Tuesday, May 2, 2017

Statistics: The Mathematical Theory of Ignorance

Be able to analyze statistics, which can be used to support or undercut almost any argument. - Marilyn vos Savant

Statistics are no substitute for judgment. - Henry Clay

The #1206 “fiction” series continues …


On a bright, sunlit Sunday in a medium-sized city typical of western civilization, children ran excitedly around, happily exploring police cars, ambulances, MedEvac helicopters and the like.  It was the annual emergency preparedness presentation offered by the city to its citizens and as usual, a fairly large contingent of families had shown up.

“It was”, as one parent mused to another as they meandered by the exhibits, “a great way to kill a Sunday afternoon.”

At one of the exhibits, a group of families listened raptly as a government official explained the government’s latest advances in emergency planning.

“Two key things to remember”, he explained to the group, “Always make sure you have three days of food, medicine, water, toiletries and personal items on hand.”

“And”, he said as he held up his mobile phone, “Always make sure you keep this charged up.  We have an app that you can download that will allow you to receive alerts from us.  You can also use your phone to communicate with emergency officials and family members.”

A tall, thin, pale man wearing dark sunglasses was observing from the sidelines.  When the presenter finished speaking, the thin man approached him.

“Everything you say may be true”, he said to the presenter, “But we all know that in times of emergency, cell phone systems often get overloaded as they are designed for enhanced coverage but not for enhanced capacity. You saw this happen for 9/11, Hurricane Katrina and other situations.  So it is likely that the people here won’t be able to make a call once an emergency event takes place because the networks will become saturated.  Some SMS messages may get through but data and voice communication may become spotty at best so your app won’t work when it is needed most.  As for duration, there have been many emergencies that were not resolved in three days.”

“Well”, replied the presenter, “That may be true but the likelihood of such an event is statistically unlikely.”

“Statistically unlikely”, the thin man said as if he were were digesting the words, “Interesting choice of words.  I will come back to that in a moment.  But back to the mobile systems, the average cell tower only has battery power for 8-16 hours so if a major power disruption of extended duration occurs, your mobile phone won’t be able to connect to anything anyway as the cell towers begin to die.  So people may be prepared for three days but their ability to communicate will cease long before this.”

The presenter frowned at the stranger, clearly agitated by the thin man with his observations.

The thin man pressed his point more firmly.

“Back to your observation regarding statistically unlikely”, he said emotionlessly, “Likelihood is only part of the equation for emergency preparedness when one must also consider the potential scale of such an event.”

“What does that mean?”, asked the presenter.

The thin man frowned at the presenter as if feeling a level of frustration with the lack of knowledge being shown by “the expert”.

“Well”, the thin man pointed out, “If I have an event that is statistically likely or unlikely, knowing the scale of the impact AND to a lesser extent, the cost of preparing for or preventing the event is what determines how much effort goes into mitigation of the event.  If I have an event that is extremely likely but has no impact of merit, then I don’t really care about preparing for it.  If the same event is also expensive to prepare for, then I really don’t want to prepare for it.”

He paused for a minute and glanced around him.  The crowd around him were staring at him in silence.

“Conversely”, continued the thin man, “If I have an event that is statistically unlikely, most emergency preparedness people use this fact as the reason to not prepare for it.  If this statistically unlikely event is also very expensive to prepare for, then emergency preparedness planners claim to have a reason to ignore it completely even if the potential event they are ignoring has catastrophic potential.”

“As they should”, interrupted the planner.

The thin man raised his hand.

“Patience”, he said, “Please allow me to continue.”

“No”, said the planner, “Statistical likelihood and cost are the key measures for emergency preparedness planning.”

The thin man began to respond but was again interrupted by the presenter.

“Give me one example where a major event was unlikely but happened with great significance or impact”, he demanded.

The thin man frowned slightly.

“The event you know as 9/11, the nuclear reactor accident in Japan, multiple suicides by pilots flying commercial aircraft, Hurricane Katrina,some of which I mentioned earlier”, began the thin man but he was interrupted again.

“But we couldn’t have anticipated those”, the presenter said, now clearly exasperated by the thin man.

“Of course you could have and many people in your industry actually evaluated the possibility of them”, replied the thin man, “But they chose the likelihood of it happening as the primary rationale for whether a response plan was needed.  They factored in cost plus the political unpopularity of acknowledging that such a thing could happen and then buried the problem as unnecessary to think about, talk about or prepare for.  To insult people even further, they claimed to be surprised by every major event that happened even though they had evaluated such events and discarded them based on the criteria I just mentioned.”

The presenter turned his attention back to the crowd of people around him.

“Thank you, folks”, he said to the now silent crowd, “I have a brochure here outlining what I shared with you today.”

Members of the crowd took a copy of the brochure and dispersed, some of them looking at the thin man uncertainly.

“What was that all about?”, the presenter demanded as he turned towards the thin man.

“Did I say anything that wasn’t true?”, asked the thin man.

“No”, replied the presenter, “But we don’t like to tell people these things.”

“So you are lying to them”, replied the thin man, “And in doing so, condemning them to some real problems when a large-scale disaster occurs.”

“Not entirely”, replied the presenter, “But putting the people in a panic doesn’t help either.”

“Informing them of reality always helps people”, the thin man replied softly but forcefully, “You need to consider them as your allies and not your enemies, now and in times of emergency.”

“Some people can’t handle truths like this”, replied the presenter.

“Possibly”, the thin man said, removing his sunglasses and looking intently into the eyes of the presenter, “But given that uninformed people become a potential liability when and not if a significant event occurs, I would rather inform the masses anyway instead of choosing to not tell anyone just because a small percentage of people aren’t mentally strong enough to deal with reality.”

The thin man’s eyes were dark and glittered in the sunlight.

The presenter suddenly felt uncomfortable as the thin man stared at him with a strange intensity.

“I think you should leave before I ask security to ask you to leave”, the presenter said.

“No need to be rude”, the thin man said as he put his sunglasses back on, “Someday you will be forced to admit the truth.  I hope that day doesn’t have other complexities that are difficult to deal with for you and your family.”

He turned and vanished into the crowd that was milling around.

The presenter stood in silence and watched him walk away.

He looked down at his cell phone, looked up at the cell tower he could see in the distance, thought about his own family and then pressed a speed dial button on the phone.

“Hi, honey”, he said when his wife answered, “We need to talk about a few things.”

To be continued.


© 2017 – Harry Tucker – All Rights Reserved

Addendum - When the Data is Embarrassing (June 9, 2017)

In a report released this week regarding the fire at Fort McMurray last year, a number of embarrassing things were revealed leading up to and during the catastrophe.  Had you asked people if they were ready to fight a fire, they would have said "yes" with great certainty.  Despite this assertion, whoever was in charge of coordinating readiness failed Fort McMurray.  The people, on the other hand, rose to the occasion as humans often do.  What are the lessons here?  Details regarding the report can be found here - Fort McMurray wildfire reaction marred by communication gaps, says report (Calgary Herald).

Blog Post Background / Supporting Data

The title of this post is a quote from Morris Kline.

Interestingly enough, this conversation actually took place a couple of years ago at an actual “emergency preparedness and planning expo”.

The facts cited by the “thin man” are true.  The possibility of disaster from a hurricane in New Orleans, the likelihood of an event like 9/11, the nuclear disaster in Japan, death of passengers by pilot suicide, and other items have all been evaluated and discarded for the reasons given – statistical unlikelihood, the cost to address and the political damage that could arise should such things be disclosed before they happen.

The facts regarding the cell phone system are also true.

How one prepares for an emergency, whether it be man-made, from Mother Nature, from outer space or from any other source, depends on how much knowledge one has.  And by the way, emergencies are not limited to large scale disasters in our society.  Emergencies can also come in the form of problems with relationships, in business, in personal health, in financial health and the like.

Probability of occurrence is only a small part of preparation.

Time, energy and money to prepare for or prevent the event are important considerations also.

The government’s public promotion of preparedness and how it considers risks ends there.

However, the scale and impact of the event is equally important when evaluating what one should prepare for.  Many times when we are overrun by something, it is because we knew there was a possibility of it occurring but we ignored it for the reasons discussed in this post.  It is the thing that we choose to ignore rather than a surprise that often presents the greatest difficulties for us.

Aristotle often referred to a life well-lived as one that finds the golden mean of excess.  For example, courage is the golden mean between rashness and cowardice.

The golden mean for emergency preparedness falls between doing nothing and exhibiting extreme paranoia.  It can only be found when one informs one’s self instead of relying on data presented by someone with an ulterior motive.

Are you informed for the reason of acquiring knowledge or do you only allow others to inform you based on their needs, intentions and motives?

The answer to this question will determine how prepared you are when, not if, a major event happens in our near or distant future.

Do you care?

How important is your family’s health and safety to you?

What do you need to do, if anything?

Series Origin

This series, a departure from my usual musings, is inspired as a result of conversations with former senior advisors to multiple Presidents of the United States, senior officers in the US Military and other interesting folks as well as my own professional background as a Wall St. / Fortune 25 strategy advisor and large-scale technology architect.

While this musing is just “fiction” (note the quotes) and a departure from my musings on technology, strategy, politics and society, as a strategy guy, I do everything for a reason and with a measurable outcome in mind. :-)

This “fictional” musing is a continuation of the #1206 series noted here.

Related musings:

2 comments:

  1. Great post, Harry. Also, without having a lot of domain knowledge on the subject, it seems to me that the math involving whether or not an event is "statistically likely" would be very hard to derive for things like this. There are A LOT of variables at play and many of them are human-related. Also, mitigation strategies aren't black and white either as they can span many aspects of the risk's potential problem domain. They would vary in cost and in effectiveness. This is a complex topic. Reasoning alone shows that presenters, like the one above, are over-simplifying. Thanks for showing facts to back that reasoning up.

    ReplyDelete
    Replies
    1. Thanks, Nathan. I have spent many a day in REAL emergency planning that doesn't make it to the public domain where I have left the meeting in a daze, not sure if I should weep for our future or throw up with the stress.

      You are right regarding the number of variables. However, they are all solvable depending on how much time, energy and money we are willing to invest.

      Interestingly enough, many of them would provide an interesting job stimulus if they were approached the right way.

      That being said, I don't expect a government to think the way a businessman does. :-)

      Even though it is complex, history has taught us that over simplifying / ignoring it is not the right approach.

      Create a great day, Nathan!

      Harry

      Delete