ODD

Answers to Odd-Numbered Problems, 4th Edition of Games and Information,

Rasmusen

PROBLEMS FOR CHAPTER 2: INFORMATION

26 March 2005. 6 September 2006. Http://[Link].

This file contains answers to the odd-numbered problems in the gourth edition of Games
and Information by Eric Rasmusen, which I am working on now and perhaps will come
out in 2005. The answers to the even- numbered problems are available to instructors or
self-studiers on request to me at Erasmuse@[Link].

Other books which contain exercises with answers include Bierman & Fernandez
(1993), Binmore (1992), Fudenberg & Tirole (1991a), J. Hirshleifer & Riley (1992), Moulin
(1986), and Gintis (2000). I must ask pardon of any authors from whom I have borrowed
without attribution in the problems below; these are the descendants of problems that I
wrote for teaching without careful attention to my sources.

200
 PROBLEMS FOR CHAPTER 2: INFORMATION

2.1. The Monty Hall Problem (easy)

You are a contestant on the TV show, “Let’s Make a Deal.” You face three curtains,
labelled A, B and C. Behind two of them are toasters, and behind the third is a Mazda
Miata car. You choose A, and the TV showmaster says, pulling curtain B aside to reveal a
toaster, “You’re lucky you didn’t choose B, but before I show you what is behind the other
two curtains, would you like to change from curtain A to curtain C?” Should you switch?
What is the exact probability that curtain C hides the Miata? Answer. You should switch
to curtain C, because:

Prob(Host chose B | Miata behind C)Prob(Miata behind C)

Prob (Miata behind C | Host chose B) = Prob(Host chose B)

(1)( 31 )
= (1)( 3 )+( 12 )( 13 )
1 .

= 23 .

The key is to remember that this is a game. The host’s action has revealed more than
that the Miata is not behind B; it has also revealed that the host did not want to choose
Curtain C. If the Miata were behind B or C, he would pull aside the curtain it was not
behind. Otherwise, he would pull aside a curtain randomly. His choice tells you nothing
new about the probability that the Miata is behind Curtain A, which remains 13 , so the
probability of it being behind C must rise to 32 (to make the total probability equal one).

What would be the best choice if curtain B simply was blown aside by the wind,
revealing a toaster, and the host, Monty Hall, asked if you wanted to switch to Curtain
C? In that case you should be indifferent. Just as easily, Curtain C might have blown
aside, possibly revealing a Miata, but though the wind’s random choice is informative–
your posterior on the probability that the Miata is behind Curtain C rises from 1/3 to
1/2— it does not convey as much information as Monty Hall’s deliberate choice.

See [Link] for a Java applet

on this subject.

2.3. Cancer Tests (easy) (adapted from McMillan [1992, p. 211])

Imagine that you are being tested for cancer, using a test that is 98 percent accurate. If
you indeed have cancer, the test shows positive (indicating cancer) 98 percent of the time.
If you do not have cancer, it shows negative 98 percent of the time. You have heard that
1 in 20 people in the population actually have cancer. Now your doctor tells you that you
tested positive, but you shouldn’t worry because his last 19 patients all died. How worried
should you be? What is the probability you have cancer?

201
Answer. Doctors, of course, are not mathematicians. Using Bayes’ Rule:
P rob(P ositive|Cancer)P rob(Cancer)
P rob(Cancer|P ositive) = P rob(P ositive)

0.98(0.05)
= 0.98(0.05)+0.02(0.95)
(1)

≈ 0.72.

With a 72 percent chance of cancer, you should be very worried. But at least it is not 98
percent.
Here is another way to see the answer. Suppose 10,000 tests are done. Of these, an
average of 500 people have cancer. Of these, 98% test positive on average— 490 people.
Of the 9,500 cancer-free people, 2% test positive on average—190 people. Thus there are
680 positive tests, of which 490 are true positives. The probability of having cancer if you
test positive is 490/680, about 72% .

This sort of analysis is one reason why HIV testing for the entire population,
instead of for high-risk subpopulations, would not be very informative— there would be
more false positives than true positives.

2.5. Joint Ventures (medium)

Software Inc. and Hardware Inc. have formed a joint venture. Each can exert either high
or low effort, which is equivalent to costs of 20 and 0. Hardware moves first, but Software
cannot observe his effort. Revenues are split equally at the end, and the two firms are risk
neutral. If both firms exert low effort, total revenues are 100. If the parts are defective,
the total revenue is 100; otherwise, if both exert high effort, revenue is 200, but if only one
player does, revenue is 100 with probability 0.9 and 200 with probability 0.1. Before they
start, both players believe that the probability of defective parts is 0.7. Hardware discovers
the truth about the parts by observation before he chooses effort, but Software does not.

(a) Draw the extensive form and put dotted lines around the information sets of Software
at any nodes at which he moves.
Answer. See Figure A2.1. To understand where the payoff numbers come from, see
the answer to part (b).

202
 Figure A2.1: The Extensive Form for the Joint Ventures Game

(b) What is the Nash equilibrium?

Answer. (Hardware: Low if defective parts, Low if not defective parts; Software:
Low).

100
πHardware (Low|Def ective) = = 50.
2
Deviating would yield Hardware a lower payoff:
100
πHardware (High|Def ective) = − 20 = 30.
2

100
πHardware (Low|N ot Def ective) = = 50.
2
Deviating would yield Hardware a lower payoff:
   
100 200
πHardware (High|N ot Def ective) = 0.9 + 0.1 − 20 = 45 + 10 − 20 = 35.
2 2

100
πSof tware (Low) = = 50.
2
Deviating would yield Software a lower payoff:
      
100 100 200
πSof tware (High) = 0.7 +.3 0.9 + 0.1 −20 = 35+0.3(45+10)−20.
2 2 2

203
This equals 15 + 0.3(35) = 31.5, less than the equilibrium payoff of 50.
Elaboration. A strategy combination that is not an equilibrium (because Software
would deviate) is:
(Hardware: Low if defective parts, High if not defective parts; Software: High).

100
πHardware (Low|Def ective) = = 50.
2
Deviating would indeed yield Hardware a lower payoff:
100
πHardware (High|Def ective) = − 20 = 30.
2

200
πHardware (High|N ot Def ective) = − 20 = 100 − 20 = 80.
2
Deviating would indeed yield Hardware a lower payoff:
   
100 200
πHardware (Low|N ot Def ective) = 0.9 + 0.1 = 55.
2 2
   
100 200
πSof tware (High) = 0.7 + 0.3 − 20 = 35 + 30 − 20 = 45.
2 2
Deviating would yield Software a higher payoff, so the strategy combination we are
testing is not a Nash equilibrium:
      
100 100 200
πSof tware (Low) = 0.7 +0.3 0.9 + 0.1 = 35+0.3(45+10) = 35+16.5 = 51.5
2 2 2

More Elaboration. Suppose the probability of revenue of 100 if one player choose
High and the other chooses Low were z instead of 0.9. If z is too low, the equilib-
rium described above breaks down because Hardware finds it profitable to deviate to
High|N ot Def ective.
100
πHardware (Low|N ot Def ective) = = 50.
2
Deviating would yield Hardware a lower payoff:
   
100 200
πHardware (High|N ot Def ective) = z +(1−z) −20 = 50z+100−100z−20.
2 2
This comes to be πHardware (High|N ot Def ective) = 80 − 50z, so if z < 0.6 then
the payoff from (High|N ot Def ective) is greater than 50, and so Hardware would
be willing to unilaterally supply High effort even though Software is providing Low
effort.
You might wonder whether Software would deviate from the equilibrium for some
value of z even greater than 0.6. To see that he would not, note that
      
100 100 200
πSof tware (High) = 0.7 + 0.3 z + (1 − z) − 20.
2 2 2

204
 This takes its greatest value at z = 0, but even then the payoff from High is just
0.7(50) + 0.3(100) − 20 = 45, less than the payoff of 50 from Low. The chances of
non-defective parts are just too low for Software to want to take the risk of playing
High when Hardware is sure to play Low.
This situation is like that of two people trying to lift a heavy object. Maybe it is
simply too heavy to lift. Otherwise, if both try hard they can lift it, but if only one
does, his effort is wasted.

(c) What is Software’s belief, in equilibrium, as to the probability that Hardware chooses
low effort?
Answer. One. In equilibrium, Hardware always chooses Low.

(d) If Software sees that revenue is 100, what probability does he assign to defective parts
if he himself exerted high effort and he believes that Hardware chose low effort?
0.7
Answer. 0.72 (= (1) (1)(0.7)+(0.9)(0.3) ).

2.7. Smith’s Energy Level (easy)

The boss is trying to decide whether Smith’s energy level is high or low. He can only look in
on Smith once during the day. He knows if Smith’s energy is low, he will be yawning with
a 50 percent probability, but if it is high, he will be yawning with a 10 percent probability.
Before he looks in on him, the boss thinks that there is an 80 percent probability that
Smith’s energy is high, but then he sees him yawning. What probability of high energy
should the boss now assess?

Answer. What we want to find is P rob(High|Y awn). The information is that P rob(High) =
.80, P rob(Y awn|High) = .10, and P rob(Y awn|Low) = .50. Using Bayes Rule,
P rob(High)P rob(Y awn|High)
P rob(High|Y awn) = P rob(High)P rob(Y awn|High)+P rob(Low)P rob(Y awn|Low)

(0.8)(0.1)
= (0.8)(0.1)+(0.2)(0.5)
= 0.44.

205