Probability management and the flaw-of-averages

Abstract

Uncertainty is ever-present and is an integral part of life. Recognising the existence of uncertainty and its possible effects on decision-making may be important for the profitability, financial success, or even survival of an organisation. A relatively new discipline, - known as probability management, - has recently emerged as part of operations research/management science. This paper will attempt to provide a brief introduction to the concepts of probability management and the motivation behind its development. One of the major driving forces resulting in this development is the phenomenon known as the ‘flaw-of-averages’. The flaw-of-averages has important consequences in industrial engineering, financial, business, and economic models. The classic newsvendor problem will be used as an illustrative example. However, the main purpose of this paper is to discuss the most important characteristics of the flaw-of-averages. It will investigate and attempt to quantify some of the generic factors that may have an influence on the existence and severity of the flaw-of-averages and its expected consequences. Various models will be developed, and experiments will be conducted using Microsoft Excel as a modelling tool and an experimental approach based on Monte Carlo simulation modelling.

Onsekerheid is altyd teenwoordig, en is ʼn integrale deel van die lewe. Die herkenning van die bestaan van onsekerheid en die moontlike gevolge vir besluitneming mag belangrik wees vir die winsgewendheid, finansiële sukses en selfs die voortbestaan van ʼn organisasie. ʼn Relatief nuwe konsep het onlangs verskyn as deel van operasionele navorsing / bestuurswetenskap bekend as waarskynlik-heidsbestuur. Hierdie artikel sal poog om ʼn kort inleiding te verskaf tot die konsep van waarskynlikheidsbestuur en die motivering agter die ontwikkeling daarvan. Een van die belangrike dryfvere agter hierdie ontwikkeling is die fenomeen bekend as die ‘gemors-van-gemiddeldes’. Die gemors-van-gemiddeldes het belangrike gevolge in bedryfsingenieurswese, finansiële, ekonomiese en besigheids-modelle. Die klassieke koerantverkoperprobleem sal as ʼn illustratiewe voorbeeld gebruik word. Die hoofdoelwit van hierdie publikasie is egter om die mees belangrike eienskappe van die gemors-van-gemiddeldes te bespreek. Dit sal sommige van die belangrikste generiese faktore wat ʼn invloed mag hê op die bestaan en die erns van die gemors-van-gemiddeldes ondersoek, en poog om die verwagte gevolge daarvan te kwantifiseer. Verskeie modelle sal ontwikkel word, en eksperimente sal uitgevoer word deur gebruik te maak van Microsoft Excel en ʼn eksperimentele benadering gebaseer op Monte Carlo simulasie modellering.