Experience taught me how invaluable it is to reflect on and write down my decision-making criteria whenever I made a decision, so I got in the habit of doing that.

Wouldn’t it be worth working out what the actual criteria are?

We had to be able to make a decision like that, without reprocessing all of our learning. So we take a small slice of information and a profile of the customer, set up a promotion, prime up the customer's history, and wait for the events. It primes up just the right amount of information in real time... [A system like that,] might not be as precise, but it would be statistically good enough. Maybe it would be wrong five percent of the time, but that's okay. It would be more like making judgment calls.

Here’s a simple, systematic process you can use to apply selective criteria to opportunities that come your way. First, write down the opportunity. Second, write down a list of three “minimum criteria” the options would need to “pass” in order to be considered. Third, write down a list of three ideal or “extreme criteria” the options would need to “pass” in order to be considered. By definition, if the opportunity doesn’t pass the first set of criteria, the answer is obviously no. But if it also doesn’t pass two of your three extreme criteria, the answer is still no.

Algorithm’ is arguably the single most important concept in our world. If we want to understand our life and our future, we should make every effort to understand what an algorithm is, and how algorithms are connected with emotions. An algorithm is a methodical set of steps that can be used to make calculations, resolve problems and reach decisions. An algorithm isn’t a particular calculation, but the method followed when making the calculation. For example, if you want to calculate the average between two numbers, you can use a simple algorithm. The algorithm says: ‘First step: add the two numbers together. Second step: divide the sum by two.’ When you enter the numbers 4 and 8, you get 6. When you enter 117 and 231, you get 174.

The first step in making some kind of sense out of this mass of raw data was to set up standards for measurement and analysis. After several attempts, it became apparent that only a few of the typical yardsticks could be applied. Therefore, assumptions or premises were made to permit identification in the sorting process, and the conclusions brought forth are only as valid as the premises on which they are based.

The function of knowledge in the decision-making process is to determine which consequences follow upon which of the alternative strategies.

Decisions should be based on facts, objectively considered — what I call the fact-founded, thought-through approach to decision making.

Strictly speaking, decision theory really is concerned only with the fourth part of the division given above, that is, the determination of the computational methods for optimization. Given the determination of the other three factors―the objective function, the range of policy alternatives, and the model―the ideal picture is that someone, presumable the firm that hires the operations researcher, hands him, on a silver platter, an objective function. By talking to the engineers, or by looking into a few scientific laws, he determines the policy alternatives available and also the model.

The function of knowledge in the decision-making process is to determine which consequences follow upon which of the alternative strategies. It is the task of knowledge to select from the whole class of possible consequences a more limited subclass, or even (ideally) a single set of consequences correlated with each strategy.

You need a strategy, and a trade or investment decision can be evaluated only in the context of that strategy.

A great deal of study has been directed to denning 'best decisions,' particularly since the pioneering work of mathematical statisticians (such as Wald), of mathematicians (such as von Neumann), of economists (such as Arrow)... The main effect of this development on the practice of OR has been the growing realization that there are decision objectives other than maximizing expected return and minimizing maximum loss. That is, in many practical situations there are criteria of optimality that are more appropriate than these two mentioned.

Now the salient characteristic of the decision tools employed in management science is that they have to be capable of actually making or recommending decisions, taking as their inputs the kinds of empirical data that are available in the real world, and performing only such computations as can reasonably be performed by existing desk calculators or, a little later electronic computers. For these domains, idealized models of optimizing entrepreneurs, equipped with complete certainty about the world - or, a worst, having full probability distributions for uncertain events - are of little use. Models have to be fashioned with an eye to practical computability, no matter how severe the approximations and simplifications that are thereby imposed on them... The first is to retain optimization, but to simplify sufficiently so that the optimum (in the simplified world!) is computable. The second is to construct satisficing models that provide good enough decisions with reasonable costs of computation. By giving up optimization, a richer set of properties of the real world can be retained in the models... Neither approach, in general, dominates the other, and both have continued to co-exist in the world of management science.

I said, “You’ve got to be able to look at the deal and know what it hinges on to know whether it works or not. If you realize that the key component works, then you use the numbers to test it. You don’t do the numbers to find out eight hours later whether it was worth starting.” I’m sure his IQ was higher than mine. But that isn’t how we operate. You have to be able to effectively assess the initial picture and see where the greatest risk is most likely to be, or you’ll spend your life doing numbers just to find out if a deal will work. And all that time lost is time you could have been looking at other opportunities.