Orders & Medals Price Archive

Where prices come from

Three kinds of source feed the archive:

Auction houses and auction archives — realised prices for lots that sold, with the sale date.
Specialist dealers — fixed prices for items in stock, and items the dealer has marked as sold.
Whole-web discovery — every search also looks across the open web for the same award on sites without a dedicated connector. Finds pass the same checks as everything else and are added to the archive, so the next search for the same award is instant.

The current list, with counts and crawl dates, is on the sources page. General online marketplaces are not yet covered.

Two kinds of price, never mixed

Sold — a realised price. For an auction this is the hammer price (marked hammer, or hammer + premiumwhere the house publishes the price including the buyer’s premium). For a dealer it is the last asking price of an item marked sold (sold at asking); the actual negotiated price may have been lower.

Buy now — a fixed price on an item you could buy today. It is what a seller hopes for, not proof of what buyers pay.

Asking prices usually sit above realised prices, often well above. Averaging the two would produce a number that describes neither market, so every search shows two separate bands and every listing carries its badge. Estimates, opening bids, current bids and unsold lots are never treated as prices.

What is excluded, and why

Only single, original awardscontribute to a price. Each listing is classified when it enters the archive, from its title and description in any language. Listings that fail are kept aside — not shown as comparables, never counted in a band — and each search reports how many were set aside for it (“Excluded from these figures…”).

Excluded	Why	Typical wording
Replicas, copies, restrikes and fantasy pieces	Not the object being valued. A copy sells for a fraction of an original, and one in a comp set drags the median down.	replica, copy, reproduction, restrike, recast, museum copy, fantasy, “Nachfertigung”, “copie”, “réplica”
Lots and groups	A price for several items says nothing reliable about any one of them — a single group sale can move a median tenfold.	lot of 3, group of medals, pair, trio, “Konvolut”, “ensemble”
Accessories on their own	A case, box, ribbon, document or ribbon bar without the award is a different, much cheaper market.	case only, empty box, ribbon bar, award document
Miniatures	Genuine, but miniature dress pieces trade at a small fraction of the full-size award and would pull its price down.	miniature, “Miniatur”, “réduction”
Non-award militaria	Uniforms, helmets, photographs, books, coins and the like are not orders, decorations or medals.	postcard, photograph, uniform, cap badge, coin
Anything that is not a price	Estimates, opening and current bids, and lots that failed to sell are opinions or hopes, not transactions.	estimate €300–500, current bid, unsold, passed

When the rules cannot decide — a title that might describe a copy, or might not — the listing is held back as unverified rather than let through. Precision matters more than volume: one wrong comparable can distort a median; one missing comparable rarely does.

How the bands are computed

Matching.The query is read as a country, award, class and year (“Interpreted as”). Listings must agree on the words that identify the award; a different class or year of the same award — a 1914 and a 1939 Iron Cross, a Knight and a Grand Cross — is a different object and is not used as a comparable.
Currency. Prices are converted to US dollars so they can be compared. The original price and currency are always shown beside each listing.
Outliers. Prices far outside the rest of the set — typically a mispriced listing or an exceptional named or cased example — are trimmed before the band is calculated, so a single freak result cannot stretch it.
Median. The headline figure is the median: half of the comparables are below it, half above. Unlike an average, one very high or very low price barely moves it.
Typical range. The middle 50% of prices (from the 25th to the 75th percentile). A narrow range means the market agrees; a wide one usually means condition, maker or variant matter a lot for this award.
Confidence. High, moderate or low, from how many close comparables there are and how consistent they are. A tight range over three listings is not more certain than a wide one over thirty. When too few listings of the exact class exist and the band has to span several classes, confidence is never better than low, and the card says so.

Each card also shows the number of comparables and the span of their dates. The price-distribution strip below the cards plots every comparable individually, so you can see clusters and gaps that a single number hides.

Limitations

Condition, maker and variant are not modelled. A pristine cased example and a worn one can sit in the same band; read the listings.
Named awards, groups and provenance can multiply value. Because groups are excluded, the archive describes the market for the single, unattributed award.
Buyer’s premiumvaries by house (often 20–30%); hammer prices exclude it, so the buyer paid more. Each listing shows the price as published, but the sold band counts a hammer price at hammer plus the platform’s typical premium (20–25%), so it compares with dealer and premium-inclusive prices; the exact premium of a given sale may differ.
Dealer “sold” prices are the last asking price, not a confirmed transaction price.
Coverage is uneven. Some countries and awards are heavily traded online and well covered; rare awards may have only a handful of comparables, or none.
Classification is automated.It is deliberately strict, but it can still be wrong in either direction. Treat any single listing with the usual collector’s caution.

Disclaimer