Google and the spectrum auction

I always wondered -as a person who has been working in the telecomms field for the past 11 years- how the firms developed their spectrum bidding strategies and how they implemented those. After all, these are billion-dollar games that have long and profound impacts in the firms as well as in the industry, and even in the economy of the country.

Spectrum auctions are sealed-bid long-run auctions where bidders -typically incumbent telecom operators- bid aggressively to get the spectrums that they need for their future infrastructures. FCC is the regulator of such lucrative auctions. FCC is incentivized by highe, successful bids and fair outcomes.

Recently, C-block auction resulted in a 8-week 20-billion dollar bidding process. FCC had a 4.6 billion dollar reservation price in this block. Before Google joined the bidding, no one knew why it was doing so. Google was actually planning to invest into the applications and handset business that would work on this spectrum block. As a result, Google needed big incumbents to invest heavily in this auction so that it can develop and sell handset and application products. In other words, Google played a billion-dollar chicken game, increasing its own bidding multiple times at points. Result: Google made sure that reservation price is reached, but it did not win. It was the outcome that Google desired. This strategic, well-played move was a big win for Google. Playing devil's advocate: What would happen if Google was the winner of the bidding process?In my opinion, they could either have dived into an industry that they are not very familiar with and failed eventually, or they could have sold the spectrum to another firm, potentially writing off some loss.

More details in the following article:

http://www.pcmag.com/article2/0,2817,2281382,00.asp

Global Climate Deal

This fresh news article of today was a mind-blowing one as I had a chance to see pretty much every aspect of Game Theory there: namely negotiation, fair division, distribution, Prisoner's dilemma, multiple players, efficiency, coalition and collective action.

As you all know, the delegates of the countries are trying to reach a global deal to reduce carbon emissions. According to some observers, there are two key players in these negotiations, developed countries and developing countries, i.e. rich and poor. According to some others, US and China are the key players. Denmark's involvement as a potential third player cannot be ignored due to the fact she is the host. From another perspective, there are hundreds of players (countries) or thousands of players (There are 15,000 people attending this conference!!) . There is also a regulator for this deal; namely United Nations.

In my opinion, it looks like China and the US will be the major key players though because they are the countries with the largest emissions and largest economical and political power.

Developing countries do not want to limit their peak emissions because they are still growing. Also, they claim that 60% of atmospheric space is occupied by 20% of most developed nations. So they think this is not fair. Developed countries, on the other hand, are afraid that their industries will be affected due to the new rules. They face strong objections locally in the industries that would be most affected. Also, they will be carrying most of the cost to support the developing countries so that developing countries can reach the targets by implementing new technologies.

This is an example of collective action multi-player Prisoner's Dilemma with prescriptive and predictive features. If everyone can agree, the total payoff for each player will be better because there is only one world. There is also only one atmosphere, and we cannot physically divide it. However, in the short-term some countries may financially be affected more. Further, short-term political pay-off structures impact the outcome.

In conclusion, by using the different expectations and payoffs of each country an efficient, balanced and fair deal is viable and has to be reached for the benefit of our next generations. However, due to the complexity of the game, it is very challenging to reach that deal.

http://www.cnn.com/2009/WORLD/europe/12/09/danish.draft.climate.text/index.html?iref=allsearch

Game Theory in Numb3rs (TV Show)

Game theory applications are used in every aspect of life including the agents' battles to locate and identify the criminals. Following video from TV show "Numb3rs" discusses the game theory implementation to find the assassin before he commits the crime. The idea is that by increasing the partial information about him and his potential moves, officers can think strategically and find out his potential moves and their probabilities, resulting in a mixed strategy equilibria.

http://www.youtube.com/watch?v=HCinK2PUfyk&feature=player_embedded

Also, the video establishes an analogy between "Hide and Seek" and "Battleship" games. Assume that the assassin would like to make this look like an accident. Then his options are limited. For instance, he cannot use bombs which limit his options. As the agents come to know more about the criminal, they can limit the options and probabilities and can eventually solve the crime.

Voting systems in each country

It seems truly amazing how different countries implement different type of voting systems. As we all know, no voting system is perfect, and none of them satisfies all five standard principles of voting theories. However, some of them yield more fair and democratic results than others. Following world map shows the most commonly used voting systems around the globe:

http://www.worldpolicy.org/projects/globalrights/democracy/maps-pr.html

The most common two methods are plurality representation(FPTP=first past the post) and proportional representation(e.g. D'Hondt method).

Below table explains a country by country voting method used:

http://en.wikipedia.org/wiki/Table_of_voting_systems_by_nation

Finally, IPU and ACE project provide rich resources for global voting and electoral issues:

http://www.ipu.org/parline-e/parlinesearch.asp

http://aceproject.org/

Eurovision Voting System

When we discussed voting theories in our SDA class, I thought about the voting system in Eurovision. Those who are from Europe or who have lived there in the past know very well that Eurovision is a major commercial, entertainment and political matter throughout Europe. It is an annual Europe-wide song contest that has long been controversial. It is broadcasted simultaneously all around the world to millions of viewers.

Previous voting system was such that only fans in each country determined the outcome by voting for their most favourite songs excluding the songs from their own countries. However, historically neighboring countries voted for each other (Balkan Block for instance). So, countries with most number of neighbors had an advantage. Also, countries who wanted to send a political peace signal to some other countries voted for them (e.g. Turkey to Armenia or Cyprus). Finally, diaspora voters became really strong as the immigrants in another countries voted heavily for their original countries. Due to these reasons as well as increasing number of new Central and Eastern European countries on the map, the chances of winning for Western countries have been significantly reduced. In protest, some of them have withdrawn from the contest. Even, petitions were started to change the voting system:

http://www.petitiononline.com/newesc/petition.html

In 2008, the voting system was changed to a half-jury, half-fan system which aimed to eliminate the criticism around the voting method. It also aimed to increase the musical value of the contest rather than political supports or visual aspects of the competition:

http://www.telegraph.co.uk/culture/music/music-news/5332681/Eurovision-2009-new-voting-rules.html

Under new rules, viewers from 42 countries vote for 25 finalists songs by phone or text. Each viewer can vote 20 times but cannot vote for one's own country. Next, judges in each country vote. The two different types of votes in each country are then combined equally, which determines the country votes for the Eurovision song contest: The favourite songs get 12, 10, 8, 7, 6, 5, 4, 3 ,2 and 1 points in decreasing order of preference. The song/country with the highest number of cumulative points win the contest. A snapshot of the points assigned by each country to others is shown below (You can also click this picture to see the URL link of Eurovision contest):


Below video shows the actual live voting process:

http://www.youtube.com/watch?v=KBqTo2lIIGs

This method is very similar to a modified Borda count; modified in the sense that jury determines one half of the votes and millions of viewers determine the other half. However, due to larger point allocation system around the best two (12 and 10 points), the results will be in more in favour of top-2 preferred songs/countries in general.

It looks like, regional and diaspora biases were mitigated to some extent in 2009 thanks to the new voting system. However, we need more samples to make a more informed evaluation whether the new voting system is fairer in evaluating the musical content. My personal opinion is that judges will be biased to a great extent too. Eurovision contest creators will need to make further changes to make the competition more around the music.

Auctions are moving online

The followiing recent article by The Economist blew my mind once again. The fact that you are opening your specific piece of art to thousands of bidders at the same time all around the world has profound economic impacts on the auctioneer side as well as on the bidder side. Major auction firms are evaluating the profitability of this option against other concerns such as brand management and psychology of the bidders. It seems like some of them already chose to go online.

Article shows that online auctioning proved to be profitable for Christie's within 1 year. Christie's is probably charging an admission fee for online bidders. In addition, it is selling the items to higher prices because of the differences in the reservation prices of bidders globally. Say, a painting might have a reservation price of ten thousand dollars to a bidder in Russia while it might be only five thousand dollars for a bidder in the United States.

Other important factor is the psychology of bidders in the bidder room. In-room bidders may have a better idea of where the auction is heading and when it might end because of their physical presence in the room.

http://www.economist.com/specialreports/displaystory.cfm?story_id=14941237

HOA Board/Committee Volunteering

I noticed during the last few weeks that HOA Board/Committee volunteering is a good example of Prisoner's Dilemma in a sequential format. Due to some criminal incidents lately, HOA Board of my community decided to establish a Security Committee and asked for help. I was one of the 6 volunteers to work in this committee. We met few times and decided on actions to take. Recently few members stopped supporting the committee. As a result, I have immediately seen an increase in my responsibilities (worse payoff due to more time/effort spent). So some of the players "cheated" causing a worse payoff for "self" while they received higher payoffs. It was because remaining committee members took care of the duties anyways and "cheaters" enjoyed"better payoff" due to less time spent on the matters. Because we are the owners of the houses and we will be playing this and other games with those players again, they will have a bad reputation and will not potentially be trusted in the future.

I created the following simplified sequential game tree where all the players other than "self" were considered to be a single player (called "them") to simplify the game structure. Two strategic moves for each player are (Contribute or not). There is a single nash-equilibrium of (Not contribute, not contribute):

However, collectively as a community, if we could have recruited more volunteers, everyone would be better off. Total payoff would be largest for the society while individual payoff would be reasonably good (5 for self and 5 for them).

In general, most volunteering situations can be considered to be a "Prisoner's Dilemma" type of game. They can also be modelled as simultaneous games.

Strategic Moves during the last minutes of an NBA game

The following two articles explain some of the strategic moves that go on in the last minutes of an NBA game. We all know that depending on the point differential and remaining time, coaches constantly change their strategies to be ahead of the game. Due to the certainty feature of this timed sport, a tie option is 3 times more heavily observed than it normally would have been as can be seen in the second link. Strategies include slowing down the game, fouling, taking breaks, changing players etc. Over the past decades these strategies have matured and came almost to a perfection thanks to the repetition of the game. Coach of a team estimates the potential moves of the other team, and calculates his best move to get the highest payoff. Since the major aim is to be strategically unpredictable to the other team, the dynamic/competitive nature of basketball results in a mixed strategy equilibrium.

http://www.sportshistory.us/uncertainty.html

http://cheeptalk.wordpress.com/2009/06/10/the-overtime-spike-in-nba-basketball

Friendship as an Assurance Game

Well, another personal example of game theory! Assume two rational players: self (S) and friend (F). Assume also that these people already know each other and they will make a decision whether to invest into the friendship with the other person or not. Similar to other interpersonal meta games like dating, this game could be simplified into a 2x2 matrix assurance game as follows:

Imagine that both players are simultaneously thinking about the other and deciding whether to invest or not, but not sharing this decision with the other person. This meta-game portion is a simultaneous, partial information, repeated (or one-shot), variable sum game with manipulable rules.

There are two Nash equilibria that can be found using Best-response analysis: (Invest, invest) and (Do not invest, do not invest). The first one has a higher individual and total relative payoffs (say, in terms of happiness) for both players. So, if the two players talk and cooperate, they might achieve higher payoffs using (Invest, Invest). If S invests, but F cheats the agreement by not investing in the friendship, then F does not get any payoffs but S is now worse off than not investing because he spent time and resources. In this case he/she may be punished in future games with S and other players.

Obviously, the actual meta-game is a lot more complex and we cannot be friends with everyone. However, when we choose the right opponent player in the "Friendship" game, we should show the game table and let him/her know that cooperation provides better payoffs for both!!

Dividing the Wi-fi

When I moved in to my new house a year ago, among so many different issues was the internet/cable requirement. I am not a big TV fan, so I decided to subscribe only to internet. I do not use internet very often, and when I do so I do not need the whole band. I feel like most average users are the same way. So, it seemed really redundant to pay $50 a month for the internet subscription that I would only use few hours a month. On the other hand, I really needed it because of the school. I had this idea to talk to my neighbors but I hardly knew them so i could not. Result: I subscribed myself and am still paying $50 a month:)

So imagine 3 adjacent houses in a townhome community Say A, B and C. Each needs a low amount of internet usage at different times. Residents at A use internet only at weekends, residents at B use it only during the weekdays because they are working from home and residents at C demonstrate a more scattered but occassional internet usage at random hours. Let's assume that they know each other and they do not have internet service at this point. Cost of internet per month per house is $50. If A, B and C have this information in front of them and have some basic Game Theory perspective they would be able to create an efficient and fair division by understanding their requirements. If A and B cooperate and get the Wi-fi together, they would share the service and the cost, each paying $25 dollars for almost the same service experience. Obviously, division does not have to be exactly like that. For a 2-person coalition, the line between -50 dollars on x and y axis forms the ZOPA and Pareto efficient frontier, similar to Goreton and Bushville discussion in the class. Their BATNA is to install their own equipment and their reservation price is $50 each.

If C also joins the coalition, then the cost per person would reduce to $17 and the service would still be almost the same. Obviously, we need to assume that the Wi-fi router is located near the window at B and provides good service to neighbors. Residents could sign "Terms and Conditions" between the three of them pertaining to rules of service acquisition, usage, payment and termination. Note that the differences in the level and timing of internet usage are key to creating and achieving this efficient frontier. Another key issue here is the access to information among the neighbors...