September 29, 2015 Decision Analysis can be used to determine the optimal strategies when a decision-maker is faced with several decision alternatives and an uncertain future events. Possibilities to consider Certainty Uncertainty Risk It is the product of the probability that the event will occur and the amount to be received upon such occurrence. Let P be the probability value and X be the amount of money. EV = P(X) If an event has several possible outcomes with probability P1, P2, ..., Pn and if X denotes a discrete variable which can assume the values X1, X2, ..., Xn then EV = P1X1 + P2X2 +...+ PnXn A fair coin is tossed. If the coin lands heads, Chito will receive P6, and pay P4 if it lands tails. Find the EV. EV = (0.50)(6) + (0.50)(-4) = P1 This means that the game is fair for Chito Consider the following game of chance. Mark pays P200 and roll a fair die. Then he receives a payment according to the following schedule. If the event A={1,2,3} occurs, then he will receive P100. If the event B={4,5} occurs, Mark receives P200. If the event C = {6} occurs, then he will receive P600. What is the average profit he can make if he participates in this game? If A occurs , then a profit will be 100-200=-100 (Mark will lose P100) If B occurs, a profit will be 200-200=0 If C occurs, a profit will be 600-200=400 Therefore we may compute the average profit as follows: ave.-profit = (3/6)(-100) + (2/5)(0) + (1/6)(400) = 50/3 That is , Mark can expect to make P16.67 on the average every time he plays this game. The ABC construction operations manager has to decide whether to accept a bid or not. If the manager accepts the bid, the construction company may gain p3.5 million if it succeeds, or lose P2.5 million if it fails. The probability that it will succeed is 30%. Find the EV if the company accepts the bid. The ABC construction operations manager has to decide whether to accept a bid or not. If the manager accepts the bid, the construction company may gain p3.5 million if it succeeds, or lose P2.5 million if it fails. The probability that it will succeed is 30%. Find the EV if the company accepts the bid. P1 = 30% P2 = 70% X1 = 3.5 X2 = -2.5 EV = -0.7 This means that the company is expected to lose. The decision-maker is to pick the alternative with the best payoff for the known event Payoff – consequence resulting from a specific combination of a decision alternative and a state of nature Best alternative Highest payoff if the payoffs are expressed in profits Lowest payoff if the payoffs are expressed as costs. Consider the payoff table which illustrates a capacity planning problem. It is helpful in selecting among alternatives because they facilitate comparison of alternatives Possible Future Demand Alternative Low High Small Capacity 300 370 Large Capacity 180 900 Present values in thousand pesos The payoffs are the present values of future revenues minus the costs for each alternative in each event. What is the best choice if future demand will be low? The best choice is the one with highest payoff. P300,000 (leads to building small facility) Assumption: Managers can list the possible events but cannot estimate their probabilities. 1. Conditions that will be used in decision rules: Pessimistic Approach Choose the alternative that is the “best of the worst.” It takes into account only the worst possible outcome for each alternative. Maximization : maximin Minimization : minimax Example: Refer to the previous table Alternative Small facility Large facility Worst Payoff 300 180 P300,000 is the worst member, the pessimist will build a small facility. Conditions that will be used in decision rules: 2. Optimistic “best of the best” “go for it” strategy that has high expectations Maximization : maximax Minimization : minimin Since P900,000 is the best member, the optimistic would build a large facility. Conditions that will be used in decision rules: 3. Laplace Choose the alternative with the best “weighted payoff” Treats the state of nature as equally likely to each event Alternative Small facility Large facility Weighted Payoff 0.50(300) + 0.50(370) = 335 0.50(180) + 0.50(900)=540 Since P540,000 is the best weighted payoff, the realist would build a large facility. 4. Minimax Regret Choose the alternative with the best “worst regret.” It seeks to minimize the difference between the given payoff and the best payoff for each state of nature. For profit: Regret Value = Highest column entry-every column entry For cost: Regret Value=Entry every column-lowest column entry Regret Alternative Low Demand High Demand Maximum Regret Small facility 300-300=0 900-370=530 530 Large facility 300-180=120 900-900=0 120 Pick a large facility to minimize the maximum regret. The manager has less information than with decision-making under certainty but more information than with decision-making under uncertainty. A widely used approach circumstances is the EV under such EV is the sum of the payoffs for an alternative where each payoff is weighted by the probability for the relevant state of nature. state of nature – possible outcomes for a chance event Which is the best alternative if the probability of low demand is estimated to be 0.40 and the probability of high demand is estimated to be 0.60? Possible Future Demand Alternative Low High EV Small Facility 300 370 0.4(300)+0.6(370)=342 Large Facility 180 900 0.4(180)+0.6(900)=612 The large facility has higher EV which is 612. A schematic model of alternative available to the decision maker along wit their possible consequences Composed of a number of nodes that have branches emanating from it. Two types of nodes a square a circle represents a decision point stands for a chance event. The branches of the tree having square nodes represent alternatives and branches having circular nodes represent chance events. The manager of the company has to decide whether to prepare a bid or not. It costs P5,000 to prepare the bid. If the bid is submitted, the probability that the contract will be awarded is 60%. If the company is awarded the contract, it may earn an income of P60,000 if it succeeds or pay a fine of P15,000 if it fails. The probability of success is estimated to be 70%. Should the owner prepare the bid? P = 0.70 P = 0.60 contract awarded prepare not awarded ▪ not prepare P = 0.40 success 60,000-5000 failure -15,000-5,000 P = 0.30 -5,000 Compute the EV backward from position. EV = 0.70(55,000) + 0.30(-20,000) = 32,500 EV = 0.60(32,500) + 0.40(-5,000) = 17,500 55,000 P=0.60 contract awarded prepare success P=0.70 32,500 failure P17,500 P=0.30 not awarded P=0.40 not prepare -20,000 -5,000 0 Decision: The manager should therefore decide to prepare the bid. Once a manager knows which decision to make, the payoff increases and is now a certainty, not a probability. EVPI = expected payoff under certainty – expected payoff under risk. Compute for the expected payoff under certainty. The best payoff for the small capacity and large capacity are P370 and P900 respectively. Then combine by weighing each payoff by the probability of that state of nature and add the amounts. The expected payoff under certainty is 0.40(370)+0.60(900) = P688 The expected payoff under risk, as computed is P612. EVPI = P688-P612=P76 Mr. X is planning to open up a new branch of Engrande Lechon at a new location in Pasig or expand the existing branch located in Marikina City. Demand on the new location is expected to be 60% high and 40% low. Fixed cost will reach the amount P150,000. If the demand becomes high, he expects to have a revenue of P250,000, however if the demand becomes low, he could only expect a revenue of P200,000. Upon analyzing the situation in his existing branch, he believes that by introducing new recipes the sales will reach the amount P120,000 if the demand becomes high. However, if the demand becomes low, he could only expect P80,000 revenue. Projection on high demand in the existing branch tends to be 55% and upon computing the fixed cost it would reach the amount of P50,000. If you were the consultant of Mr. X, what would you advise? Ruth has to decide whether or not to open a small dress shop near the U-belt, a few blocks away from her. Her options are to open a small shop, a medium-sized shop, or no shop at all. The market for a dress shop can be good, average, or bad. The probabilities are 0.20 for a good market, 0.50 for an average, and 0.30 for a bad market. The no profit loss for the mediumsized or small shops for the various market conditions are given in the table. Building no shops at all yields no loss and no gain. What do you recommend? Alternatives Good Market (PhP) Ave. Market (PhP) Bad Market (PhP) Small Shop 85,000 20,000 -30,000 Medium-sized shop 120,000 45,000 -70,000 No shop 0 0 0 Probabilities 0.20 0.50 0.30