Thursday, May 15, 2003

Of Hands and Cookie Jars. 9:21:14 PM    comment []

Apparently a student was just asked to leave the MBA program after their P4 exam results were found to be less than.. shall we say... independently generated.  "Flagrant cheating" was the message that the Dean used in his message to the student body. I am glad that the appropriate message has been sent, and that the situation has been used as an opportunity to reinforce the expected behavioral norms of school.  Perhaps most people take the ethical character of their colleagues for granted; this is a reminder that there are always those who will cut corners, and that those transgressions are taken seriously, at least at this institution.  For all the numerous flaws in the education here, we do know that the ethical standards are maintained at a high level.   8:06:24 PM    comment []

Two sample negotiation problems from Nego Analysis. 1.  Bidding for $1 A bill of one dollar is up for competitive sealed bidding.  You are bidding for this $1 prize against N opponents chosen at random from this class.  In the table below and on the computer, fill in your bid as a function of the number of opponents you face.  Your strategy will be pitted against those of all other class members.  Your raw score will be the average payoff from all these contests. Your payoff for a winning bid b (in cents) is 100-b.  Six rounds will be played, with the number of bidders varying as N=1, 2, 3, 5, 10, 25. What would you bid for each round?   2.  Vaxco-Dambit In this bidding exercise, you will submit a sealed bid for a prize of uncertain value.  It is equally likely that the value of the prize is $0, $10, or $20.  There is one other bidder. Lucky constructed a model (and yes, this is all the information that you are given), ran the numbers and believes that the best bid is $5.  What do you think? In a variation, assume that Dambit knows the value of the prize, and Vaxco does not.  As Vaxco, what is your bid now?  Lucky thinks it's $3.33, but really has no idea. As Dambit, what bid do you submit when the prize is $0?  $10?  $20?   12:16:28 AM    comment []

Ingemar Dierickx's Quotes of the Day from Negotiation Analysis (commenting on a 1-cent bid, the lowest possible in a game where high bids win) "Why do we see bids of one cent?"  student: "Collusion!"  Prof: "In true form, an Italian suggests this!" When modeling anything, the uniform distribution is better than the normal distribution because "it captures humility and ignorance". "Unaided intuition does not help solve these problems." "Developing a game-theoretic model to a real-life situation is well beyond the capability of anyone in this room."  There were 75 people in the room. (continuing a long tradition of denigrating the Organizational Behavior department at school) "This is not an OB class, where anything goes." (reminding people that they can talk their way into supporting any particular position with words unsupported by quantitative evidence) "There is a saying in French: on peut prouver tout chose, et son contraire." (speculating as to the required number of people choosing a Nash equilibrium in a multiplayer auction forcing all others to adopt the same strategy) "How many mathematicians does it take to ruin a party?  One."  [Ed. - actually, this number is about 5 in a class of 75.] (following a long monologue on how some situations are structured and some are unstructured, and how the world of espionage is of the latter aspect, and again returning to the hard science theme) "This course contends that OB largely does not matter, except for spy stories.  In spy stories, the OB dimension is critical." What did I learn in this class? Always expect variance.  People are different. Do not expect other people to do what you find rational.  Other people are not like you. Think in terms of pictures.  Optimize pictures and you will gain a greater understanding of the problem you are trying to solve. Most important: do not try to reason your way through these problems.  There is only one way to reach a conclusion: crank the numbers!   12:07:02 AM    comment []