first edition — the 1947 Contact printing. Those eight have a middle price of €1,825. The other 54 lots have a middle price of €150.
The two groups are, in every practical sense, different books wearing the same title. If your copy is a 1948 third printing, the €7,500 result is not your comparable; the €81 typical for 1948-dated copies is much closer. Sorting the pile is what turned a useless €20-to-€7,500 range into two usable numbers.
That is the whole method. Find the group your copy actually belongs to, then read the middle of that group.
What the match percentage is telling you
Every result on this site carries a match score. It is not a quality rating and it has nothing to do with price. It measures one thing: how closely that lot's description resembles what you typed.
The search works on meaning rather than keywords, so a query for a Dutch cookery book will find Dutch cookery books even when the words differ. The trade-off is that it always returns its nearest neighbours, even when the nearest neighbour is not very near. A deliberately obscure query still fills the page.
So read the score before you read the price:
- High match, sensible title. A genuine comparable. Use it.
- Low match, high price. The most dangerous line on the page. A weakly matched lot at €900 is not evidence about your book; it is evidence that the search ran out of close options.
- High match, but the title mentions something yours lacks — a signature, a portfolio of plates, an extra volume — it is a comparable for a different object.
The practical rule: build your estimate only from results whose titles you could honestly describe as the same thing as mine. Usually that is five or six lines out of a page of ten, and searching a second time with a tighter phrase gets you a better page than squinting at a bad one.
Read the date, then discount it slightly
Every lot carries the date it was recorded. The database runs from 29 May 2023 to 31 July 2024 — a fourteen-month window, and a lopsided one: roughly two thirds of the sales were captured in May and June 2023.
That matters in two directions.
It is still good evidence. The typical sale was €61 in the 2023 lots and €57 in the 2024 lots — a market that barely moved over the period covered. Book prices are not volatile the way equities or trading cards are; a mid-range title that made €70 in mid-2023 is very unlikely to be a €300 book now.
But it is a record, not a quote. Auction results tell you where a book sat when it sold. Taste moves, an author dies, a film adaptation lands, a private collection floods the market with twelve copies of something previously scarce. If a title has been in the news since 2024, treat these figures as the baseline and expect the current market to sit above it.
Condition is invisible in the price — and it explains most of the gap
This is the single most common misreading. Two lines on your results page show the same title and the same year at €40 and at €400, and it looks like one seller got lucky.
They almost never differ by luck. They differ by condition, by printing, or by what was physically present in the lot.
The database has a clean illustration. Two copies of Kenneth Grahame's The Wind in the Willows, both dated 1922, sold in the same period: one for €94, one for €1,600. Same title, same year, seventeen times apart. A 1908 copy — the first year of publication — made €750. Across all 36 Wind in the Willows lots in the data the typical sale is €45, so both of those figures are outliers in opposite directions — and nothing in the price list on its own tells you why.
What you cannot see in a price list, and must supply yourself:
- The jacket. For a twentieth-century book the original dust jacket is frequently the majority of the value. A jacketed copy and an unjacketed copy are not the same lot.
- Which printing. A first printing and a fourth printing look identical on a shelf. This is where checking the number line and copyright page pays for itself, and it is also why the words "first edition" in a listing are worth so little — pooled across the whole database they add only about eight per cent, because anyone can type them.
- What else was in the lot. Sets, extra plates, a signature, loose ephemera. A signed copy carries a real premium — 1.63× within the same author — so a signed comparable is not a comparable for your unsigned copy.
- Damage. Foxing, damp, cracked hinges, a supplied plate, a library stamp. Any one of these can take a book from the top of its range to the bottom.
When you see a wide gap between two apparently identical lots, assume the difference is real and go looking for it in your own copy. It is nearly always findable.
Currency and fees: the number is not what changes hands
Two honest caveats about the figures themselves.
They are in euros. The database is drawn from European auction activity and is heavily Dutch, Belgian, French, German and Italian. If you are pricing in pounds or dollars, convert — and be aware that a book with a strong local following in one country can be worth noticeably less outside it.
They are the sale price as recorded, not the money that moved. At auction the buyer normally pays the hammer price plus a buyer's premium, and often VAT on that premium and shipping on top. The seller normally receives the hammer price minus a seller's commission, and sometimes photography, insurance or lotting charges as well.
Both figures vary by house, by country, by lot value and by whether you are a trade or private seller, so I am not going to invent a percentage — anyone quoting you one flat number is guessing. The principle is what you need: a buyer pays more than the number, a seller receives less than it, and the gap between the two is not small. A €200 result does not put €200 in your pocket.
The five-minute routine
- Search the fullest description you have — author, title, and year if you know it — on the search page.
- Throw away every line whose match score or title tells you it is a different object.
- Sort what is left by price and take the middle one. Ignore the top result.
- Ask what separates the top of that range from the bottom: printing, jacket, signature, completeness. Then find which end your copy belongs to.
- Adjust for fees in the right direction, depending on whether you are buying or selling.
That is genuinely all of it. If the answer comes back under €100, that is not a failure of method — most books are worth less than people hope, and knowing it in five minutes is worth more than a fortnight of hoping. If it comes back over €500, you are in the top three per cent of the market and it is worth running the full buyer's checklist before you do anything else.
