Course Outline

Teaching Philosophy

Grading

Daily Assignments

Session  1: Monday, January 18/19
Session  2: Wednesday, January 20/21
Session  3: Monday, January 25/26
Session  4: Wednesday, January 27/28
Session  5: Monday, February 1/2
Session  6: Wednesday, February 3/4
Session  7: Monday, February 8/9
Session  8:Wednesday, February 10/11
Session  9: Monday, February 15/16
Session 10: Wednesday, February 17/18
Session 11: Monday, February 22/23
Session 12: Wednesday, February 24/25
Session 13: Monday, March 1/2
Session 14: Wednesday, March 3/4
 

Course Nature and Purpose:

 Most of what managers do is make decisions. This course aims to improve your decision-making ability. There are two essential ingredients to making good decisions. One is making the best choice among the alternatives. Equally important and less understood is getting the right alternatives to be put under consideration.

 This course will present a systematic approach to decision making. We will concentrate on the analytic, decision tree method, worked out thoroughly from assumptions and basic principles to increasingly advanced applications. This approach is most useful when one is faced with a complex environment. The basic idea is to take a complex environment and break the problem into pieces, use judgment, intuition, and facts to analyze the pieces, and then reassemble the pieces into a logically coherent whole. Thus this method is well suited to the worlds of business and personal life, for in both cases many factors and personalities affect even the simplest decisions.

 One important source of complexity is uncertainty, which arises because of future events (prices, market demand, new technologies, fickleness of consumers, political elections) whose outcome is unknown at the time a decision must be made. A second important source of complexity is strategic interaction, which arises whenever one's own outcomes depend upon the choices of other decision makers, whose attributes, motives, and decisions are unknown. We first apply the analytic method to uncertainty with the help of probability theory, and then extend the method to the case of strategic interaction, including the specific issues that arise in negotiations.

 Our goal is to help you become a better decision maker, and to help you understand the importance of developing a coherent approach to problem formulation, judgment, and strategic planning.

 Two further elements round out our approach. At intervals, we will consider the psychological factors that either impede or contribute towards the effort to make good decisions. And we will consistently strive to link our methods with real world applications. Like many things, this has become easier with technology. Thus, new this year to the course is the integration of Excel plugins into the assignments.
 

 The Course is divided into two parts, with the midterm and spring break separating the two.

  Part 1. Single Person Decision Analysis.

 a.    The Organization and Evaluation of Decision Problems. (4 classes) Decision trees, strategies, expected value, rollback.

b.    Probability and its Application to Decision Analysis. (6 classes) Review of basic probability, Bayes' Law, reversal of conditioning, the value of information, sensitivity analysis, modeling, Monte Carlo analysis, conditional independence..

c.    Utility and Risk. (4 classes) Attitudes towards risk, certainty equivalents, constructing preference curves, expected utility analysis, psychological anomalies.

d.    Review and Midterm. (2 classes)
 
Part 2. Strategic Interaction.

a.    Introduction to Game Theory. (4 lectures) How to frame decision problems with more than one decision-maker. The definition of equilibrium, rollback, imperfect information, bayesian updating, applications to models of signaling and entry.

b.    Applications of Game Theory. (4 lectures) Applications to auctions, pricing, voting, and cooperation.

c.    Bargaining and Negotiations. (3 classes) Game theory and psychology of bargaining and negotiating, in-class workshops.
 
Grading

There will be a midterm, a final, and frequent problem sets. Your grade will be based on your performance in each of these three dimensions. Your homework will be particularly relevant if you are on the border between two grades.

 Teaching Philosophy

 Before the course begins, I think it is important that you have some understanding of my teaching philosophy. I like to teach using an interactive or Socratic format. I will try to make the material lively and interesting. I will work to find problems that are relevant to business and to your life more broadly.

Your active contribution is what makes the class discussions sophisticated and lively, so reading and thinking before class, and debating during class are important. I will actively call on people. I expect you to come to class prepared. Accidents and emergencies do happen. I don't need to know about them and don't need to hear an explanation. If you are not prepared, leave me a note at the front of the class or speak to me beforehand and I won't call on you. As your classmates will tell you, remembering names is not my forte. Please bring your namecard to every class.

  My goal is to present you with what I consider to be fundamental tools for any manager to make decisions. Learning new material, especially formal or mathematical techniques, can be difficult, even painful. It means giving up some old assumptions about how to approach problems and it means learning how to use rather than memorize formulas. The hardest part is not solving equations but knowing how to set up the problem. Many of the problems I will give you are deliberately presented in an awkward way. That's the way life is. Just like reading comprehension is something you learn by reading, a structured approach to decision making is something you learn through practice.

  Much of our classroom experience will be focussed on working through and practicing setting up problems and case studies. When I call on you and you do not know the answer, my guess is that you will feel uncomfortable. This is not my goal upon calling on you and persisting. My goal is to make certain that you understand the material. Thus, I will continue discussing the problem with you until I believe that you understand what is going on. In fact, once you have shown that you understand the answer, I am likely to move on to someone else who may be having more trouble. I do not want people getting lost. Much of what we will do builds on what we have done. Fortunately, everything we do in this course, we will do more than once, so if you do not understand it completely the first time, you will have a second and maybe more chances.

  If you think the material is incomprehensible, please come talk to me and it is my responsibility to do something about it. Do not wait until you are lost before seeing me. To help me to gauge your progress, there will be at least one assignment due each week (and sometimes one for each class). Experience has shown that people who do not work in groups have a much much much harder time understanding the material and solving the problems. I have not gone so far as to assign people to groups. If you are not part of a study group, I believe you are making a big mistake.

  Here's to making better decisions.

 Daily Assignments

The most important materials for this course are the classroom activities. The concepts which you are expected to learn in this course will be covered in the readings, lecture notes, and in class discussion; you are responsible for everything in the notes and everything which arises in discussion in class, except where specifically told otherwise.

There are three types of assignments.

1.    READ --- this refers to reading to be done before class. All readings are from the Holloway notes unless otherwise indicated.

2.    PREPARE --- this refers to materials to be prepared for class that day; prepare ahead. The usefulness of our discussions, and of the entire class period, depends on your preparation.

3.    HOMEWORK --- this refers to materials to be turned in on the day due. These will be looked at by a grader and returned to you. There will be a few homework assignments which are especially important, and I will indicate these to you.
 
Some problem sets will have two types of questions, those that you should be able to do on your own (so as to make sure that you understand the material) and more challenging problems that you should work on in a group.

 Any cases or numbered assignments which you have not yet received will be handed out in class. As a rule, check for handouts each class. All handouts, problem sets, and solutions will be posted to our website. As a second rule, always review all solutions to the assignments and make sure you understand them.

  NOTE: The reading list for this course, handouts, problem sets, an online exercises can be found at http://mayet.som.yale.edu/coopetition/teaching97/CC1997.html.

Session 1: Monday, January 18/19. Introduction and Overview; Better decision-mkaing through science; A Simple Decision Tree.

IN CLASS: Survey Questions, knives, and 2,4,6.
READ Holloway chapter 1.
At end of class, please form a study group.

Session 2: Wednesday, January 20/21. The Analytic Method; The Formal Decision Tree Approach.

READ: Holloway chapter 3 (first half);
READ: Hugh Courtney, Jane Kirkland, and Patrick Viguerie, "Strategy Under Uncertainty,'' Harvard Business Review.

HOMEWORK (due at beginning of class): 1. Please turn in the names in your study group;
2. Gore's decision tree.
PREPARE Monday Morning Quarterback

Session 3: Monday, January 25/26. Seeing the Trees from the Forest. Decision Trees, continued.

There are 5 assignments for class 3:

1. READ Holloway chapter 3 (second half); Swersey Chapter 3. 

2. Familiarize yourself with tree.xls. Read Treeplan documentation. 

3. HOMEWORK due Tyler (A). All cases due (at beginning of class). Draw the tree. Write down a complete list of strategies, checking the list twice. 

4. PREPARE Wildcat (A1). 

5. PREPARE Medical News Network. 

Session 4: Wednesday, January 27/28. Criteria for Making Decisions; Strategies. 

READ Holloway chapter 5, pages 109-113, 116-118.
HOMEWORK due: Warren Agency. 

Session 5: Monday, February 1/2. Review of Probability: Conditional Probability, Probability Tables, Independence, Probabilities on Trees, Bayes' Law and Reversal of Conditioning (Tree Flipping). 

REREAD Swersey chapter 2 (this should be a review from Data).
READ Introduction to Probability and Simulations in Spreadsheets.
PREPARE Assignment #1, question 1; Assignment #3. (No assignment #2.)
HOMEWORK due: Tyler (B) and Tyler (C), Probability problems A. 

Session 6: Wednesday, February 3/4. Value of Information. 

READ Stokey and Zeckhauser "A Primer for Policy Analysis" pp. 219--236;
READ Swersey chapter 4
HOMEWORK due: Redding (A), Probability problems B. 

Session 7: Monday, February 8/9. Sensitivity Analysis, Tornado Diagrams. 

READ Holloway chapter 14.
HOMEWORK due: Redding (B). Familiarize yourself with tornado.xls. 

Session 8: Wednesday, February 10/11. Monte Carlo Analysis. 

READ Holloway chapter 9, pages 206-214; chapter 15.
PREPARE Scotia case.
HOMEWORK due: Wildcat (A2), Assignment #4. (Those who want more practice on decision analysis and the value of information should work through Ramirez Farms (problems 14.6 and 14.7 in the Holloway Notes.)  

Session 9: Monday, February 15/16. Incorporating Risk into Decision-Making; Attitudes Towards Risk; Certainty Equivalents; Compound Risk. 

READ Frank, Microeconomics and Behavior/, Chapter 6;
READ"The Next American Tort Crisis," it WSJ/, 12/28/89.
READ Holloway chapter 5, pages 100-108;
HOMEWORK due: Tyler (D). (For those who want practice, try Tyler (E).) 

Session 10: Wednesday, February 17/18. Constructing Certainty Equivalents; Assessing Preference Curves; EU; EU Rollback; Why EU Works.

PREPARE Assignment #7.
READ Holloway chapter 6, pages 128-136;
READ Stokey and Zeckhauser, "A Primer for Policy Analysis", pp. 237--254. 

Session 11: Monday, February 22/23. The Allais Paradox; Loss Aversion; Reference Points. 

READ Frank, Microeconomics and Behavior/, Chapter 8.
READ Shiller, "Human Behavior and the Efficiency of the Financial System" 

Session 12: Wednesday, February 24/25. Learning, Procrastination, Flexibility, Commitment, and the Sunk Cost Fallacy. 

PREPARE: Go to the Internet and do the online exercise.
READ: Staw and Ross, "Understanding Behavior in Escalation Situations". 

Session 13: Monday, March 1/2. Modeling; In Class practice problems; Review for Midterm.  

READ: Geanakoplos and Gray, "When Seeing Further is not Seeing Better;"
HOMEWORK due: Modeling problems. Toro Snowblower. 

Session 14: Wednesday, March 3/4. NO CLASS: MIDTERM To Be Arranged.
 

 Session 2. Gore's Decision Tree

 In July, 1991, Senator Gore calls you into his office and asks for some advice. He understands that you are an expert in decision-making. He would like some help regarding his decision about whether or not to run for president in the 1992. His objective is to become president of the United States. He is concerned that the Republicans will win the election if the Democratic primary is contentious and not quickly resolved. (A long primary exhausts financial resources and leads candidates to make too many promises to special interest groups which are costly in the general election.) To do this assignment, do a Lexis/Nexis search and read some of the newspaper stories at the time about the political environment. (Bush had a 90+% approval rating following the victory in the Gulf war.) Then draw a decision tree to present to Senator Gore.

 Session 3. Medical News Network

 Prepare the following case: Entrepreneur Chris Whittle is best known for the Edison Project, a system of for-profit public schools, and for Channel One, which beams educational programs along with commercials into school classrooms. Another of Whittle's ventures was Medical News Network (MNN). Transmitted directly into doctors' offices, MNN carried programs of special interest to the medical profession--news of recent clinical developments, discussions of healthcare policy, advice on office management techniques, and the like--along with sponsored messages from pharmaceutical companies.  

Whittle hoped that the drug makers would find MNN a much more effective way to reach doctors than the old approach of sending out "detail men" to make personal presentations. He persuaded Roy Vagelos, Merck's then CEO, to invest $17 million to help MNN get off the ground. 

MNN began broadcasting in March 1993. The test got off to a ow start, as fewer physicians than expected signed up to receive MNN. Of the drugs advertised on the network, most showed some increases in sales, but some registered declines. Still, Whittle pronounced the experiment a great success.

In November, the trial period specified in the contract between Whittle and Merck came to an end. Whittle declared that it was now time for a verdict. MNN would make ten shares available for sale at roughly $33 million each. From that point on, only pharmaceutical companies owning shares in MNN would be able to advertise on the network. Companies who are not part of this network will be at a significant disadvantage in defending their market share and in introducing new drugs. 

Merck had been working with MNN from the outset to help make the concept a reality. But, given the ambiguous results of the test, the team wasn't convinced that the network was effective, never mind cost-effective. Merck was considering buying three shares in MNN, and brings you in to help evaluate the situation. You don't feel comfortable investing another $100 million in MNN without more data. Although you might want to finance further trials, Whittle rejects that option. He'd rather sell shares in MNN to someone else today.

Draw a decision tree facing Merck. Why might you proceed even if the expected value going forward is negative (even ignoring sunk costs)? Think out of the box.