Lying By The Numbers: Potemkin Numbers And You

When Joe McCarthy told a stunned America that there were 57 card carrying members of the Communist Party in the United States State Department the allegations were treated as fact. After all, 57 is a very specific number. It must reflect an actual counting process, and if something can be counted, it must exist, right?

Wrong. Joseph McCarthy's 57 (or is that 102?) Communists is a classic example of what Charles Seife, in his must read book "Proofiness," calls a "Potemkin Number." Like the fake buildings supposedly put up by Prince Potemkin to fool Catherine the Great into believing that the serfs were living prosperous lives, a Potemkin number is a hollow façade designed to deceive.

It is natural to assume that numbers have meaning. Without context, however, numbers can be meaningless. In "The Hitch-hikers Guide to the Galaxy," a supercomputer reveals that the answer to the ultimate question is "42." The problem is, no-one is certain what the actual question is. Context.

Anyone can invent a number and use it to create the illusion of reality. You might doubt that there are suspicious people hanging around on the street, but if I tell you that I can see three suspicious looking people hanging out on the corner, you might take my claim a little more seriously. The specificity gives weight to the claim.

We want facts in order to understand things. "Shots may have been fired" sounds wishy-washy. "Shots have been fired," is more affirmative, but "six shots were fired" is best of all. "Six shots," six is a number, that makes it a fact, right?

Wrong. Without evidence, "six shots were fired" is nothing more than a claim, not a fact. Before you believe that six shots were fired, you need to know who is making the claim and how they obtained the information. Were they present when the shots were fired, or are they merely repeating something someone else said? If you trace the claim back to its origin, you also want to determine how accurately the shots were counted. Did someone count each shot as it was fired, or did they make a guess based on their recollection? Did they count muzzle flashes or sounds? Each method has its own strengths or weaknesses.

This may seem like hair splitting, but it is actually a very important part of critical thinking. Numbers without a context can be used to lie. There were not 57 card carrying Communists in the State Department, yet McCarthy was able to convince people that there was a massive conspiracy afoot.

Whenever a source provides a number, ask yourself how they arrived at that number. If you cannot see any evidence that shows how that number was arrived at, it may be a Potemkin Number,a number plucked out of thin air to persuade you that a lie is based on solid fact.

Here is a handy guide to spotting Potemkin Numbers:

1. Is the number plausible? Does it seem too high or too low?

2. Is the number too precise? In the early stages of a disaster, one can speculate that "hundreds are at risk." 349 at risk would be too specific.

3. Is the number rounded appropriately for the context? The Apollo program cost around 110 billion dollars, as opposed to $110,204,683,468.79

4. Is the method used to determine the number transparent? If it is a statistic, is the sample size significant? Is the methodology explained?

Be on the lookout for Potemkin Numbers, the propagandist's best friend.

a reply to: DJW001

Proofiness? Sounds like he owes the term to Stephen Colbert and his "truthiness".

I'll definitely have to pick up this book. The topic sounds very interesting. I'm very math focused and teach math for a living and the biggest lesson I learned when I taught myself statistics is that it's very very easy to play with numbers.

We all seem to think that just because a " trusted " source gives us information it must be true. It's a shame we get the subjective summaries and pretty much zero access to the raw data these days.

On the other hand, refusing to use numbers means that you are not open to new data.

So, If I try to tell you that 2/3 of US firearms-inflicted deaths are in fact suicides... what can you do with that information?

Does it seem too high or too low? Does it seem convenient for my arguments, whatever they may be? If so, then you must immediately distrust everything I say.

How convenient for you.

23. 'Nuff said.
Investigate at your own risk.

a reply to: tovenar

I think the point of the books author and the OP is that you should ask where the number came from and how it was obtained. Why should we trust what you say? Why should you trust what I say?

a reply to: tovenar

On the other hand, refusing to use numbers means that you are not open to new data.

Not necessarily. Not everything is amenable to mathematical treatment. On the other hand, an inability to provide figures where figures are needed makes for a weak case.

So, If I try to tell you that 2/3 of US firearms-inflicted deaths are in fact suicides... what can you do with that information?

Unless you can explain how that statistic was determined it is meaningless; in fact, it looks likely to be a Potemkin number.

Does it seem too high or too low? Does it seem convenient for my arguments, whatever they may be? If so, then you must immediately distrust everything I say.

That would depend upon whether or not you can provide a source for the statistic. If you are unable or unwilling, it would certainly cast doubt on your claim.

How convenient for you.

I'm not sure what you're getting at here. Try applying the techniques I have suggested to figures and statistics that lead to conclusions you don't like. Better still, apply them to figures and statistics that support your own beliefs and be prepared for your mind to expand!

a reply to: tavi45

Proofiness? Sounds like he owes the term to Stephen Colbert and his "truthiness".

He acknowledges Colbert in the preface.

a reply to: DJW001

Very nice. Colbert is a wildly intelligent man.

a reply to: tavi45
Nah. That's just smartiness.

It's his writers that are smart.