http://www.forbes.com/free_forbes/2007/0507/124.html

The above article by Lyor Cohen, American Chairman and CEO of Recorded Music for Warner Music Group, details the elements necessary to be able to make rap superstars. Cohen didn’t always work for Warner Music Group, he got his start as a club promoter, eventually DMC’s road manager, manager of Rush Management, and eventually Def Jam Records. Through his experience he notes that “You need two networks to sell music: one to find the artists, the other to market them. If one of the networks doesn’t exist, you must invent it. In the case of rap music, a quarter of a century ago neither existed. So we built our own networks to discover artists and promote them.”

This presents an interesting scenario from a networks perspective, because what you have happening is one node who knows people in the marketing network and builds an artist finding network. As this node stands as a bridge between these networks, he can control the flow of information between them and exercises a lot of influence over the decisions of both networks. By strategically positioning himself in a gate-keeping role, Cohen set himself in a position that eventually led to his upward advancement in the music industry.

Cohen then transitions his discussion of hip-hop networks into the current state of modern hip-hop networks. He talks about how now that hip-hop is mainstream that finding artists is usually done through a network of lawyers, artist managers and other insiders. “They have the access to power, and they shop their artists constantly, setting up showcases and starting bidding wars between the music companies to sign their artists.” What is happening here is that there are middlemen standing as a local bridge and they are taking advantage of the very restricted network to make a lot of money by driving down earnings of artists and driving-up prices for record companies. With the artist discovery network becoming price saturated (price bidding wars driving profits to zero), there have been new independent networks popping up that perform the same function as the artist discovery networks. However, unlike the old networks these new networks create their own network of fans and supporters from scratch (so they haven’t had all of the profits sucked out by having many middlemen in a layered network) and if larger companies find these networks they stand to make a substantial profit by singing these record labels. Another advantage to using the new smaller networks over the larger established networks is that the smaller networks build a loyal fan-base, which payoff for more profits in the long run. “The A&R guys feel compelled to sign bands that resemble junk food. Their music goes down easily but doesn’t offer any lasting satisfaction. These bands (we all know the ones I’m talking about) may sell a lot in the short run, but they burn out quickly. That’s costly, because over the long run we lose the opportunity to cultivate the best audience a band can ever have: a loyal base of discerning fans who stick with an artist for decades and sell others on the music, too.”

www.pipelinepub.com/0207/pdf/Pipelinev3i8Ariticle3.pdf

Above is a link to an article published in Pipeline magazine (http://www.pipelinepub.com/) called “Avoiding Future Schlock: The Real Life Application Of ‘Game Theory’” by Alana Grelyak and Tim Young. It’s a pretty quick read, and the most important thing to know before you get into it is to not get intimidated or give up on it if you don’t recognize a bunch of the terms used in the body of the paper. There are a lot of insightful points in the article, and the take home themes in particular can be (and in some cases, are) applied to many sectors in business and economics. The article analyzes how game theory applies to you first from a micro perspective (”how can I play games”), then takes a more macro stance in predicting trends in certain business sectors and giving advice to anyone who plans to pursue a career where knowing the market is helpful.

A particularly interesting comment that Grelyak and Young make is in reference to the role that Game Theory plays in both evolutionary biology and modern business. “Each kind of ecology/environment and period of time had specific winning strategies, so to use this game

theory, you just need to match the biological game conditions to a comparative economic environment and period.” The theory behind this is that if you look at the history of organisms on this planet, accepting Darwin’s “Survival of the Fittest” notion, the battle between any two (or more) species for dominance and/or survival in an ecology is basically a game, and if you take all of the games played between species throughout history, the winning “team” generally follows a strategy (or set of strategies) that can also be applied to business trends. For example, Grelyak and Young talk about “following proven success” and how it has worked well both for evolution as well as business, but also make a point of knowing when it’s time for something new. Some other metaphors that the authors use to describe strategies include, “Joining a herd”, “Injuring the competition”, and “boxing in the water hole” (having a regional monopoly and hoarding a resource).

The authors then apply these thoughts to mostly technology-based trends in business and try to predict where the world will go based on the theory that the new strategy taken by businesses will be “out with the old… this is the year of…” It’s cool to realize that with at least some of their predictions or realizations, they seem mostly right. So it’s great that Grelyak and Young are providing us with an understanding of many different strategies that can be adopted in game theory, but how do we know which ones to choose when? This seems especially important when considering a game where you opponent(s) either have some information about how you’re going to play (and vice versa) and/or the game will be played multiple times - the answer is to pay a lot of attention to the factors and players that matter in the game. This includes your competition, customers, partners, etc. Say you’re a project manager for Apple, and you had the idea for the iPhone. You have to decide whether to go with this new, risky strategy that could either be a great success or a embarrassing failure. If you do some market research and get the opinions of potential customers, you’ll have a much better idea about how it’ll be received. Also, if you look at past successes and failures and those of other competitors, you’ll get a better sense of what needs to be done.

“Collaboration is the essential strategy to having all parts of the product definition, launch, operation and management environment in place to win the futures game”

Looking through the topics that we have covered thus far in lecture I became particularly interested in Game Theory. In my personal opinion, games govern most of our lives, whether directly or indirectly. Since this idea is present during most social interactions I thought it would be interesting to find an article about this topic. When looking for an article that related to this idea I was able to find one quite intriguing, Game theory and the Cuban missile crisis.

This article deals with the historic event of the Cuban Missile Crisis and how it can be viewed through the definitions of Game Theory. Although it would be kind of inaccurate to relate such an important even as a game, since a lot was at stake, Game theory and the Cuban missile crisis points out how it reduced to the basic ideas of games. It begins by stating the strategies that each superpower had (U.S. and the Soviet-Union) and their respective payoffs.

Soviet Union (S.U.)
		Withdrawal (W)		Maintenance (M)
United States (US)	Blockade (B)	Compromise (3,3)		Soviet victory, U.S. capitulation (1,4)
	Air strike (A)	“Dishonourable” U.S. action, Soviets thwarted (2,2)		“Honourable” U.S. action, Soviets thwarted (4,1)

As shown in the image above, the crisis behaved somehow as a game since each nation was looking to maximize their payoff. Furthermore, from the graph one can see the Nash Equilibria (Air Strike, Maintenance) & (Blockade, Maintenance) and the Social Welfare Maximizer (Blockade, Withdrawal). Looking at this we see that, as discussed in lecture, the Nash Equilibrium is not always the best set of strategies. In fact, the crisis was handled by choosing the Social Welfare Maximizer as the best set of strategies, but as the article shows, that was not that simple to choose. The articles continues to demonstrate that Compromise was chosen as the best set of strategies since both nations were taking into account not the particular best payoff of each nation but the overall best set of responses that maximized the sum of the individual payoffs. Moreover, the article follows with a discussion of the “Theory of Moves” to show that both nations were somewhat forced choose the best joint strategy since each nation had to take into account the possibility of the other nation playing a better strategy. To explain this one must notice that no nation had a particular best response, thus each nation had to choose their strategy based on what the other nation did. In such a case, as the article describes, the nations would choose their strategies based on the other nation’s strategy and continue to do this until they come to an agreement, which would occur once both nations agree on a set of strategies and notice that no other set of strategies would make them better off.

Such an article clearly connects to the definitions and ideas discussed in lecture. This particular article covers the topics of “Game Theory” and “Theory of Moves” through the example mentioned above, which views the Cuban Missile Crisis as a game. It shows, as in class, how in a game each player must take into account not only their payoffs but their opponent’s possible moves (since it is based on this that their payoff will be defined) and, additionally, how not having a best response can affect the outcomes and play of the game.

http://www.respondentdrivensampling.org/

http://www.migrationinformation.org/USfocus/display.cfm?ID=441

Respondent Driven Sampling (RDS) is a technique used by sociologists and other researchers to study hidden populations.  It is especially hard to identify and reach such populations in order to conduct research on them because data assigning people into these demographics is not readily available.  Some examples of hidden populations include second generation immigrants, injection drug users, sex workers, and other groups.  RDS uses the concept of graph theory and the related characteristics that define social networks to draw conclusions about a given population.  The success of RDS in representing a target population is based on the connectivity of social networks and homophily within the social network of the chosen hidden population.    

The process works by identifying a small group of seeds that meet the criteria of the group to be studied.  The seeds then complete the survey and refer a specific number of their friends to complete it as well.  The process continues until the desired sampling size has been attained.  In the process a directed graph is formed with waves of referral that shows the distance between members of the group.  The network created is then structurally analyzed and statistically inferences are made from the pattern that was formed. 

This method of chain-referral has been criticized for its inability to provide accurate statistical data on the sampling; however, the above article suggests that if inferences are made from the social network encompassing the population instead of directly from the population itself unbiased estimation techniques may exist.  There is evidence supporting the validity of RDS that supports the idea that people will recruit their friends randomly when participating in the study.  RDS has been used to combat the problem of acquiring willing interviewees since it is common that members of hidden populations are engaging in illegal activity.  When people are referred by a friend with an established notion of trust they are more likely to participate in the study.  RDS relies on the insular nature of hidden populations to draw conclusions about their characteristics and has been useful in obtaining otherwise unknown data.

http://www.nytimes.com/2008/01/29/science/29tier.html?scp=10&sq=game+theory&st=nyt

At the end of the day, it can be well argued that love is what we all strive for as individuals. Unfortunately, love is a tricky thing and by that, I mean difficult and hard to find just the right person who will love you and all your weird tendencies. In response to this, online dating websites have persisted in competition for years claiming to be able to find your one and only soul mate. In a New York Times article by John Tierney, the idea of algorithms for finding love is examined. Such algorithms have been developed by psychologists and sociologists who have utilized personality tests as the means to match individuals on traits, such as emotional temperament and social approaches.

Hence, the concept of the online dating service can be viewed as a network constraining how the buyers (those seeking love) and sellers (those profiting from your lack of love) interact. With the goal of a perfect matching, most online dating services have reverted to these algorithms instead of allowing individuals to choose mates themselves. As stated in the article, when given the choice to choose mates on their own, individuals have been shown to date less than 1 percent of the profiles studied and such dates have ended unsuccessfully as well. Furthermore, it is likely that with similar ideal traits, individuals would choose the same mates over and over again, resulting in a constricted set among the buyers. As a result, the use of algorithms can be viewed as a mechanism to remove constricted sets as a central planner (the online dating service) eliminates competition by collecting buyer valuations and assigning the matches with the maximum payoffs.

Thus, it is highly subjective whether or not these algorithms, in which coordination of personality tests are used to assign individuals to different mates, achieve perfect matchings (market-clearing prices). I am arguing neither pro or con as companies have yet to demonstrate scientific evidence that such methods are even accurate. What I am arguing, however, is that these online dating sites have become a permanent fixture in our society today and several implications arise from that. We must consider them in light of accessibility as not everyone has access to a computer, which inherently creates power imbalances. We must also ponder the notion that such services aim for social welfare maximization not in the name of love, but quite possibly for the call of ‘money, money, money.’ Lastly, we must acknowledge the fact that love, no matter the means to achieve it, is difficult to attain, sometimes tricky to hold onto to and surely impossible to forget.

http://arstechnica.com/news.ars/post/20071017-comp-sci-researchers-use-economic-theory-to-disrupt-malware-black-markets.html

This article by Justin Berka discusses an interesting campaign underway by researchers from Carnegie Mellon University, the International Computer Science Institute, and the University of California, San Diego. The goal is to analyze and befuddle online black market interactions.

According to Berka, the researchers employ two similar strategies to choke Black Market trade. Both methods operate by creating false “signals”, which are “transfer[s] of information between the parties in [a] market”. By either posting misleading information about individuals (a “slander attack”), or by creating false identities, the researchers aim to destroy the credibility of illegal transactions.

This project invokes interesting issues in light of our classwork. In particular, it involves our work on trade theory. In our discussions, we have only lightly touched upon the concept of credibility; in this model, however, credibility is paramount. In terms of our class models, credibility may be likened to a player’s expected payoff. The greater the credibility of a trader, the more likely a buyer or seller is to receive his end of the deal. Conversely, low credibility means that there is a greater chance that a trader (by which I mean a buyer or seller, not an intermediary) will be cheated, or simply have a bad experience. The latter case corresponds to a decrease in expected payoff, which is essentially the goal of this research project.

If the researcher’s activity makes other buyers or sellers seem less credible, the apparent expected payoff to a trader would decrease. The introduction of fake traders, on the other hand, would actually decrease the reliability of the black market’s interactions, and consequently a trader’s expected payoff as well.

Both of these forces exert a potentially global influence upon the market. If enough fake traders and enough “slander attacks” infiltrate a market, it can cast a shadow upon the market as a whole. Although this possibility exists, one must consider the nature of the players in a black market. I posit that players in such a market are somewhat reckless, perhaps even irrational. After all, how trustworthy can illegally operating traders be, much less a whole market of them? And what if one was caught? These two considerations seem to point to a certain daring nature that would characterize traders in a black market.

In our third assignment, we considered the effect of an irrational player in an auction. It seemed that the net effect of an irrational player was a decrease in expected payoff for all bidders in the auction. If we extend this example to a Black Market where traders can be considered irrational, the ploys of the researchers seem less effective. Someone who already hurts their expected payoffs may not be overly concerned with low credibility. Or perhaps such a person has established a well-known network of traders (a sub-set of the market) already. The latter case would allow a trader to better navigate the increased complexity of a Black Market under siege by the researchers. In any case, I feel that the initiative is clever and will likely enact some effect upon illicit trade, due to a decrease in expected payoffs.

http://www.msnbc.msn.com/id/23335173/

In the above article, the Associated Press states that housing sales dropped .4 percent this month, the slowest pace observed since 1999.  The median price of a home has also dropped 5% in price since this time last year.  With this news, it becomes apparent that a bipartite preferred seller graph can explain what is happening.

With the current housing situation, perfect matching is not occurring;  in fact, there is a constricted set of sellers which results in a lack of buyers, hurting the seller’s chances of making a transaction.  Part of the reason that there is a constricted set is because the sellers prices are too high to allow for perfect matching and to clear the market.

Because of the inflated prices in the market, for there to be perfect matching, the seller needs to lower its prices to entice the possible buyers to purchase which turns this situation into a descending (Dutch) auction.  The sellers will continue to lower the prices of their homes until a buyer arrives at his reservation price and (assume breaking even means acting) will then purchase the house.  If the reservation price of the buyers are too low, however, the seller may quit trying to sell the home and wait until the market is more favorable.  This is the current economic state we are in: less houses are being sold, and the ones being sold are sold for lower prices.

Descending Auction

The recent release of the new website Juicycampus.com has created much controversy on college campuses across the country. It is a public blog that allows students on college campuses to make anonymous posts regarding current gossip that’s going around. Students are able to search the sight by simply typing in one’s name to see what turns up. Most of the blogs deal with drugs, sex, drinking, and rumors floating around. Very personal and private information has been posted on these blogs for everyone to see, ruining people’s reputations and feelings. People are worried that their name may turn up on a Google search, linking them to this information. Juicycampus has pledged that it blocks its discussion boards from being indexed by search engines. However, debates have arisen as a result as to whether or not this sight is appropriate. Colleges are attempting to band it access on their wireless networks on campus, as there’s little they can do otherwise. The recent release of an article by CNN, “College Website Posts Sex, Gossip, Hate, Rumor” looks into the current controversy spreading across college campuses. The first portion of the article expresses the reaction of a student from Cornell University, after witnessing posts about him.

Clearly this sight is a social network that connects people, very similar to other online networks such as Facebook. It closely relates to some of the material we have covered in class. An issue of game theory definitely exists, as people have options of whether or not to make posts in retaliation to comments made. There’s always the option of being the bigger man and not post or attempts could be made to improve the situation by posting the correct information. Choices will have to be made by individuals, but an individual acting in their own best interests could mean anything. The internet serves as a means for people to connect to one source, providing this somewhat controversial information. Through a network of friends, information is passed through a series of edges. The edges connect to nodes, which represent people. As long as the edges between nodes are positive then a friendship exists. These individuals serve as local gatekeepers in this network, as they monitor information passed within the group of friends. They exists only as local gatekeepers as the information can be attained through another path. They serve the purpose of providing a path between two neighboring nodes or friends for the information to pass through. Nodes can also represent a group a friends, and the means of which information travels between them may be between a certain node. This could be represented through gatekeepers, as the information has to pass through him or her to get to another group. This person may exist as a gatekeeper, as the information would have to pass through this node in order to get to the other group of friends. The decision to post or not to post, as well as the transfer of information within groups of friends and between the groups of friends exists on the Juicycampus network.

College Website Posts Sex, Gossip, Hate, Rumor. February 18, 2008.

http://serendip.brynmawr.edu/bb/berman/P4S3.htm

http://plato.stanford.edu/entries/game-evolutionary/

Evolutionary game theory brings a mathematical tool to analyze the concept of evolution and survival of the fittest. The websites above question the rigidity of game theory when applied to the biological or cultural side of evolution. Using the prisoner’s dilemma as a model for behavior implies that there is an intrinsic selfishness in humans. This selfishness can be explained partly because of the human desire to live and to keep their species alive. The first website mentions the motherly tendency to sacrifice her own life for the safety of her children. This variance from the prisoner’s dilemma we studied is slight but does vary from the predicted outcomes of selfishness. Genetics definitely instills an innate drive to live, but humans use other cues like loyalty and love which undermine selfish strategies that might otherwise be predicted. The Prisoner’s Dilemma makes more sense when applied to survival of the species, if the desired outcome is not individual survival, but survival of one’s offspring. This model could then include other strategies, such as cooperation, morals, or charity, to name a few.

The second site begins to address this, describing how evolution in regard to game theory is not only biological, but that cultural evolution is also valid. The notion of changes in beliefs and cultural norms leads to the evolving and differing strategies of humans. Looking at evolutionary stable strategies provides some insight to biological behavior. For example, like we saw in class, proof of the hawk strategy being evolutionarily stable over the dove strategy can explain many predator-prey relationships in the natural world. But, it seems to be inconsistent with human tendencies. Studies show that changing the payoff matrix leads to different stable population outcomes. While one example may have the population leading to hawks, another may have a more fluctuating outcome. This spawns from the fact that humans do not have equivalent preferences and values, making Nash equilibrium difficult to calculate for some games, where rationality is not necessarily the deciding factor. The ability of humans to behave differently from genetics is shown through things like loyalty and altruism from cultural evolution, which change the expected behaviors of people in social networks. At a higher level, these differences in strategies within a network become a species survival strategy in itself as individuals try various ways to survive within a group.

As many people already know, poker is not just a card game. As other gambles, opponents’ decision matter, and statistical probability also matters when one expects the result. However, different from many gambles that players have to play regardless of their hands, in poker, players can fold after looking at their hands. Also, different from some gambles that player cannot bet (score is counted to decide how much one earns), in poker, players can raise or just call. This is probably why many people say the outcome of a poker game depends more on skill rather than luck.

When I play with my friends, it is interesting to see that each of them have different beliefs about the best strategy to play poker. Some claim that it is about his action and others’ reaction, and some claim that it is more about probability. Yes, probability is important in poker games, but I never heard of poker champion who graduated from MIT, majoring in math (poker and blackjack is different). Poker is a game that one needs to read others’ mind and also think about probability.

In class, we learned that Game Theory can be applied when people have to make decisions in scenarios where the outcome is also affected by the decisions made by others. We also learned theories called Prisoner’s Dilemma, which shows that one’s outcome depend both on his and other’s decision. However, poker is especially complicated than some cases such as sports or on-line shopping. I agree with the article “High Stakes for Poker as a Learning Tool”.(http://www.nytimes.com/2007/12/12/nyregion/12poker.html?_r=1&scp=3&sq=poker&st=nyt&oref=slogin) , which supports the idea that poker involves a complex decision making and says poker can be used as a tool for education.

In poker, you have to consider about five opponent’s decision, rather than one. Another reason poker is complicated is that it is very hard to expect others’ action. . Your expectation of other players can be totally changed by one player’s random behavior. Also, unlike examples like ‘Prisoner’s Dilemma’ or Hawk-Dove Game”, which one can expect the other part will “confess” or be “Hawks,” in poker there is no answer. In poker, while players may have certain styles, you cannot expect almost anything because you do not know what other players’ cards are. Suppose a situation you have a quite strong hand, and from experience, you think that you’re chance is about 70-80% to win. So you raise money, and one player called and others hold. Now flop is opened, and none of the cards opened help you. And the other player goes all-in. What would you do in this case? At the first stage (before flop), you had 70-80% to win, but now you have to decide whether to give up money you put before or call. As shown here, poker requires players to make tough decisions by each of their “Game Theory” about 2-3 times for each deal. (Because players have at least 4 chances to bet, including pre-flop betting)