When I will talk Tajikistan Email List Pareto, or the Pareto principle in this article, it will not be to refer to the Pareto Law – a law of statistical distribution – but to the pseudo-correlation which derives from it and which proposes that 20% of causes to a phenomenon would be enough (too) often (for it to be a coincidence) to explain 80% of this phenomenon. Some often used examples: 20% of the population is responsible for 80% of the costs of [insert what you want] (health care, justice, insurance…) 20% of your social interactions / friends are 80% of your happiness (I promise I heard that before …)

I am therefore preparing to speculate, foment, conspire against this principle of 80/20 which is even found in “productivity tutorials” against a background of personal development and which, without going that far, is often a significant “theorem” in business. . We ourselves at ISLEAN, we often use the methods which derive from it (problem solving, ABC Method…) for our clients. This article is not the object of an in-depth study but of a critical questioning of which I encourage you to abuse the foundations. Imagine. You are faced with a problem and you are looking to solve it. For this you diagnose the causes of your problem and

Oakland airport (California)

, each situation being different, each of you arrives at different cause / problem breakdowns where: Therefore, from what thresholds can we consider that the Pareto principle was applied? Since the “discrete” value of 80/20 does not apply in our continuous world, what limits must be set at this proportion so that we can reasonably say to ourselves: “there, a form of organizing entity has meddled in my affairs and paretoized my problem ”? We often suggest that between 70/30 and 90/10, we are still in the Pareto principle (it’s arbitrary but I’m not the only one to say it; see more here , here or here ). If we accept these limits, how does this translate? Let’s schematize:


Another way to visualize this “scope of applicability” is to reconstruct the Lorentz curve for different values ​​of k the Pareto index Abscissas : increasing cumulative share of the causes; Ordinates : increasing cumulative share of effects In short, and with the limits currently chosen to speak of the “Pareto” effect, we see that the principle addresses in fact 40% of all the possible scenarios imaginable a priori for the cause / effect distribution. To speak of a true Pareto “effect”, particularly recurrent – or of a law of nature – it would therefore be necessary to observe its application in much more than 40% of the cases where one seeks to break down an effect. by causes.

Shenzhen Airport (China)

In this diagram I propose to observe only half of the field of effect / possible cause distribution cases, the missing half being symmetrical (if 70% of the causes explain 30% of the effects, it is because the remaining 30% explain 70 % of effect). From my personal experience, through all the analyzes that I have been led to carry out, I cannot affirm that I have found this effect (in the limited sense 70 / 30-90 / 10) in a proportion significantly greater than 40%… Worth what it’s worth, that is to say little – certainly not a study result – but what about you?

Whatever the case, the 80/20 principle does not come from a study result either, the burden of proof therefore falls on those who put it forward! Until then and given that the phenomenon in fact covers almost half of the field of possibilities, I cannot help thinking that we are facing a form of tautology … So far, I’ve been a bit naive. Today, in the Lean toolbox, the Pareto principle is not so much put forward as a “law of nature” but rather as a form of rule to have in mind in order to attack a problem by its causes. the most important. This is called the Pareto method or the ABC method.

Leave a Reply

Your email address will not be published. Required fields are marked *