9781118041581

A decision maker should acquire costly information if and only if the expected

value of the information exceeds its cost.

Before leaving this simple example, let’s check exactly where the

increase in expected profit came from. Refer once again to Figure 13.1. As

the tree indicates, only good sites are drilled. But now consider the effect if

all sites, good and bad, were drilled, as would be the case if the test infor-

mation were not available. From the tree, we see that the partners would

lose $200,000 from drilling bad/dry sites. Thus, the advantage of the test is

that it saves this amount by screening out these sites. This savings occurs 60

percent of the time because this is the frequency of dry sites. The partner’s

expected gain from the seismic test (compared to always drilling) is

(.6)(200,000) $120,000. This is exactly the EVI calculated earlier. The test

allows the partners to resolve the uncertainty, wet or dry, before committing

to a decision—to drill or not to drill. In this way, they save the cost of drilling

dry sites.

Imperfect Information

Although illustrative, the preceding example is unrealistic because it is unlikely

that the partners ever could obtain perfect information before drilling. We now

consider the decision to drill in light of an imperfectseismic test. Again, we

assume the test results are categorized as “good” or “bad,” but now the test is

imperfect. The partners are aware of the recent record of test outcomes, listed

in Table 13.1. We address the same questions as before: Should the partners

invest in the test, and, if so, what drilling decision should they make based on

its result?

Table 13.1 provides a record of 100 past sites (judged to be roughly simi-

lar to the current site) where seismic tests have been conducted. It provides a

two-way classification of each site’s outcome: the result of the test (good or

bad) versus the true state of the site (wet or dry). The top left-hand entry shows

544 Chapter 13 The Value of Information

TABLE 13.1

Past Seismic

Test Record (100 Sites) Actual State of the Site

Wet (W) Dry (D) Total

Seismic Good (G) 30 20 50

Result Bad (B) 10 40 50

Total 40 60 100

c13TheValueofInformation.qxd 9/26/11 11:02 AM Page 544