Opinion / 17 September 2019, 3:00pm / Pali Lehohla
JOHANNESBURG – Systems are planned and implemented by people. They feel and think. They even have additional assistive tools to secure their confidence in what they do.
But doubt is a problem. Doubt is a field of statistics that informs probability. So everyone becomes akin to being a statistician as we compute doubt and factor it into decisions. It is a passive, but powerful attribute of life.
Having now left the office as the statistician-general of South Africa for almost two years, I continue to discover new things. This morning I had an appointment for my US visa, something that would have occurred smoothly two years ago without me realising it.
Previously, these were taken care by my back office at Statistics SA. Understanding the trappings of power and the frills that come with incumbency can mute your sensitivities of self-reliance.
The appointment was scheduled for 9am. I calculated that if I left home at 7am I would have enough lead time to get to Sandton. I also knew that I could rely on Google Maps to assist me. So I took off.
As I joined the highway, I still had a full hour and a few minutes to get to Sandton.
The traffic slowed to a snail's pace and Google Maps said I would arrive at my destination at 8.59am, but then said 9.03am. I was freaked out as I had waited a long time for this appointment.
I wasn't sure if I could trust the evidence of the time or arrival and my insecurities at missing the appointment were overwhelming, because the next appointment I could have was October 23 and beyond my date of departure.
Now I had to choose whether I should get off the motorway and get on to the Gautrain at Midrand or Marlboro or continue driving.
Google won the day and I arrived at the Consulate at 9am.
What Google was doing was to remove my subjective probability based on fear of consequence and provide hard facts for me.
At the consulate another act of planning interested me as to how they dealt with people.
They moved people in batches of 10 and this rhythm was kept throughout several stages that each of the batches had to go through.
And like clockwork the queueing worked. For example, one station for fingerprinting had 10 people.
How much time you spend in traffic or in a queue is a researched subject by operations research.
In transport systems, the whole subject of transit systems is a feat in modelling. Queueing is expensive, because it is a resource-wasting process, yet it is a necessary condition to go through. Statisticians and operations research look at these all the time in resolving systems efficiency and effectiveness.
It is always a pleasure when your subjective probability is one with Google’s fac-based probability. But as humans we have strong feelings and we need to remain being human in the era of the Fourth Industrial Revolution.
Dr Pali Lehohla is the former statistician-general of South Africa and the former head of Statistics South Africa. Meet him on www.pie.org.za and @palilehohla