What Price Danger?

Evaluating Statistics in the light of human reason

I am always on the lookout for the presentation of statistics or other facts in a way that perverts the argument. And I heard a wonderful example in a recent Triggernomety podcast.

The scientist being interviewed, and I will leave it to the reader to decide if they think he merely made a mistake or was being deliberately disingenuous, presented the following example to support the idea of scientific consensus:

“If 95% of engineers told you that a given bridge would fall apart, and you would fall to your death… and the other 5% told you it would be fine, would  you cross the bridge?”

If any rational and non-biased person were to spend more than a few seconds examining that statement, they would see many problems with it. But in case I am wrong, I will lay it out, point by point:

1) Turn the problem around:

- “If 5% of engineers told you that a given bridge would fall apart, and you would fall to your death… and the other 95% told you it would be fine, would  you cross the bridge?”

Unless you are an idiot, or unless the bridge is the only way for you to save your life or win some battle, the rather obvious answer is ‘no’. Under ordinary circumstances, a one in twenty chance of dying is not something most of us do. It is something most of us avoid doing. If we didn’t, we would be dead.

Humans, like all animals, are wired to avoid deadly risks. A gambler, risking a few green pieces of paper and hoping to gain more than he risks, will certainly plunk down on something that has a 95% chance of winning. But even he won’t tend to bet on something when he has a 5% chance of dying!

2) Now make the problem neutral:

-If 5% of doctors said a given treatment would save your life, and otherwise you would die, but 95% of doctors said the treatment would kill you, and you would be fine without it… well, obviously, you wouldn’t take the treatment… but you wouldn’t be happy either.  You would keep doing rather frantic research.

3) Now make the problem unimportant:

-If 5% of flower arrangers said that a given arrangement was nicer than another, and the other 95% said the opposite… well, you wouldn’t care. You would go with whatever you liked and ignore the flower arrangers.

AND

4) What 5% and what 95%?

-Are these just random scientists you pull off the street, or do they have some agenda? Scientists are people, too, and have beliefs and prejudices and the like.

I think the speaker was trying to argue that we should follow the scientific ‘consensus’. If so, he failed miserably at presenting a logical case. If he wasn’t trying to present a logical case, if he was trying a bit of sophistry, then I’m afraid he succeeded.

There are far, far more important considerations in making a decision than whether ‘95% of dentists agree’. Indeed, that statistic is all but irrelevant to the discussion. And in our current age, post the massive lies of the covid hype, everyone should know this. But they probably don’t.