Snipers, Shills and Sharks

I recently read Ken Steiglitz's new book: Snipers, Shills and Sharks: eBay and Human Behavior.

Auctions come in many forms: English, Dutch, Japanese and Reverse and variants of these. They have been used since at least 500 BC. But the real quantitative understanding of auction science has come from online auctions, and especially the explosion of eBay in the last dozen years. This has connected more buyers to more sellers than every before. Its also constructed a test bed for understanding the science, both quantitative and psychological, that drives auctions.

Steiglitz, a computer science at Princeton, has written an excellent overview of auction science. The first 180 pages are completely non-mathematical. The appendix that follows contains the math details if necessary. There he does a good job of doing the proofs, which are usually of fairly idealized situations. He starts with the history of auctions, and then follows with a careful description of the different types that exist. Also a number of sometimes surprising revelations about the specific dynamics of eBay auctions, and how that influences buyers and sellers.

He then describes examples of consumer behavior on eBay, tracking how people bid and the resulting outcomes. Most of his examples come from ancient coin categories. These outcomes are driven by participants sometimes weak understanding of the technology, their own internal economic models, and the psychology of the bidding process. Not all is rational here! Along the way you do learn some remarkable things about how eBay works.

This book is not meant to be an introduction or guide to doing business on eBay, a quick search on Amazon shows dozens of such books, from the 'dummy' books to thick works on the best ways to make money there. This book, as its title suggests, is about how human behavior is represented in auctions. Along the way, Steiglitz suggests that a good way to study auction dynamics is using agent models. I think he is right ... this is not about the idealized 'Homo Economicus', or his student volunteers, but rather about rather irrational beings who display all kinds of interest, including emotions like 'spite' when they bid at an auction.

He uses his classes to experiment with these kinds of auction behavior. That limits the breadth of what can be experimented with. I wish he would he extended his interest in agent models to show some examples to prompt more explorations in this area. Good book for those who want an introduction to the more technical aspects of auction science.