A recent article by Andrew Coyne (The Vancouver Sun – System makes losers out of winners) comments on how the “first past the post” process for elections creates uncertainty in making election predictions by pollsters and other pundits (most notably in the recent U.K. elections). In Canada and more recently in my neck of the woods, British Columbia, we had a similar situation where the public polls and those apparently used by one of the parties suggested that the ruling Liberals were in for an election defeat. The result was dramatically different with the ruling party getting a large electoral victory.
I have thought about and studied how uncertainty and risk play out in our lives and posted many of my musing in this Blog. Andrew’s column brings to mind another way that we through our actions create uncertainty where there in principle it needn’t be so. How we look at, for and measure things create opportunities for uncertainty in our lives.
How the pollsters measure voter preferences and predictable voting patterns can and has masked what is really going on. In one sense in the U.K. election they were reasonably on as regards the overall voting patterns for each party (within 3% which is within their measuring margin of error). What they missed was how the distribution and location of this voting pattern fallout within each riding. Local factors were powerful ‘;skewers”, for example the Scottish National Party (SNP) got 4.7% of the popular vote but 56 seats whereas the UK Independence Party garnered 12.6% of the popular vote and got one (1) seat.
I have explored how we look at our world with our various world views (a.k.a. loose theories or hypotheses) enables us to see with incredible precision nuances and subtleties while at the same time being fundamentally deficient (even wrong and misleading) in other ways. When we stereotype people we run this “risk” all the time (and virtually every time).
When we do strategy like work, we often put in place means of measuring future meaningful (positive and negative) environmental events and patterns that signal to us the need to continue, modify, or even abandon our plans.
We often use statistical or stochastic type measures. These are powerful, BUT, potentially dangerous. An example is when we rely on an “average” and how it moves over time as the signal for whether we are making progress. Yet this form of measurement is very prone to the pollsters problem. An average is a calculation that can be derived whenever we have two or more items that we are counting. An average gives us a single value which makes comparisons to other average calculations very easy. But what if the actual “distribution is “bi” or even “tri” modal? When we have these kind of distributions ( and we are not aware/or don’t pay attention to them) our calculated average may very well mask what is going on. Remember the UK election: the SNP vote because it was geographically very localized, hence concentrated acted as if it was a unique modal area within the overall voting pattern.
Other times we know we cannot measure what is important directly so we rely on “proxies”. Think of how we tory and measure more qualitative attributes such as happiness, satisfaction, pain, etc. Although we may be limited to using proxies, we should always remember that these are indirect and approximate in nature. They can always mask or misdirect us from something that is important.
So how does all this relate to uncertainty and risk?
Fundamentally, risk is about dealing with variances. Is it too hot or cold? Is it too fast or slow? All these situations are about staying within some medium, median, or average place that is safe and alright. So if how we measure has underlying exposure to misdirection or misinformation then even though we think we are dealing with risk, we are not!
We are dealing with uncertainty! Uncertainty is about surprises in our lives. The difficulty for us when dealing with surprises, we basically not prepared for them. This is what it means to be surprised – we did not expect it and we are not anticipating it. Expectation means we know it happen and we are at least somewhat on guard for it (this is what fire insurance does for our home) and anticipation means we are monitoring the situation for its occurrence. When we have these two conditions in place we are dealing with risk not uncertainty.
So how we choose to monitor and measure things can mask what it is that we really want to know, or need to know (not the same thing sometimes)
Yet we have to measure, monitor or in some systematic way observe our environment(s). And, we are prudent in knowing how our choices in this area can lead us astray.
A smart person knows they can be wrong.
A clever person knows how, if and when they are wrong.
A sensible person will respect the above.