Until recently I’ve always understood voting to be very simple for the citizens; you pick a candidate whose views best represent your own, maybe you try and talk some of your friends and family into agreeing with you, then you head out to the polls on election day and put in your vote. Apparently it’s not so straight-forward. Consider a simplified example: Say there are 3 candidates: A, B, and C, such that A and B agree on a good deal of the issues, but do have significant differences, and C is completely different from both. Maybe recent polls predict that C will get 40% of the vote, and A and B will each get about 30%. If everyone goes to the polls and picks their favorite, C will win, and 60% of the population will be extremely unhappy. But if the voters who support A and B can play some sort of game, they could both do better. A’s supporters can do better by collaborating and voting for B, they’d be much happier with B in office and they can make it happen. The same goes for B’s supporters. A and B are pretty much playing a battle of the sexes, where they would absolutely hate having C win, which is what will happen if they dont collaborate, but neither will be too happy if they have to give up their candidate. I really know nothing about US History, but maybe this concept is what started the 2-party system…both parties realized that they could gain an upper hand with fewer candidates.

It can also get much more complicated. This article talks about possible strategies in the upcoming election, as well as in past elections. http://www.foxbusiness.com/markets/industries/government/article/games-played-vote_482985_18.html

It’s interesting to see how much strategy matters behind the scenes in a campaign, and it’s strategies like these that manage to put people in office with less of the popular vote. What makes this related to what we study in class is that these campaign decisions dont just depend on statistics alone, they depend mostly on what the other campaigns decide to do. For instance, it’s not just because big states like California have a high population, or that they’re more likely to vote one way or the other that makes everyone focus their efforts there. It’s because every other player is already focusing so hard there that each player needs to in order to stay alive. If no one else was putting much effort into California and you could take it without too much funding, putting all your effort there is just wasting resources. And if the republicans decide to try this thing in Wisconsin and all vote for Hillary, maybe Obama supporters will see this as an opportunity to do the same thing and vote for Huckabee, since it’s unlikely that Obama will be able to beat Hillary supporters AND McCain supporters. If this happens then all the campaign’s will have to think and decide if they can make a better play. What this article doesnt mention about Lincoln’s campaign either is what Buchanan, Fillmore and Fremont did in response. I think it’s unlikely that Lincoln just got unlucky and his idea didnt work, but rather Buchanan played his best response to Lincoln’s strategy and it worked out in the end.

What makes this interesting is that campaigns clearly aren’t just about getting the majority to vote for you, it’s about using the votes you have effectively and getting the votes that count most. What makes this more interesting is that these strategies depend mostly on how the other campaigns play their cards. Maybe this is why incidents like watergate happen. In a way I guess it sucks that the best strategist wins over the candidate with the most votes, but it’s not horrible that our politicians are forced to understand strategic decision making.

Researchers at New York University recently studied the correlation between articles, blogs, and MySpace friend counts to the success and volume of album sales. Professor Vasant Dhar has come to the conclusion that if 40 or more normal people had written blog posts prior to the album release, the rate of sale of the album ended up being three times the average album sales. Though not as substantial an effect as blog postings, Dhar addressed the fact that the more MySpace friends a band had, the greater their album sales were. Still Dhar reported that the traditional reviews from mainstream critics still have the most potential to affect album sales as “a good review in Rolling Stone can’t be beat.”I think this topic is very important and brings an application to some concepts discussed in early lectures.  For example, the Strong Triadic Closure Principle, used in this example, will make a huge impact in spreading the word of a new album release. Similar to MySpace, Facebook, very popular with companies and college communities alike, contains applications which allow users to post their favorite artists and albums.  Then, enhancing the spread of information on this social network, the “newsfeed” feature reports the recent activities of the user’s friends. Though not mentioned in class, this could be viewed as the rapid creation of edges on a social network.An important thought brought up was the question of manipulation.  For example, could band managers manipulate blogs in hopes of generating more sales in the declining music industry.  Or worse, could rival bands use blogs to send out negative review’s of their competition’s album?  Especially with the rampant piracy and widely available online downloads, blogs and social networks could soon over-power traditional sources of album publicity.

Network World is a weekly IT publication that provides news and information to network executives.This Network World article (http://www.networkworld.com/community/node/25502) describes plans for modular satellite technology, which is intended to replace present day all-in-one, monolithic spacecrafts. The project sponsored by Defense Advanced Research Projects Agency (DARPA) hopes to solve issues of scalability and reliability by splitting the standard satellite systems and allowing the different “fractionated modules” to change function if necessary by replicating functions of each other. Modules can be physically connected once in orbit or can be linked together via a wireless network to create a virtual satellite.

The advantages of the proposed multi-module system can be described using the concepts we have studied in class. Essentially, we are taking a single node network, or on a micro scale, a system in which a failure of a single component implies a failure of the entire network, and we are changing it into a network, consisting of individual modules (nodes). Since all nodes are interconnected, failure of one node does not cause the entire network to fail. Segmentation of a network makes it more robust. 

Multiple networks might have identical or similar components, which can be shared in case of emergency. For example, in an undirected graph shown above, two networks, Network 1 and Network 2, node A and node B are either identical or serving the same purposes in their respective networks. If node A fails, B can temporarily replace node A by embedding itself into Network 1. Multiple components (subnetworks) may choose to use the same module, for instance to save money, however in this case this module spans a structural hole, and its failure will damage both networks at once, so this kind of sharing should be used only in emergency situations.

Modular network of satellites displays a lot of flexibility, since any new launched element can quickly fit into an already existing structure by forming new edges with already existing fully-connected component (as seen in the example above). Multiple new nodes that have strong links to an existing component will be more likely to establish new edges between each other, which can be linked to the Strong Triadic Closure Principle. Naturally, we expect modules linked to the same modules to have links between each other. This is just like homophily and proximity in social networks.

Other advantages of modular satellites include upgradeability and incremental deployment of capability to orbit, which is related to how easy it is to add another node to an already existing network, and reduced integration and testing due to subsystem decoupling, which is related to ability of each node to exist independently of the network.

According to Network World, the first companies and their partners awarded contracts for this project this week included Orbital Sciences ($13,648,758); Northrop Grumman Space and Mission Systems ($6,159,866); Boeing ($12,891,049); and Lockheed Martin Space Systems ($5,762,781). It will be exciting to see these new virtual satellites being developed and launched in the near future. (…or even better: to work on designing and developing them).

Article link: http://arbitrarian.wordpress.com/2008/01/28/nba-game-theory/

From this article about NBA Game Theory, I find out that in basketball game, there is a high correlation between assists ratio and field goal percentage. If players try to find the other open players, open players will be likely to score with high percentage shooting. However, if players try to shoot even though defense players tightly guard in front of them, the field goal percentage will be dropped significantly. Therefore, in the view as a team, high field goal percentage will lead team to win more. However, as a view of players, players should score more and make their statistics look good for future salary contracts because salaries are highly depended on statistics. Therefore, like shown in the article, in some team like New York Knicks, players face prisoner’s dilemma between passing to other players and shooting in front of a tightly defense player. The best outcome from this dilemma is passing to other players, but since some players are selfish enough to think more about their salaries over team’s wins, they just try to shoot the ball on their own with low field goal percentage. These are differences between winning teams and losing teams in NBA.

Like a prisoner’s dilemma above, game theory can be applied to strategies of offense and defense players. San Antonio Spurs was a champion team last year. In addition, they won four of last 9 seasons with their best player, Tim Duncan who is an arguably best power forward in NBA history. Since players like Tim Duncan dominate their opposing defense players easily, it’s almost impossible to guard him by a single defense player. Therefore, many times other teams try to guard Duncan with two players, but at that moment, Duncan passes the ball to other players and they score with ease. In this situation, game theory can be applied.

Tim Duncan’s filed goal percentage is 50%. When he is guarded by a single player, he can score with 70%, while if he is double teamed, percentage is dropped to 30%. Also, shooters in every NBA team can shoot 3 pointers with 70% chance, if they are opened. If they are guarded, the percentage is dropped to nearly 30%. Therefore the payoffs of Duncan’s play will be 0.6(if he shoots with double teamed), 1.4(if he shoots with one defense player), 2.1(if he passes when double teamed), and 0.9(if he passes when one player guards him).

Double-teamed Not double teamed

Shoots (0.6, 0.7) (1.4, 0.3)

Passes (2.1, 0.3) (0.9, 0.7)

In this situation, there is no Nash Equilibrium. This is the reason why there is a no a dominant strategy for opposing teams to guard Tim Duncan or dominant offense strategy for Tim Duncan. And, that’s why basketball is complicated and has many different offence and defense strategies. Surely, in the situation above, Tim Duncan will look for other players when he is double-teamed. If he is not double teamed, he will shoot by himself. Those two strategies make most payoffs for him. In contrast to Tim Duncan, opposing teams will try to defense Duncan with two players while he is shooting, and will try not to guard with two players if Duncan tries to pass. For sure, the team that uses its beneficial strategies more fluently will win the game. But, as you can see above, Tim Duncan’s team has a higher chance to make efficient plays and that’s why San Antonio Spurs are still top ranked in NBA standings.

(http://sports.yahoo.com/nba/standings)

In class we have learned about Braess’s Paradox and how the removal of a six-lane highway in Seoul, Korea actually improved the flow of the transportation.  The plan of removing the motorway was a part of the historic restoration project of Cheonggyecheon. Last winter break, I actually had the opportunity to visit the newly renovated Cheonggyecheon public park which is why I initially decided to delve into this topic.   Searching for information on the restoration project, I came across this article http://www.guardian.co.uk/environment/2006/nov/01/society.travelsenvironmentalimpact which discusses how the transportation plan was successfully implemented.  The author of this article explains how Braess’s Paradox can be applied to the restoration project and emphasizes how other metropolitan cities can improve the flow of traffic by removing various highways and paths.

 To an older generation of Koreans, Cheonggyecheon, means “Clear Valley Stream”, and it is a location that was known as a public rest place that has witnessed the many joys and sorrows of Seoulites for centuries.  This image and history has changed drastically since the end of the Korean War and many plans were executed to restore the Cheonggyecheon area.  In 1970, in attempts to find a means to compensate for this severe increase in traffic, the stream was covered by an elevated road which was composed of a six-lane motorway.  This road quickly came to symbolize industrialization and became an area that was infamous for the oversaturation of cars, flea markets, and homeless people.  Realizing the urgency of the situation due to the rapid degradation of the area, the Seoul Metropolitan Government and the citizens of Seoul, cooperated to execute the restoration project. Under the leadership of Lee Myung Bak, who was the mayor of Seoul at the time and is currently the President, a plan was instigated in order to remove the freeway, and to restore the stream. The amazing aspect of this project was how the issue of transportation improved despite the fact that a major freeway was removed. 

Professor Kee Yeon Hwang, the mastermind of the traffic plan, realized through his analysis of past cases that the removal of the freeway would increase the flow of transportation. He stood by his decision and just as Braess’s Paradox claims, the existence of the freeway actually resulted in longer traveling time for commuters.  Also the implementation of the Bus Rapid Transit system helped cut the automobile usage even more. The restoration project of Cheonggyecheon of Seoul, Korea may serve as a model to other metropolitan cities that have endlessly expanded in developing their cities without proper planning.  For many urban planers, it may seem practical to build more roads as possible for a health city growth; however, Braess’s Paradox indicates that this is not always the case.  

http://web.cs.wpi.edu/~ypisan/virtualworlds/readings/tree-house-to-barracks.pdf

With 10 million active accounts, World of Warcraft (WoW) is the most successful massive multiplayer online (MMO) game out there on the market.  One reason for the success of the game is that it caters to many different types of players.  People that want to solo, people that want to tackle the end-game dungeons (player vs. environment aka PvE), people that want to kill each other (player vs. player aka PvP), and everyone in between has a place in WoW.

The interesting thing is that in order to progress through all the different dungeons  in PvE and/or move up in tournament rankings in PvP, players need to form groups (anywhere from two to forty players) and interact with other players.  On top of all that, players need to cooperate together if they want to succeed in whatever endeavors they choose to undertake.  Getting groups small groups of people (five or less) to all do the same thing might sound trivial.  However, getting groups of forty to all act in unison with everyone playing some vital role is simply amazing.  To put things into perspective, baseball teams only have nine players working together, soccer teams have slightly more at eleven players working together–this is nowhere near the forty that is needed for some situations in WoW.  Another way to look at it is to imagine working on a group project (where social loafing is not possible because the group as a whole will be penalized for one loafer) with thirty-nine other people.

With so many people working (or playing depending on how you look at it) so closely together, it’s not surprising that graph theory can be applied to the WoW playerbase.  The article linked above discusses how all different sorts of relationships between players arise from simply playing the game.  It is not uncommon at all for players to join guilds so that they can have an adequate number of people to do things.  What’s intriguing is that all the different guilds can basically be seen as clusters or components on a larger graph.  Players in a guild generally tend to have at least a weak link to every other player in the guild.  And, the longer a player stays in a guild, the more strong links that player will have with the other members.  It is also not uncommon for players in guilds to just have links to other players in their component.  Only a few players in a guild act as bridges to the larger network of players.  What’s more, in order to make sure that the guilds are stable, it is also not uncommon for guilds to have a blacklist of other players or guilds.  Therefore, every member of a guild will have a negatively weighted link to those that are on the blacklist in order to keep the peace.  The article (as well as what I can think of off the top of my head) goes on, but this is just the gist of it.

With MMO games steadily gaining popularity, it will be interesting to see the networks that develop in them.  It will also be interesting to see if there can be other uses for MMO games besides modeling social networks.

A college education in an Ivy League university is supposed to inculcate strong critical thinking skills and the ability for independent thought, but recent research suggests that  perhaps deep down we aren’t that different from sheep that need to be lead. Researchers from Leed University conducted an experiment where large crowds of up to 200 or more people were told to walk around in a hall randomly. Inside this crowd of people was a small percentage of individuals who were given more specific instructions with regards to where to walk. Communication was forbidden between all participants, but they were told that they had to stay within arms length of other people. The result was that in all cases, the “informed individuals” became the de-facto leaders, with the crowd following them and even forming a “self organising, snake-like structure”.

A small percentage of people were all that was needed to influence the rest of the group to follow - 5% were all that were needed to get the crowd to follow in a particular direction. This is not particularly surprising, as it is consistent with earlier theories such as Malcom Gladwell’s Tipping Point or Duncan Watts’ “No one group” rebuttal of the Tipping Point . This is, after all, the entire reason why companies spend obscene amounts of money on celebrity endorsements and trying to get the “Influentials” to spread interest in their product. What is interesting, however, is the fact that such behavior was observed even in the absence of overt communication between the participants. The leaders didn’t say “Follow me”, but everyone just started following in their wake. This is even more curious considering that there wasn’t any particular reward associated with following the leaders - there wasn’t anything to be gained from walking the way the leaders walked, no payoff matrix to observe and deduce a Nash Equilibrium from, but such “Follow the leader” behavior simply appeared in the crowd. In fact, according to the researchers, it seems that most of the people in the crowd didn’t even realise they were following anyone.

Another experiment showed a related outcome was conducted in the University of Calgary: There, researchers noticed that while working on a task, if there was someone else in close proximity doing a different job, progress on the task assigned would slow, presumably because the participants were mentally modelling themselves as doing the same task as the other person close by. Such behavior was due to our mirror neuron system, suggesting once again that overt communication was unnecessary to get others to follow in your footsteps.

I’m sure the two articles provide some comfort to all the other students out there who’ve ever struggled with their homework while their roommates are playing video games - it’s not our fault we automatically feel like playing video games too!

Many people have recently become involved with internet networking sites. The most popular of these being myspace and facebook. These tools help people to organize different networks to which they belong. On facebook users can be separate friends that they have by the different colleges and the different geographical regions they are from. The graphs that can be created following the patterns of friends show that very strong connectedness exists within the network. Facebook and Myspace are major vehicles examples of facilitators for the small-world phenomenon. It would be very difficult to sit down and think about all the people who you knew but these sites help to make this more clear. Seeing the level of connectedness that each person has with others and how far across the world these ties go. These social networks also allow for people to make a connection with both long time friends and casual acquaintances. It is a place that allows for both strong and weak ties.

Another feature of internet networking sites is that they allow marketers to reach a wide distribution of people. Although this is true there has been some concern about the effectiveness of this marketing. In the article “Networking sites MySpace, Facebook losing popularity” written by The Grand Rapid Press, the effectiveness and the future of such internet networking sites comes into question (http://www.mlive.com/news/index.ssf/2008/03/networking_sites_myspace_faceb.html). People when asked say that they do not even notice that the ads. Marketers have found that as few as four in ten thousand visitors click on the ads that appear on the sites. This is a major problem for these marketers because these large networks offer a rare opportunity to connect with millions of users.

In our last lecture, we talked about the internet as an information network built to function in a manner similar to natural thought (as laid out in Vanevar Bush’s As We May Think). This would seem to be an extremely logical way to order data, yet there are many systems that defy this natural order. One, for instance, is television. With the rise of YouTube, video data has achieved a slightly more neural-network style progression with the advent of “related videos”, but television in general is ordered by stations with shows appearing when the stations say they should, not when the user wants them to.

However (as reported here: http://news.bbc.co.uk/2/hi/technology/7259339.stm), it appears as though this could be about to change, at least in Europe to start. The European Union has decided to embrace the technology of peer to peer file sharing to create a peer to peer TV network. This technology is perfect for creating the system of TV for the same reason it is so popular in the illegal file sharing underground: very few servers are required to store and serve the information. In the traditional server/client relation, the server must send content to any client requesting it, creating a directed graph with one node pointing to others. In the case of TV, this graph would be very lopsided as one server (a TV station) would point to millions and millions of clients (viewers).

Peer to peer solves this problem by making everyone a server for everyone else, eliminating the one sided nature of the classical internet map when applied to TV. The system works with peer to peer technologies for things like that, but it will also incorporate live feeds for things like sports events – essentially replacing all of TV with the internet. If this becomes widespread, perhaps we could see the evolution of TV sometime in the immediate future.

Recently in class we have discussed the ultimatum game and how its results have differed from those that would be expected from purely rational and profit-maximizing individuals. In recent years, the game has been played in numerous different scenarios and has resulted in a number of unique and interesting results. For example, behavior of participants in the game remains relatively constant among Western cultures, but large variations of outcomes have resulted in other societies.

In a recent paper (available online at http://www.pnas.org/cgi/content/full/104/40/15631?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&fulltext=cesarini&searchid=1&FIRSTINDEX=0&resourcetype=HWCIT ) published in an October issue of the Proceedings of the National Academy of Sciences, a Ph.D student at MIT and three Swedish researchers discussed the role of genetics on the ultimatum game. The researchers compared the results of identical twins (share the same genes) with fraternal twins (do not share the same genes) and found that the identical twins were more likely to play the same strategy. In fact, their research has shown that around 42% of the variation in responder behavior is attributable to genetic effects.

This study, along with others which examine the ultimatum game are important because while neoclassical economics assumes utility-maximizing and profit-maximizing behavior, there seem to be a number of factors, one being genetics, which complicate this belief. Better understanding what drives economic decision-making by recognizing the role of genetics in this process can help better explain the results of network exchange theory studies and can increase the effectiveness of the models designed to predict human behavior.