Many numismatists hold auctions, and the prices realized therefrom in very high regard, nearly akin to Holy Scripture. These people live in a world where the auctioneer, a paragon of virtue and honesty, presides over an open competition where all participants have complete knowledge and make their offers based upon wholly rational preferences. The resultant prices, according to the high priests of the auction cult, thus reflect "true" or "correct" market price for numismatic collectibles.
People with knowledge concerning the operation of auctions will realize this Polyanna fantasy for what it is. Like the examples in many modern economic textbooks, this theoretical version is based on several very strong assumptions. Casual observation suggest these assumptions are too strong to be tenable.
People often assume auction bidding to be handled by calm, cool, calculating professionals, but this is not always the case. While professional dealers usually make up a large portion of auction bidders, collectors also participate, and while professional dealers can (usually) decide when to let an item go, collectors sometimes cannot. Collector buyers occasionally contract "gotta-have-itis", pushing the price up to (what would seem to most observers) ridiculous levels. Envy, ego, and competition amongst would-be owners can also play a part in producing distorted (high) prices.
Prices can and do face distortions from auction manipulation. A common form of auction hi-jinx involves collusion of one sort or another. It is not unheard of for a group of buyers to get together before the auction to decide who will bid on what, thereby reducing competition and prices. Collusion can also work to increase prices. It's possible for a dishonest auctioneer to submit phony mail or "book" bids to jack up prices. It's also possible to hire a shill or house bidder to run up prices on the floor. The auctioneer, using pre-arranged signals, surreptitiously instructs the shill to bid, in the hope of raising prices. Shills, while illegal in most jurisdictions, are often employed by unethical auction houses to stimulate increased bidding activity. Sometimes shills can be exposed: Does a bidder always seem to be bidding, but never wins anything? Does the bidder often participate early in lots, then drop out once things get rolling?
The best defense against a suspected shill is to simply stop bidding and let the shill win the lot. Shills are usually employed to push up the prices of items consigned (owned) by the house itself. Auctioneers do not make money buying their own merchandise, so shills who get stuck with too many purchases will not receive further bid signals.
Ultimately, the single biggest factor in auction price distortion is knowledge, or the lack thereof. Potential bidders include those who are very knowledgeable, those with some knowledge, and those with little knowledge of numismatics and/or numismatic pricing. Lack of knowledge may be due to poor communication, namely bidders not knowing exactly what is up for auction. Misattribution (the item is not what the description says it is), misgrading (either over or under), inadequate condition information (either over or under-emphasis of flaws, eye appeal, etc.) all can be factors in the final price. Bidders not viewing the items for sale can be at a distinct disadvantage. Add to this unethical auctioneers who puff their own merchandise by over-grading or slanting lot descriptions in their favor and the information situation becomes very cloudy indeed.
Clearly, the requirements needed to produce "true" market prices, or prices which exactly reflect the quality of merchandise offered for sale, cannot be met. This news will not make any friends with auction price book publishers. However, it will hopefully help to dispose of some of the myths surrounding numismatic auctions. These auctions are composed of people, with the same imperfections, the results of which are open to as much interpretation as there are points of view.