[SOUND] Today we're going to be talking about the case study activity called the hiring decision. And in particular we're going to be using decision analysis to try and decide what the right decision is, in this case. So what's the core choice here? Really there are two options. The first is to offer the job to Bob. The second is to offer the job to Susan. What makes this decision a dilemma? Well it's a dilemma, because it's a choice between a sure thing and a uncertain outcome. Bob is the sure thing. We know if we offer the job to Bob, that Bob's going to take the job. There's no risk there at all. Susan, however, represents a risk. There is some uncertainty about whether Susan will take the job if we offer it to her. And it turns out there's a little additional uncertainty because if we offer the job to Susan and she says no. Bob may not take the job if he thinks he's been offered a job after Susan was offered the job. So we have some action options. We have some uncertainty. What are the outcomes that attached to these action options? Well first of all on the sure thing side, we know that if we offer the job to Bob, he will take the job. But Bob has only about 80% of the potential of Susan. So this sure thing is only about 80% of what we're really looking for. On the uncertain side of the equation, we know if we offer Susan the job, there is some probability she'll take the job, which would be a home run, 100% of the value we're looking for. And some probability she'll say no, at which point we would offer the job to Bob. But once again, Bob is only about 80% of the potential of Susan. What makes this a dilemma, of course, is that the best outcome is Susan taking the job, 100% of what we're looking for. But the worst outcome is neither taking the job, and in-between, of course, is Bob. Bob is not the worst outcome, but Bob is not the best outcome. So both the best and the worst outcomes are uncertain. The certain outcome, Bob taking the job, Is between the best and the worst or uncertain outcomes. So let's address the uncertainty here. We know that the probability of Susan taking the job, if we offer it to her, is 25%. And we also know that there's a 33% probability that Bob will say no if Susan also says no. So we can work through the equation to figure out what our decision should be. Our sure thing is that Bob will take the job, but only at 80% of what we're looking for. The uncertain outcome, is that Susan has a 25% chance of taking the job at a 100% of what we're looking for. 75% chance that she'll say no, in which case there is the 67% chance that Bob will take the job at 80% of the value we're looking for. When we worked through these equations, what we find is that the certain alternative is 80% of what we're looking for, and the uncertain alternative is 65% of what we're looking for. That's decision analysis using the action options, using the outcomes that attach to those action outcomes, and calculating through the uncertainties to decide which decision has the best expected value. The concerns of course in this case are that there is a lot of estimating going on here. We're estimating that hiring Bob is only 80% of the value of hiring Susan. That's pretty much a guess. We're also estimating that the probability of Susan taking the job is only about 25%. That's also a guess. And we're estimating that the probability of Bob saying no if Susan turns the job down is 33%. That's also a guess. So there's a lot of uncertainty here that we have to estimate. We've also limited ourselves to only two action options, that is hiring Bob and hiring Susan. Now of course, if Susan says no, we could negotiate with her. We could try to do things to make the job more attractive, so it's not simply a yes or no proposition here. And finally we have to consider whether zero captures the value of both Susan and Bob saying no. So there's a lot going on here that really isn't captured by the information we have. In that sense, decision analysis is great when we have great information, but when we're doing a lot of estimating, when we're doing a lot of guessing. Decision analysis is really just a framework for thinking about how we might make this decision. So decision analysis provides a very useful framework for thinking about how tough decisions can be made. Decision dilemmas, a choice between a risk and a sure thing. Decision analysis presents us with the framework for objectively integrating action options, outcomes and probabilities. The usefulness of decision analysis is always limited by the availability of critical information. Often we have to do a lot of estimating and often we have to constrain the universe of choices we're actually considering in order to be able to work through decision analysis and reach the decision. In that sense, decision analysis provides us a nice framework for thinking about decisions. But at the end of the day, it's simply a framework for helping us think about how to make better decisions.