We will find out, that ebay is also running all this sniping companies under their flag :-), to make another 1-2 billion renvenue

:

:

:

:

:

: On eBay, it pays to snipe

: Posted 6/25/2006 6:21 PM ET

: E-mail | Save | Print Thank
: goodness for science. How else would we know
: the best way to nab those barely-used weed
: whackers, dumbbells or duck-shape salt
: shakers on eBay? In a study that gives the
: lie to the notion that eggheads don't like
: to eyeball online auctions like normal
: folks, a study by South Korean physicists
: confirms via some elaborate mathematical
: modeling that "sniping" — waiting
: for the very last second to submit your bid
: on that Elvis-shape throw rug — is indeed
: "a rational and effective strategy to
: win in an eBay auction."

: Founded in 1995, eBay is the king of online
: auction sites. Sellers put up items for sale
: and buyers bid up the price. Thanks to the
: Internet's lack of state sales tax and the
: public's thirst for other people's garage
: sale items, the company has grown into a
: firm that amassed $4.55 billion in revenue
: last year. The service sets a deadline on
: bids for items, which has given rise to the
: practice of "sniping," bidding at
: the last minute to deny other bidders time
: to outbid you.

: Savvy buyers have taken to the practice in
: swarms. Some companies even exist to snipe
: for you. Sellers, however, have grumbled
: that the practice keeps winning bid prices
: lower than they would be in a more
: open-ended auction, in which prices may be
: driven up by competition between buyers. If
: nobody bids until the last second, it's
: inevitably just a (relatively) low-bidding
: person who puts in the highest-price bid and
: walks away with the item.

: To test whether sniping is a smart way to do
: things or just truncates normal bidding, the
: South Korean team at Seoul National
: University produced a "master
: equation" for how bidding proceeds
: (it's nk(t+1) — nk(t) = w(k-1)(t)*n(k-1)(t)
: — wk(t)*nk(t) + sigma(k,1)*u(t), if you
: really want to know), and then tested it
: against a massive number of auction records,
: some 264,073 items sold in one day on eBay
: and another 287,018 items sold in one year
: by eBay's Korean partner.

: Plugging all those data into the model and
: testing the outcome in terms of how the
: auctions turned out, the team found that the
: probability of submitting a winning bid on
: an item indeed drops with each bid.
: "Our analysis explicitly shows that the
: winning strategy is to bid at the last
: moment as the first attempt rather than
: incremental bidding from the start."
: The study appears in the current Physical
: Review E journal.

: The finding is no surprise to Harvard
: economist Alvin Roth, who has studied
: sniping from an economics viewpoint since
: 2002 with colleague Axel Ockenfels of
: Germany's University of Cologne. They came
: to similar conclusions. "I think you
: might do the most good if you advise bidders
: to form an opinion of how much they are
: willing to pay for an item, so that they
: don't get caught up in a bidding war and pay
: more than they will be happy with,"
: says Roth, by e-mail. "But, that being
: said, if they know what proxy bid they want
: to submit, it won't hurt them to submit it
: very near the end (but neither will it help
: them much, or often ...) So, sniping is a
: good strategy, for those with the time to do
: it," he adds.

: A statement on the eBay site says: Sniping is
: part of the eBay experience, and all bids
: placed before a listing ends are valid —
: even if they're placed one second before the
: listing ends.

: BONUS MATERIAL: Dan Vergano's Q&A with Alvin
: Roth and Axel Ockenfels

: 1. Do you view sniping as a problem? Some
: eBay sellers have complained that sniping
: works to artificially lower auction prices.
: What is your view?

: Ockenfels: Sniping can help bidders to get
: better prices on eBay. But sellers too can
: profit from sniping, because the possibility
: of sniping may attract more bidders. For
: instance, sniping can lead to more bidding
: from experts, because by bidding late,
: experts can avoid giving information to
: others through their own early bids. Sniping
: can also increase the excitement and
: entertainment value of bidding, which again
: attracts more bidders.

: Roth: Sniping is a feature of the auction
: that eBay bought into when it chose to have
: a hard close. They must think it adds enough
: to the auction, in entertainment value, in
: allowing experts to protect their
: information, etc. to make up in increased
: bidders what it loses in lost bids and
: bidding wars.

: 2. If your work and this South Korean paper
: show that sniping is rational and effective,
: why doesn't everyone use the strategy?

: Ockenfels: On eBay, not the last bid but the
: highest bid wins. Furthermore, last-minute
: bids sometimes come in too late, after the
: close of the auction. So, it can be a
: perfectly sensible strategy to submit a bid
: early. In fact, depending on the situation,
: game theory supports both early and late
: bidding strategies. However, we are also
: seeing a lot of non-rational, naive
: behaviors on markets such as eBay. There is
: no reason to suppose that everybody always
: behaves in a rational and effective way.
: This is especially true for eBay, where many
: experienced and sophisticated traders
: interact with many unexperienced, naive
: bidders.

: Roth: EBay isn't an English auction, it is a
: second price auction with proxy bids. If
: you're a busy guy, you might find it better
: to put in an early proxy bid, high enough to
: have a chance of winning. The winning bidder
: isn't the last bidder, it's the bidder with
: the highest proxy bid (and the earlier
: bidder in case of ties). So sniped bids only
: get lower prices when other bidders would
: have been willing to raise their proxy bids,
: but don't have the chance. That happens
: often enough so that sniping is a good
: strategy for those with the time ....

: 3. How does this new study's approach strike
: you compared to the one you published in
: 2002? My understanding was that it rested on
: game theory, so I'm just trying to see how
: you see things.

: Ockenfels: We closely intertwine game
: theoretical, laboratory and field analyses.
: Taken together, our studies help in
: understanding how the market microstructure
: qualitatively influence participants'
: strategies and overall market performance.
: The new study looks at bidding phenomena
: from a very different perspective and thus
: takes a very different approach. It
: quantitatively analyzes statistical
: properties of dynamic bidding patterns on
: eBay — without addressing institutional
: complexities or equilibrium aspects of
: behavior.

: Roth: We take a lot of approaches,
: empirical, theoretical, experimental. And so
: we are able to look into the multiple causes
: of sniping, and how they are influenced by
: the auction rules (and why, therefore,
: there's so much less late bidding on other
: kinds of auctions, for example.) But the big
: divide between physics and economics is that
: physicists tend to study processes that
: don't have any human volition in them.
: Molecules do what they do without forming
: opinions about what other molecules do.
: Sometimes this physics approach can also
: yield some insights into large markets,
: where each player is small enough to be
: inconsequential. And eBay must have looked
: that way to the authors of this article,
: since they report that in one day they have
: data from 264,073 auctions involving 384,058
: distinct bidders. On the other hand, when I
: look at those numbers, what strikes me is
: that there were fewer than 2 bidders per
: auction in their data. To put it another
: way, a lot of auctions in their dataset had
: only a single bidder. Obviously conclusions
: about sniping are going to be different in
: such auctions (and in our analyses we
: normally exclude them).

: 4. Do you see a better auction strategy for
: online auction?

: Ockenfels: What is a good bidding or selling
: strategy in online auctions depends on the
: context, such as the degree of competition
: and the available information about the
: value of the object. However, there is a
: fast-growing applied literature in economics
: on auction/market design and bidding
: strategies. (See Roth:
: http://kuznets.fas.harvard.edu/~aroth/papers/engineer.pdf
: .)

: Roth: I think you might do the most good if
: you advise bidders to form an opinion of how
: much they are willing to pay for an item, so
: that they don't get caught up in a bidding
: war and pay more than they will be happy
: with. But, that being said, if they know
: what proxy bid they want to submit, it won't
: hurt them to submit it very near the end
: (but neither will it help them much, or
: often&) So, sniping is a good strategy, for
: those with the time to do it. (You can also
: pay a fee to third party sniping software on
: the web, like esnipe.com or others ....)

: 5. What do you see as the key point(s) to
: make to readers about a study like this one?
: How do the results apply to other auctions?

: One of the general lessons that comes out of
: our research in "economic
: engineering" is: details matter! For
: instance, our studies demonstrate that
: replacing eBay's hard close by a soft close,
: which allows bidders to always respond to
: late bids, would remove the strategic
: incentives to snipe and thus substantially
: affect bidding behavior. Bidders respond to
: incentives, and incentives can be strongly
: affected by the details of the auction rules
: and algorithms. This is true for all
: auctions, including, for instance, spectrum,
: electricity and procurement auctions.

: Each week, USA TODAY's Dan Vergano combs
: scholarly journals to present the Science
: Snapshot, a brief summary of some of the
: latest findings in scientific research. For
: past articles, visit this index page .

:

: Posted 6/25/2006 6:21 PM ET

: E-mail | Save | Print

:

:

: Link to original article on USAToday follows...

: -- Chuck