Game theory is a useful technique for companies to rapidly respond to changes in products, technologies, and prices. It helps companies pay attention to interactions with competitors, customers, and suppliers, and induces companies to focus on the end-game so that their near-term actions promote their long-term interest by influencing what these players do. The theory is reasonably straightforward to use. There are two competitors, Ace and Smith. Ace expects Smith to enter the market and is trying to understand Smith’s likely pricing strategy. To do so, Ace uses something called a payoff matrix. Each quadrant in the matrix contains the payoffs or financial impact to each player for each possible strategy. If both players maintain prices at current levels, they will both be better off: Ace will earn $100 million and Smith will earn $60 million (Quadrant A). Unfortunately for both Ace and Smith, however, they have perverse incentives to cut prices. Ace calculates that if he maintains prices, Smith will cut prices to increase earnings to $70 million from $60 million. (See the arrow moving from Quadrant A to Quadrant B.) Smith makes a similar calculation that if she maintains prices, Ace will cut. The logic eventually drives them both to Quadrant D, with both cutting prices and both earning lower returns than they would with current prices in place.

This equilibrium is unattractive for both parties. If each party perceives this, then there is some prospect that each will separately determine to try to compete largely on other factors, such as product features, service levels, sales force deployment, or advertising. But it is necessary to have in-depth knowledge of the industry before game theory is truly valuable. Whether the goal is to implement by fully quantifying the outcomes of a payoff matrix or by more qualitatively assessing the outcome of the matrix, it is necessary to understand entry costs, exit costs, demand functions, revenue structures, cost curves, etc. Without that understanding, the game theory may not provide correct answers. The following are the rules to observe to make the best use of the theory:

• Examine the number, concentration, and size distribution of the players. Industries with four or fewer significant competitors have the greatest potential for using game theory to gain an edge because (a) the competitors will usually be large enough to benefit more from an improvement in general industry conditions than they would from improving their position at the expense of others, and (b) with smaller numbers of competitors it is possible for managers to think through the different combinations of moves and countermoves. Similarly, the number of customers, suppliers, etc., affects the usefulness of game theory.

• Keep an eye out for strategies inherent in one’s market share. Small players can use “judo economics” to take advantage of larger companies that may be more concerned with maintaining the status quo than with retaliating against a small entrant. In 1992, for instance, Kiwi Airlines got away with undercutting Delta’s and Continental’s prices between Atlanta and Newark by as much as 75 percent. The reason: When Kiwi first entered the market it represented less then 7 percent of that route’s capacity, and the cost of a significant pricing response by the incumbents would have likely exceeded the benefits. Conversely, large players can create economies of scale or scope. Companies such as United and American have used frequent-flier programs to create switching barriers, whereas most small airlines would not have the route structure required to make their frequentflier programs very attractive.

• Understand the nature of the buying decision. If there are only a few deals signed in an industry each year, it will be hard to avoid aggressive competition. In the jet engine industry, for example, three manufacturers (GE, Pratt & Whitney, and Rolls Royce) compete ruthlessly for scarce orders. If a producer loses several large bids in a row, layoffs will be likely, and it might even go out of business. In this kind of situation, the challenge for game theory is to improve the bidding process to shift the power balance between the industry and its customers.

• Scrutinize the competitors’ cost and revenue structures. Industries where competitors have a high proportion of fixed-to-variable cost will probably behave more aggressively than those where production costs are more variable. In the paper, steel, and refining industries, for example, high profit contributions on extra volume give most producers strong incentives to cut prices to get volume.

• Examine the similarity of firms. Industries where competitors have similar cost and revenue structures often exhibit independently determined but similar behavior. Consider the U.S. cellular telephone industry: The two providers in each market share similar technologies, and have similar cost structures. Given their similar economic incentives, the challenge is to find prices that create the largest markets and then to compete largely on factors such as distribution and service quality.

• Analyze the nature of demand. The best chances to create value with less aggressive strategies are in markets where demand is stable or growing at a moderate rate. For example, even in oil-field services in the early 1980s after drilling activity had plummeted, declining demand did not lead to lower prices in all sectors. In those more-technology-demanding parts of the industry where there were only a limited number of competitors (e.g. open-hole logging and well-pressure control), prices were more stable than in other sectors. Done right, game theory can turn conventional strategies on their heads and dramatically improve a company’s ability to create economic value. Sometimes it can increase the size of the pie; on other occasions it can make a company’s slice of the pie bigger, and it may even help do both.

DELPHI TECHNIQUE

The delphi technique, named after Apollo’s oracle at Delphi, is a method of making forecasts based on expert opinion. Traditionally, expert opinions were pooled in committee. The delphi technique was developed to overcome the weaknesses of the committee method. Some of the problems that occur when issues are discussed in committee include:

1. The influence of a dominant individual.

2. The introduction of a lot of redundant or irrelevant material into committee workings.

3. Group pressure that places a premium on compromise.

4. Reaching decisions is slow, expensive, and sometimes painful.

5. Holding members accountable for the actions of a group. All of these factors provide certain psychological drawbacks to people in face-to-face communication. Because people often feel pressure to conform, the most popular solution, instead of the best one, prevails. With the delphi technique, a staff coordinator questions selected individuals on various issues. The following is a sample of questions asked:

1. What is the probability of a future event occurring? For example, by what year do you think there will be widespread use of robot services for refuse collection, as household slaves, as sewer inspectors, etc.? a. 2000 b. 2010 c. 2020 d. 2030

2. How desirable is the event in Question 1? a. needed desperately b. desirable c. undesirable but possible

3. What is the feasibility of the event in Question 1? a. highly feasible b. likely c. unlikely but possible

4. What is your familiarity with the material in Question 1? a. fair b. good c. excellent The coordinator compiles the responses, splitting them into three groups: lower, upper, and inner. The division into groups may vary from one investigation to another. Frequently, however, the lower and upper groups each represent 10 percent, whereas the inner group takes the remaining 80 percent. When a person makes a response in either the upper or lower group, it is customary to ask about the reasons for his or her extreme opinion. In the next round, the respondents are given the same questionnaire, along with a summary of the results from the first round. The data feedback includes the consensus and the minority opinion. During the second round, the respondents are asked to specify by what year the particular product or service will come to exist with 50 percent probability and with 90 percent probability. Results are once again compiled and fed back. This process of repeating rounds can be continued indefinitely; however, rarely has any research been conducted past the sixth round. In recent years, the delphi technique has been refined by the use of interactive computer programs to obtain inputs from experts, to present summary estimates, and to store revised judgments in data files that are retrievable at user terminals. The delphi technique is gradually becoming important for predicting future events objectively. Most large corporations use this technique for long-range forecasting. Some of the advantages of the delphi technique are listed below:

1. It is a rapid and efficient way to gain objective information from a group of experts.

2. It involves less effort for a respondent to answer a well-designed questionnaire than to participate in a conference or write a paper.

3. It can be highly motivating for a group of experts to see the responses of knowledgeable persons.

4. The use of systematic procedures applies an air of objectivity to the outcomes.

5. The results of delphi exercises are subject to greater acceptance on the part of the group than are the consequences arrived at by more direct forms of interaction.

Delphi Application Change is an accepted phenomenon in the modern world. Change coupled with competition forces a corporation to pick up the trends in the environment and to determine their significance for company operations. In light of the changing environment, the corporation must evaluate and define strategic posture to be able to face the future boldly. Two types of changes can be distinguished: cyclical and developmental. A cyclical change is repetitive in nature; managers usually develop routine procedures to meet cyclical changes. A developmental change is innovative and irregular; having no use for the “good” old ways, managers abandon them. Developmental change appears on the horizon so slowly that it may go unrecognized or be ignored until it becomes an accomplished fact with drastic consequences. It is this latter category of change that assumes importance in the context of strategy development. The delphi technique can be fruitfully used to analyze developmental changes. Functionally, a change may fall into one of the following categories: social, economic, political, regulatory, or technological. The delphi technique has been used by organizations to study emerging perspectives in all these areas.

One drawback of the delphi technique is that each trend is given unilateral consideration on its own merits. Thus, one may end up with conflicting forecasts; that is, one trend may suggest that something will happen, whereas another may lead in the opposite direction. To resolve this problem, another forecasting technique, the cross-impact matrix (discussed later) has been used by some researchers. With this technique, the effect of potential interactions among items in a forecasted set of occurrences can be investigated. If the behavior of an individual item is predictable (i.e., if it varies positively or negatively with the occurrence or nonoccurrence of other items), the cross-impact effect is present. It is thus possible to determine whether a predicted event will have an enhancing or inhibiting influence upon each of the other events under study by using a crossimpact matrix. Recent research shows that the use of the delphi technique has undergone quite a change. The salient features of the revised delphi technique are (a) identifying recognized experts in the field of interest; (b) seeking their cooperation and sending them a summary paper on the topic being examined (based on a literature search); and (c) conducting personal interviews with each expert based on a structured questionnaire, usually by two interviewers. Feedback and repeated rounds of responding to written questionnaires are no longer considered necessary.