9781118041581

tree, which incorporates the seismic results. We start by emphasizing the

sequence of events in the tree. As in Figure 13.1, the first event is the test

result: good or bad. This is represented by the chance node (the circle) from

which the possible test results emanate. After seeing the result, the partners

must decide whether or not to drill. These decisions are denoted by the two

squares appearing on the good and bad test branches. Finally, for the “drill”

option, the tree shows the chance event, wet or dry, along with the revised

probabilities. After a good test, striking oil carries a 60 percent chance; after

a bad test, the chance is only 20 percent. (In contrast to the case of a perfect

seismic test, a good result does not guarantee oil, nor does a bad result rule

it out.)

With the decision tree in hand, the optimal decision strategy is easy to

determine. Consider the upper decision square. If the test is good, drilling

offers an expected profit of $280,000 and should be undertaken. After a bad

test, drilling has an expected profit of $40,000 (since the chance of finding

oil is only 20 percent). Thus, the partners should employ a contingent strategy:

Drill the site if the seismic test is favorable and forgo it if the test is unfavorable.

How much profit do they stand to gain using this strategy? To answer this ques-

tion, we simply calculate the expected profit at the initial chance node, before

the test outcome is known. As we noted earlier, a good test is expected to occur

50 percent of the time, in which case drilling earns an expected profit of

$280,000. The other 50 percent of the time, a bad test occurs, no drilling takes

place, and the profit is $0. Thus, the expected profit at the initial chance node

546 Chapter 13 The Value of Information

FIGURE 13.2

An Imperfect Seismic

Test

A “good” seismic test

boosts the chance of

striking oil to .6. A

“bad” seismic test

lowers it.

Do not drill

Drill

Do not drill

Good test Drill

Bad test

0

280

.5

.5

140

Imperfect

test

–$200

$600

Wet

Dry

Wet

Dry

0

0

–$200

$600

–40

280

.2

.8

.6

.4

c13TheValueofInformation.qxd 9/26/11 11:02 AM Page 546