Todd is a local of Central New York and will soon be attending Cornell as a non-traditional undergraduate. Within the past year, he has enthusiastically become a member of the greater eBay community. His combination of life experiences as well as his new favorite past-time of buying and selling on eBay made him an excellent choice for an interview as we can see how his story and many others create the unique online network of eBay.

As we know, eBay is an online auction site. Even with this simple definition, we can already see how the eBay phenomenon fits within many of the course’s topics. Over the past semester we have discussed online communities as well as auction mechanisms and as my interview will show, many buyers and sellers on eBay demonstrate the near magical strength between weak links.

Weak Links

Todd began listing items on eBay at the urging of one of his friends. Orignally living with a group of friends in a pair of apartments, he found that many simply abandoned their things when their leases expired. From his original set of sheets, he has made over two hundred sales with items that people simply abandon or items people sell through him on comission. In the Network context, Todd acts as a contact point for people with little time or knowledge of e-Bay. Todd finds many of his items in basements he is sometimes hired to clean and offers to sell the item 50/50. In Todd’s own words, he’s happy to take whatever cut he can because for him, its simply “Free Money.” Just as in the real world, Todd also takes the intiative to contact buyers who appear to like his merchandise and thus, wider communities form between those who buy and sell similar items. A link is created through him between those without the time or energy to understand eBay and those who do.

Market Mechanisms

Another exciting aspect of buying and selling on eBay is the remarkable efficiency and price stability of similar items on eBay. A quick look at the listings of say, iPods, show a general price range where differences in age and condition is taken into account. Todd has to be careful about the pricing of his merchandise, down to the shipping and handling costs because the fees that eBay applies to its sellers can add up. It is not free to list on eBay. Luckily, Todd uses a popular service that eBay offers that consolidated information about past sales. This service powerful tools for sellers to keep track of current market prices, which in effect makes the market more efficient.

Todd has also observed other interesting “economic” activity. For instance, it is often to see the prices of auctions rise rapidly right before an auction closes. Bidding wars are common for select products and often unexpected products (like Aviation Magazines and old Board Games) and it often takes a few initial bids before more bidders are attracted. The last phenomena may be explained with information cascades: that only items desired are items worth bidding on.

Reputation

Finally, an interesting point made in class discussed the ownership of an eBay member’s reputation. Todd highly values his reputation and so far has a perfect score out of 138 responses. “With so many unknowns, the system is heavily based on feedback,” Todd explains. Quite simply, the number of feedbacks represent the likelihood that a transfer is to succeed. This has quite a host of game theory applications as probability of success can be factored in the strategy a player chooses. More advanced game theory explains behavior based on past experiences, such as the past experiences of a potential seller.

The quantity as well as quality of feedback also has wide applications as it is safe to assume that the seller has completed twice as many sales. “So what if a seller has a perfect score of 4, he only did 4 things right,” Todd emphasises. While Buyers can receive feedback, the reputation of a seller is his most prized asset. At times, Todd is more than willing to take a loss just to satisfyg his customers.  If we associate a cost or a certain risk related to one’s confidence in a transaction as part of the framework of a game, I’m sure we could prove that the cost of a bad reputation is higher in the long run.

Thus, Todd is very happy with his hobby. Todd feels like he is part of a community of buyers and sellers and enjoys the challenges of getting things sold.  Todd feels familiar with many of the topics we’ve discussed in our course because he lives networks every day with eBay. He also finds a joy finding “free money” on things that people either abandoned or allow him to sell. I should know first hand, Todd sold a pair of my older mint condition 56k Tivo compatible RCA Wireless Modems for $20 each!

Earlier this evening, I navigated to the popular digg.com website earlier to see what was buzzing around the Internet tonight. I was quite surprised to find that there were few, if any, legitimate stories on the front page. Every story was a posting of the (fairly) recently discovered key to “crack” the encryption of the AACS Digital Rights Management content protection system. It’s a string of 16 hexidecimal numbers, a total of 128 binary digits. I won’t publish them here, as it is a criminal act, or at least something that will get you served with a DMCA takedown notice and threat to sue.

The point here is not the key itself or even that people can use it for piracy now. What is interesting is the huge buzz around the key. The lack of any legitimate digg stories rising to the top, the edit-locked Wikipedia articles, and the mass of blog posts suddenly containing nothing but the key are the interesting part. While the crack was discovered back in February, there were no real problems until today. Suddenly, it’s all anyone can talk about (see this digg story). What was the tipping point in this story?

To me, it appears to be one or two stories that were linked to on the Digg page earlier today (here and here) that deal with censorship. Digg users have risen up before about censorship, so it seems like it’s a rather important issue to them. Soon after, many stories were submitted to Digg and apparently deleted as well (see here). This was the tipping point. Once possible, users submitted as many stories as possible, and dozens of them populated and filled the front page in barely any time at all.

Who knows what the page will look like tomorrow morning; perhaps they’ll all be deleted, or perhaps it will fade into obscurity. In any case, for a short time, this was all anyone on the Internet could talk about, and this showcases the effects of bandwagons. So few of the users could really care about the code; they won’t be the ones to use it, as it does not do anything for a layperson. As one of the linked Digg stories says, “the processing key is only useful for sophisticated programmers capable of writing software to rip and decrypt the discs, much like how a blue-print drawing of a door key is not sufficient to unlock a door without an engineer first making a physical key based on it.” Yet for something that seemingly should be important to every digger, they all join the cause, be it illegal or not.

It’s just another example of how bandwagons can get out of control, and end up as bad as a mob.

http://www.infozech.com

Infozech, an India based company, is a company pioneering the field of telecom billing. This means that it provides ways for companies that provide telecommunication services to bill and manage their customer networks. Large companies (most frequently in India or nearby countries) consult with Infozech to find unique solutions to satisfying every customer. The cases that Infozech confronts are very interesting in the context of our course because multiple graphs of networks can be created for a population of customers. Simply examining the aspect of communication between nodes in the graph, each customer communicates only with the telecommunication company (aka telco). So an edge would be created from every customer to the telco. By means of this connection, each customer simply takes care of managing their subscription to the service.

But obviously, Infozech is concerned with affairs from the point of view of the telco. Even though from a communicative standpoint there are no edges connecting any customer to any other customer, there are numerous network effects that Infozech has to deal with (on behalf of the telco) that are a result of geography and preexisting telecommunication infrastructure. For example, if a telco is a provider of a suite of wireless phone services, Infozech must write a program that can handle the individual situation of every single customer. This means that if the telco provides services nationwide, but telecommunication infrastructure in different parts of the country result in different charges, Infozech must account for these. Infozech’s software can track each individual phone call or communication of any sort that occurs between any two customers on behalf of the telco. It can then analyze each call and determine what type and what quantity of charges were incurred to the telcos customers.

Another challenge that the company must deal with that is less related but does require note is that they must integrate their software into any framework the telco already has. Judging from the information at their website, they are always able to meld their services such that customers can track their own usage of the telco’s service in real time with seamless integration.

 http://www.bioone.org/perlserv/?request=get-document&doi=10.1666%2F05008.1#I0094-8373-32-1-1-F01

Species extinction is a big topic of discussion these days. Whether global warming, habitat destruction, or other various sources are being pinned as the culprits, loss of biodiversity is indeed a serious issue that we are facing. However, what are the mechanisms driving this extinction cascade once a disturbance occurs and how do ecosystem parameters contribute to profiles of extinction throughout a food chain? The cited article attempts to create a testable model of species extinction in mass extinction events, using species richness, functional diversity, and trophic link distributions as parameters. Of course, defining these aspects in an ecosystem is a tough task in itself. Recent blog posts have cited papers addressing the issue of how interconnected some food chains actually are by defining mean distance between nodes.

Most mass extinctions do not involve direct destruction of species, but instead a bottom-up cascade of secondary extinctions as primary production is disrupted. However, one’s initial instinct to carry a linear relationship through the food chain through cause and effect relationships is not completely accurate. Assuming reasonable behavior, one would expect consumers to alter their interactions with remaining prey groups. In the presence of these conditions, which indeed is included in the model, there is a secondary, top-down stress on mid-level prey taxa. This can be easily seen in that if a predator feeds off a variety of prey, and one of the prey groups is reduced in number, the other groups increasingly become a source of food.

This model demonstrates that is impossible to know how a disturbance in primary production can effect a particular species without a nearly complete knowledge of their place in the food network. Species diversity, one would assume, would increase the stability of the system and its ability to hold up against mass extinction simply due to the fact that potential changes and concentration of food sources has less of an effect on each group if there are more of them. Thus, current species loss, especially at the extremely high levels witnessed in today’s world, serves not only to directly destroy current ecosystems, but make them much more vulnerable to potentially catastrophic disturbances.

With the advent of the internet, the world seems infinitely smaller than it has ever been in the past. When the experiment was performed, doubtless six degrees of separation seemed infinitely short, unbelievable in a world with six billion people. As times have changed, however, this fact seems less startling. In a world where I can now find my own public high school records with a quick Google search, it doesn’t seem so far-fetched at all.

Put into terms of finding ‘any random person’, things begin to find new meaning. If you were to give me the name and address of someone in the US to make contact with, I certainly wouldn’t try to decide who of my closest friends to contact, or try to go through any job that I might have. Use of Google in conjunction with someones residence and/or career makes it quite likely that one can find the person on the first page of a search. If they aren’t, it is a simple step to find a well-connected person in their place of residence whom you can pass on the request to.

Put into terms of the Watts-Strogatz model, the internet magnifies the value of ‘k’, our random connections, immensely. We are ‘randomly’ have the capacity to find literally millions and millions of people, instead of the few hundreds whom we are connected to otherwise. In a similar vein, with the scales of resolution model, the model changes in the sense that while our local circle remains small, the outer circle becomes massive, leading to fewer ‘changes of resolution’ necessary to find a person.

Its not hard to envision that the average, middle-income person is connected to another within a degree of connectivity, or perhaps two, when anyone can find anyone else in their own town/city/block, and its easily possible find someone from their town/city block.

Its a small world.

http://www.winsupersite.com/showcase/zunestory.asp

www.zune.net

Late last year, Microsoft launched its new mp3 player, “Zune”, with an interesting catchphrase - “Welcome To The Social”. Its new, killer-app-wannabe technology was a system to interact wirelessly (using 802.11b) with other Zune users to download and listen to music that they had on their own Zunes.

Unfortunately for Microsoft, they’re fighting a hugely uphill battle against Apple’s iPod, attempting to defeat a network effect (iPod’s are cool) that has so far defeated all other comers (Apple’s marketshare has remained sky-high, despite numerous attempts by other manufacturers to break into the mp3 player market). And to do so, Microsoft is relying on its own own network effect (I can share my music with my friends).

Though very much a long shot, the idea is quite simple. A few friends get Zunes, which can share music with each other wirelessly. They benefit from this sharing, and others perceive the benefit, so they decide to get Zunes. This repeats, and soon Zune has replaced the iPod as the market leader.

Of course, we don’t see this happening right now, for a handful of reasons.

1. Wireless sharing isn’t as compelling as Microsoft hoped it would be. People haven’t been demanding this feature, and are quite content with their wireless-sharing-less iPods right now.
2. Wireless sharing isn’t as useful as it could be. Microsoft is restricting sharing to certain songs, and those can only be listened to (on the 2nd Zune) 3 times over 3 days.
3. The iPod is too strong, is only growing stronger, and has saturated the market. That is, Microsoft has to not only persuade new mp3 player users to buy Zunes, but to persuade current iPod users to switch (an ironic direction, isn’t it?).

So to “win”, Microsoft needs to either 1.) drastically improve its wireless feature (to take advantage of the network effect) or 2.) set itself apart from the iPod with other, superior features (which it already has some of). As the article linked to at the top concludes, TiVo was an insanely popular device that is in trouble now from commoditization and the introduction of other comparable devices. That’s what Microsoft needs to be hoping for.

The point here is that while it may be impossible for Microsoft to dislodge the Mac faithful from their iPods, that’s a very small group of people anyway. The iPod, ultimately, is like the TiVo. That product, too, was off the scale for customer satisfaction. But today, the DVR pioneer is struggling because the DVR market is becoming commoditized thanks to integrated DVR solutions offered by cable and satellite providers and PC-based solutions like Windows Media Center. See a trend anywhere? If it can happen to TiVo, it can happen to the iPod too.

The only problems here are the high popularity and penetration of iPods, and the network effects surrounding them. Sure, TiVo was cool, but it was hardly a fashion accessory.

A question one would like to think about is whether the small world phenomenon has been changed with the increase usage of the internet. With online communities, the ability to meet people who you have never actually physically met, does this make our world even smaller? Watts and colleagues at Columbia University (New York) conducted an email experiment to find out whether there was any effect [http://www.newscientist.com/article.ns?id=dn4037]. They found on average 5-7 emails had to be sent in order for the subject to be found. This was the sort of results that Milgran found out when he first proposed the experiment, so perhaps the world wide web does not affect the size of the world.

Thinking about online communities, it is true that we are able to meet more people than we were before. However, as we do not see each other face to face, it is possible that perhaps people do not tell the whole truth of who they are. Someone you were randomly chatting to on an online forum would not be someone you would think about sending the letter to in order to find the subject. I would argue that the world wide web’s impact is not so great as the impact of improved transportation systems around the world. Nowadays, it is possible for someone in China to be connected (through 5-7 steps) to someone in South America without the world wide web. This is just because of the transportation system that allows many people to travel globally. If this experiment were to be done before it was common for people to take airplanes, or even to take ships, the chance of someone in New York being connected to someone in Australia is minimal.

We recently spoke in class about information cascades and the impact of having all of your friends move to a new type of device. We used the example of two different instant messengers: A and B. Assuming all of your friends used A, we then evaluated the likelihood of people switching to B. This decision was predicated on the number of your other friends who also switched to B. There was some kind of a tipping point, at which everyone would move to B.

This was an interesting example, but it became even more interesting when we applied the same principle to a real example. We used Windows and Netscape. While Microsoft already had Internet Explorer, it had to decide how to balance its interest in making it difficult for Windows users to use Netscape with their vested interest in supporting their own product. If Microsoft made it impossible for Windows users to have Netscape as their internet provider, then they would run the risk of having Netscape loyalists switch to a different operating system.

Microsoft found the right balance and the Netscape lovers didn’t defect from Windows in high enough numbers to create any type of legitimate threat to Microsoft. But there is a new and very legitimate threat to Microsoft that has emerged as a serious player on the tech market. Apple has become instant messenger B from our in-class example. Windows was the dominant player in terms of computer operating systems for the better part of the past 15+ years and now is dealing with the impacts of a very real information cascade problem.

Apple has successfully marketed its Mac operating system as not only a more user friendly product, but as a cultural phenomenon. Mac users have tremendous disdain for PC’s and aren’t shy about sharing their opinions. They are so smitten with their hip machines that they actively recruit new members to join “the movement.”

Here is a great example of some of the talking points that Mac users love along with a simultaneous satirical parody created by a PC user:

http://www.youtube.com/watch?v=4Tq7yykR-DM

Well have they made a significant impact on the market? Has it been significant enough to scare Microsoft? Absolutely. That became crystal clear when Microsoft released its new Windows substitute called Vista. Vista is a Windows/Mac hybrid intended at becoming more user friendly like a Mac without changing settings so much so that traditional Windows users become too frustrated or confused. This is a clear and real indicator that Microsoft recognized there was a information cascade happening very quickly and that they needed to adapt to changing circumstances, before Mac was able to convert an even more substantial percentage of the market. However, it was at a huge risk. What if Windows users eventually decide that they don’t like Vista? Maybe they will think that Vista is so similar to the Mac operating system, that they may as well make the full switch to a computer, “void of viruses.” This is an interesting example of information cascades and it will be fun to track its progress over the next few quarters.

http://www.americanscientist.org/template/AssetDetail/assetid/45930;jsessionid=baagGcSyTPpJry?fulltext=true

According to Professor Paul Turner, “the temptation to cheat appears to be a universal fact of life. In the struggle to survive and reproduce that drives evolution, selfish individuals may be favored over cooperators because they are more energy efficient.” The simple fact that taking shortcuts means better survival is the basis for Turner’s research into evolutionary game theory which suggests that the more a virus “cheats” the better it grows; however, this evolutionary technique eventually reaches a dead-end. Although the immediate effects of this “cheating” are very favorable, it requires a great deal on other “cheating viruses” to participate in surviving. The example of paracites-on-paracites is made to help illustrate this fact. Turner claims that these pairs assist each other by neutralizing the effects of certain vaccines on each other but end up requiring them to spend even more energy to counteract against other vaccines otherwise it becomes impossible to survive independently.

Ultimately, we see the rise of a “prisoner’s dilemma” in this virus game where “cooperators/cheaters” act in a quid-pro-quo relationship in order to survive.

My high school friend was recently a contestant on NBC’s game show “Deal or No Deal.” As any responsible engineering-student friend would do, I asked her (after they already filmed her episode) if she researched the game theory involved. Her response was had something to do with makeup and interviews, and boiled down to “no”.

Searching for “Deal or No Deal Game Theory” yielded some interesting literature, one of which came from the Wallstreet Journal.

For those of you who have not had the pleasure to witness an episode of Deal or No Deal, the way it works is there are 26 closed briefcases, each holding a distinct dollar amount from $0.01 to $1,000,000. The contestant chooses a case which remains closed until the end of the game. He/she then opens the remaining briefcases, one by one, narrowing the possibilities of what could be in his/her chosen case. Periodically, “the banker” will make the contestant an offer, based on the expected value of the contestant’s case given the already-revealed values (”deal”), or continue to open cases (”no deal”).

The data from Deal or No Deal is more valuable than other game shows because of its simplicity. Contestants are making the choice between calculated odds and a guaranteed payoff. As the article states, “there’s no trivia. No vowels to buy, no wheels to spin.” The players’ decisions are directly motivated by their immediate payoffs.

Thierry Post, a professor of finance at Erasmus University in Rotterdam, the Netherlands, studied 53 episodes of Deal or No Deal and found evidence to support prospect theory as it relates to risk aversion. Findings which shed light on human behavior in high stakes decision making can be used to tweak financial portfolio strategies.

While this article only addresses the contestants’ behavior in the game, there is other literature which discusses the “banker’s” offer strategy, which is arguably just as interesting.