Shrinks Or Quants: Who Will Improve Decision-Making

Abstract

Abstract Uncertainty is the underlying cause of the failure of many decisions to deliver their promised technical and economic metrics. Are these failures, despite the increased application of tools and processes for decision-making under uncertainty over the past decade, due to bad decisions or just bad outcomes? This paper examines quantitatively the validity of the assumptions underlying current decision tools and processes, and the extent to which their application accounts for the risks that arise from uncertainty. We also quantify the impacts of prevalent human biases and cognitive limitations on our ability to make decisions that are logically consistent with our objectives. We conclude that the above factors can have a large impact on the outcomes of decisions and therefore bad process, not just bad luck, is a major cause of disappointing results. More awareness is needed of the capabilities and limitations of common decision-making making techniques, and of the psychological and judgemental factors that influence their application. Introduction This paper examines the consequences for the outcomes of decisions of two schools of thought on decision-making: those who take a traditional normative approach (known as "quants") and those who focus on the psychological aspects (the "shrinks"). We will show that both have important roles to play in improving the outcomes of decisions, and thus the creation of value in our industry. Conversely, ignoring either school results in loss of value. Business Context Every major petroleum company has faced the disappointment of failing to return the predicted technical and economic metrics that formed the foundation of their investment decisions - implying a systematic over-estimation of returns and/or under-estimation of the risks of loss. This applies to major capital expenditures decisions such as to develop a new field, purchase a competitor, build a pipeline, etc. It also applies to operational decisions associated with the search for greater efficiency in exploration, appraisal, development and production activities, such as decisions to acquire seismic, core a well, perform a production test, drill an in-fill well, carry out technical studies, invest in R&D, etc. Whilst many forces influence these failures, we believe that the single underlying cause is uncertainty - uncertainty around current "states-of-nature" (e.g. OOIP) and uncertainty around predictions of future "states-of-the-world" (e.g. oil price, recovery efficiency or even the effective implementation of decisions). In addition to the evidence for uncertainty under-estimation from the outcomes of historical decisions, there is strong evidence that people tend to grossly under-estimate uncertainties, particularly:the number of uncertain factors,the magnitude of the uncertainties,the complexity of the relationships among uncertainties, andtheir consequences. Background Business under-performance in the oil & gas industry, the failure of many decisions to return expected results, and the advent of increased availability of decision-making tools and software, has led to considerable promotion of techniques for so-called "normative" decision-making, such as decision-tree analysis. (For now, normative decision-making is defined as a process for making optimal decisions consistent with ones objectives.) However, adoption of these techniques has been retarded for several reasons, includingresistance from older, entrenched methods (basically the status quo)their perceived inability to account for the risk tolerance of decision-makersfailure of the normative toolkit to yield expected results due to violatedassumptions about people's logical behaviorconditions of application We have previously addressed the requirement for a holistic approach to evaluating complex decisions. - an approach that incorporates all of the key elements that may influence the decision, their dependencies and an ability to assess uncertainties in an integrated fashion1. We have also addressed the need for an awareness of the many human biases that can effect the assessment of uncertainties, and of the cognitive limitations of the human mind in evaluating complex problems and in making decisions that will achieve their objectives2.