Uncertainty & Risk – Eliminating misconceptions

Uncertainty & Risk – Eliminating misconceptions

Speakers: Srinivas Atreya, Chief Data Scientist, Cigniti

  • Here is the Transcript

Srini: Hello, folks. This is Srini Atreya again. Apologies for the longest gap, just that some other things have gotten in the way. But here I am now promising to be regular in the future. So how are you all doing? Hopefully, you’re all at home minimizing unnecessary travel and staying safe. Hopefully, this will all pass away very soon, and we will all be back to work.

In the meantime, I’ve seen and read quite a few data science articles on Covid 19 and this has gotten me thinking about uncertainty and risk, and that’s what we’ll be talking about today. These are fundamental concepts to anyone working in applied statistics space, but there still seems to be quite a few misconceptions. Today’s talk will be a bit meandering and probably touch quite a few areas, but that is just to give a holistic picture. As usual, please feel free to leave any comments, suggestions for improvement, and thanks for listening in. Everything is vague to a degree you do not realize till you have tried to make it precise, said the great Burton Russell in his article The Philosophy of Logical Atomism.

As data scientists who happen to influence organizational and societal decision-making, it is imperative that we try and understand these concepts from different vantage points. Today we will talk about uncertainty and risk from an economic standpoint. In later discussions, we’ll try and explore this from a mathematical, statistical, engineering, and even psychological viewpoints.

Okay, so what are the odds that your new idea will succeed? If it does, what will the return to you be? One of the problems that we have in business and in life is that we often can’t know the answers to questions like this in advance. And this drives us nuts. At least it does to me. Consequently, a lot of people invest a great deal of effort into reducing uncertainty. There are two problems with this approach. The first is that we often do not understand uncertainty very well. And the second is that profitable opportunities only exist where outcomes are genuinely uncertain.

Frank Knight wrote about this in 1921 in his seminal work called Risk, Uncertainty, and Profit. He distinguished between two types of uncertainty. The first type is when we know the potential outcomes in advance, and we may even know the odds of these outcomes in advance. Knight calls this type of uncertainty risk. An example of risk is rolling a pair of dice. Before we roll, we know in advance what the odds are for each possible outcomes, provided that the dice are fair. Knowing these odds forms the basis for all the games of chance that we can play. Dice are relatively simple, cards are a bit more complicated, but we can know all of the odds with them in advance too. Genuine uncertainty is a bit different. This occurs when we don’t even know the possible outcomes in advance, let alone their probabilities. Genuine uncertainty occurs in complex systems where a lot of actors interact over time the economy, for example. This is also understood as ambiguity.

The important point that Night makes is this real opportunity for profit only exist in the face of genuine uncertainty. Which means that if we want to innovate successfully, we not only have to deal with uncertainty, we must seek it out. We’re not the most rational decision-makers in the first place and confusing uncertainty with risk makes things worse. Here’s how we act like everything is just a risk. Here is how Nike puts it business decisions, for example, deal with situations which are far too unique, generally speaking, for any sort of statistical tabulation to have any value for guidance. The conception of an objectively measurable probability or chance is simply inapplicable. The confusion arises from the fact that we do estimate the value of validity or dependability of our opinions and estimates. And such an estimate has the same form as a probability judgment. It’s still a ratio expressed by a proper fraction, but in fact it is meaningless and fatally misleading to speak of the probability in an objective sense that a judgment is correct.

When we act like everything is a risk, we greatly increase the chance of failure. However, the opposite can also be a problem. We act like everything is unknowable. Uncertainty often gets blamed for inaction. The next time you hear someone mention uncertainty, ask yourself this how much less do they actually know about the future today versus what they knew last week or year? How much less do they think they know? We can’t use not knowing as an excuse to not act because we never know. It seems like we either suppress uncertainty and act overconfidently or we overemphasize uncertainty and don’t act at all. Both are bad outcomes.

Managing risk is pretty straightforward you match up your investment to the odds of it paying off. Managing uncertainty is trickier. Fortunately, there are a few things that we can do. It’s important to cope up with uncertainty effectively because doing so allows us to go where the opportunities are. Here are some strategies aggregate when I get in my car to drive to work tomorrow, it’s impossible to know if I will have an accident or not. And yet I have car insurance. How can this be? Insurance works because when you add up enough individual cases of uncertainty, you might end up with probabilities.

When we are talking innovation, this means that Lioness Pauling was exactly correct when he said that the best way to have a great idea is to have a lot of ideas. Volume can reduce uncertainty. This is what venture capital firms do too, when they make a wide range and high number of investments. Understand that when we face uncertainty, some ideas will fail. But this is what drives progress.

Joseph Schumpeter, Austrian political economist process of creative destruction can only proceed by trial and error. We see that which is created through the lens of survival bias and ignore the hopeful monsters that economic evolution has spawned and left behind in metamorphical emulation of Darwin’s process of natural selection. No doubt every one of them was launched on the basis of an exercise in forecasting future revenues, costs, and an expected value to be compared with a rough estimate of the cost of capital. As Schumpeter well knew wastage is a measure of the inescapable uncertainty that attends the practice of doing capitalism. We need to only visualize the situation of a man, let’s say, who considers the probability possibility of setting up a new plant for the production of cheap aeroplanes which would pay only if all the people who drove motor cars could be induced to fly. The major elements in such an undertaking simply cannot be known. Neither error nor risk expresses adequately what we mean.

Waste measures uncertainty and uncertainty has a potential for profit. It’s a little tricky but getting your head around this is critical to actually understanding the very nature of uncertainty. It is very very important to understand the distinction between risk and uncertainty. The two situations require different responses and if we confuse the two, we will not use the right approach. Risk is objective uncertainty which can be managed, whereas subjective uncertainty or true uncertainty is also known as ambiguity. And this cannot be really managed in the traditional sense, especially in times like these where there’s so much information about COVID-19 but so little is actually known. It’s important to understand the nature of uncertainty and deal with it differently than risk. Not knowing this difference will only increase the risks we face.

Hopefully this has helped a little in understanding the difference between uncertainty and risk and ambiguity to just to confuse it a little more. I guess the next time we’ll actually talk about this in a more mathematical sense rather than an economic sense, we’ll try and formulate this in a little more mathematical sense and then we’ll probably look at it from some other vantage points too.

Once again, thanks very much for listening in and hope all of you have a very great day.