game theory Archives

Moving Fuji

A couple of weeks back I mentioned that I was reading ‘How would you move Mount Fuji‘ by William Poundstone. It’s an interesting read but no doubt pretty ‘scary’ in terms of the facts it is trying to convey. Poundstone presents not only the puzzles that are offered in interviews conducted by Human Resource departments of ‘talent-intensive’ firms but also a discussion on whether this methods work to hire people that these companies need. He highlighted that companies thinks bad hires are way too high a risk that they would rather risk losing a good hire than getting a bad one by mistake. As a result the interview processes are all skewed towards rejecting people (more or less) and identifying somewhat extraordinary people.

Poundstone identifies an important question HR people should be asking themselves when posing puzzle questions and that is whether the candidate’s response is going to make an impact on your decision to hire or not. If it doesn’t then there’s no point asking the question, like this particular interesting one. Indeed, from a HR point of view, interview time is precious and missing out on good hires is not a pleasant thing either, so structuring an interview to be efficient is very important. If a puzzle can help to bring out key information then it should be used; otherwise, it might just be an abuse of the interviewer’s position.

Well if you’re not interested in the software industry or the finance industry then the puzzles in the book would be nothing more than just a great past-time. This makes Poundstone’s book more interesting because his section of answers to the puzzles given as examples in the book launches into elaboration discussion of how one should approach the puzzles. He attempts to give us some general guidelines on solving these puzzles. He gives nine steps (mainly for interviewees though):

1) Decide what kind of answer is expected (monologue or dialogue)
2) Whatever you think of first is wrong (that’s why it’s a puzzle)
3) Forget you ever learned calculus
4) Big, complicated questions usually have simple answers
5) Simple questions often demand complicated answers
6) “Perfectly logical beings” are not like you and me
7) When you hit a brick wall, try to list the assumptions you’re making. See what happens when you reject each of these assumptions in succession
8) When crucial information is missing in a logic puzzle, lay out the possible scenarios. You’ll almost always find that you don’t need the missing information to solve the problem.
9) Where possible, give a good answer that the interviewer has never heard before

Overall, Mount Fuji is a good read, but I’m guessing most readers would want to speed past the explanations and go straight to the answers themselves. The best puzzles always have the most counter-intuitive solutions and elegant answers. For more practice, you might like to visit some of Poundstone’s reference sites: here, here and here.

Game Theory & Politics

Christopher Beam on Slate.com framed the Senate (or any democratic deliberative body) as “the world’s greatest collective-action problem“. In a way, it is. Debating on issues and surfacing potential problems stakeholders might face and arguing on the different consequences on different parties is one thing about parliaments and national assemblies but then decision-making is another.

In democracies, debates and discussions are known to hold up decision-making and the same is reflected in bureaucratic bodies where power is shared across several individuals. This dispersion of power calls for coordination to get anything done and thus allow game theoretical analysis to dissect the dynamics involved in any of those coordination outcomes (ie the final decision).

In some sense, this is a trade-off; deliberation this way that involves the coordination game ensures that the outcome cannot be entirely fair though it might provide an illusion of it. In the first place, reality includes a spectrum or even several dimension of opinions and no system can be designed to capture and aggregate this complexity. The authors of Thinking Strategically mentions this in one of the chapters on elections. As a result, we are left with the political game that is manipulating the legislative structure although everyone hates to admit it. In some sense, Singapore’s structure might churn out better results in terms of efficiency and do ‘the right thing’. The idea then, is to move the game away from the ballot box in the first place, to somewhere further and higher.

Fox vs Time Warner

Days ago I stumbled upon a recent dispute between Fox and Time Warner Cable. The basic idea of the dispute was that Fox wanted more money from Time Warner for carrying their channels and Time Warner didn’t want to. The whole thing ended up with publicity campaigns on both sides (Rolloverorgettough.com for Time Warner Cable and Keepfoxon.com for Fox) to make use of TV viewers’ support to raise their bargaining power. They eventually settled the dispute so viewers will continue seeing Holmer Simpsons munching on doughnuts.

It is interesting how Lauren Collins explained in The New Yorker how Time Warner Cable was basically using a forced-decision device since there’s a spectrum of other options available to them. Time Warner Cable could have just absorbed the price increases and sacrifice their profits. By running the Ad campaign, they’re signaling to Fox that they’ll not accept any changes to the pricing of the deal – either get paid the same or no more screenings of Fox programmes; effectively introducing a Morton’s fork. At the same time, like what Collins mentioned in the article, “The strategy in a nutshell: couch potatoes as human shields.” The company handling Time Warner Cable’s campaign, Purple Strategies is pretty amazing; they are basically specialist in positioning stand in public for organizations or political bodies in ways that allow them to maneuver themselves under different circumstances.

The incorporation of strategic movements in corporate lives is going to become increasingly common, which gives us more reason to check out Dixitt and Nalebuff’s Thinking Strategically or their newer Art of Strategy.

Thinking Strategically

As I was mentioning a couple of weeks back, I have been reading Thinking Strategically by Avinash Dixitt and Barry Nalebuff. This is a pretty old book, being first published in 1991 and the version I was reading is the 1993 paperback re-issue – there was no more revisits to this book by the authors since then but it’s been in print until now. I believe it’s largely used as readings for undergraduate economics students as well as students of business or management schools.

The 2 authors are great teachers of Game Theory in Princeton and Yale and have often adapted the principles this somewhat mathematical subject to the less mathematical real world. Thinking Strategically is a great attempt at discussing strategic thinking that follows from game theoretical analysis for the layman.

The good thing about ideas on strategic thinking is that their principles hold even when the examples they are attached to often become obsolete or arcane – that is not to say that Thinking Strategically features arcane examples. Most of the examples used to bring ideas across in the book are simple, often bordering trivia but they illustrate the essence of the concepts and can be used to explain the principles for similar but more complex issues. One of the case studies brought up that I particularly love is the one about a three-way duel where we have 3 shooters of varying abilities.

Each shooter fires at someone (or something) each round; there’s is fixed order as to who gets to shoot first. The one who’s allowed to shoot first is a poor shooter with an accuracy of only 30%, the second has an accuracy of 80% and the last is a sharp shooter who shoots with an accuracy of 100%. The question is that if you’re the first shooter and allowed to go first, who would you choose to shoot?

An analysis of this “game” gives us a surprising but convincing result. If you choose to shoot the average shooter, and succeed, you will definitely lose because the next in line would be the sharp shooter and he would shoot you. If you choose to shoot the sharp shooter and hit, the average shooter will shoot you, leaving you with a 20% chance of survival. And even if you survive, you only have 30% chance of hitting him later. You might say, this mediocre shooter is so lousy, he’ll probably have to lose anyways. But you can actually raise your chances of winning by choosing a more intelligent strategy: To fire into the air.

This way, the average shooter will get his turn and attempt to shoot the sharp shooter since shooting you and succeeding mean he’ll have to die when the sharp shooter’s turn comes. If he succeeds, the mediocre shooter gets to try his hands at killing the average shooter. If he fails, the sharp shooter will immediately kill him and that once again, leaves the mediocre shooter with a chance of 30% to kill the sharp shooter. The somewhat counter-intuitive strategy of shooting at no one raise the chances of the mediocre shooter winning substantially.

The principle alluded by this example is that if you’re a weak player; it is wise to allow the stronger players to make their moves and get rid of all each other before making a move and fire your best shot at the one left standing. Now that we surface the principle, the logic of such a choice becomes more intuitive.

Thinking Strategically is a great read for students who likes to think and don’t mind re-reading some of the statements in the book a couple of times to understand the explanation behind some strategic moves. It teaches an important skill of looking forward and reasoning backwards and shows you the power of its application in all sorts of “games”. The book might make you feel like you’ll become smarter but trust me, it’s not that easy to apply strategic thinking that quickly in real life and often, we need a degree of foresight that we would almost definitely lack.

Cafe Ironies

In an entry with the same title on my blog, I detailed my experience at The Coffee Bean recently that didn’t quite start nicely but ended off pretty intellectually. I’m quoting the gist here:

I approached the counter with a maths worksheet in my hand (I was planning to work out some problems there while I sipped on the tea since I had some spare time on my hand and needed to exercise my mind) and made my order. The young man serving me immediately asked if I intended to sit around to study.

I commented that I’ll probably be around for half an hour and asked if it’ll be a problem. He replied that there’s an event downstairs and they anticipate a crowd so they discourage people from studying at the cafe. I kept quiet and took my receipt. I thought that the Large Chai Latte should at least buy me 30 minutes of time at the cafe.

This post is a discussion seeks to answer the question: “If a cafe wants to maximize their profits from a crowd and yet is limited by their available seats, how do they discourage people from studying there besides using attitude (which I assume is something I experienced)?” So before answering that question, we list out the factors that would encourage one to study at the cafe and see if there’s anything we can do to manipulate them:

Things that encourage one to study at a cafe

Good lighting; makes reading comfortable

Extremely hot or cold drink; takes you longer to drink and the taste of the drink don’t change that rapidly over time which means more excuse to stay at the cafe longer

Quality drinks; makes for nice beverage while you study and you probably won’t mind ordering one more and staying longer

Comfortable seats; allows you to study comfortably and sit at the cafe for longer time without feeling discomfort physically.

Technically speaking, removing any of these would help to reduce the time people stay around the cafe and also discourage studying. On the other hand, attitude (on part of the service staff) won’t help to reduce the determination of people studying at the cafe. In fact, it turns off people who genuinely just want to chill at the cafe for a while without discouraging the studying students. When one plans to stay at the cafe for a long time to study and sip on drinks, service at the counter makes up only a small part of the experience, whereas for customers who are interested to get a good drink and sit for a while, the service at the counter makes up half of the experience. In other words, giving people attitude is the worst possible solution compared to removing any of the above.

I strongly recommend the dimming of lighting, which doesn’t harm people out to relax but makes studying tedious and difficult. But this is not always possible since The Coffee Bean that I went to utilizes the in-building lighting that they probably have no control over. It’s not wise to compromise the quality of the drinks since it sends out the wrong messages and is disastrously difficult to control. That leaves us with modifying the seats.

I got this idea when I was at Saizeriya Restaurant at Liang Court; they feature a drinks bar where you pay about $6 bucks or so and get to drink lots of different drinks and it’s free flow – literally a drinks buffet. At first, I wondered why the seats there were so narrow and small; the cushion were thin and not exactly comfortable with prolonged sitting. Later I rationalized it as a means to get people out of the restaurant as soon as they’re done with the food. Of course it’s not going to put off people determined to try all the drinks, but at least they’ll finish with their affair and get out fast.

So here’s some prescription for cafes who hopes to attract people there for a drink and to sit around for a while (after all, if the cafe was empty you might think the drinks suck) but discourage students from spending their entire day studying there, the best move would be to adjust lighting according to your needs to adjust demand. Where this is not possible, modify the seats to make prolonged sitting uncomfortable.