More Proofiness

In the last post, I discussed a new book, by Charles Seife, Proofiness. Since it must go back to the library tomorrow (I tried to renew, but someone put a hold on it — a reader?), I have to hurry to share another fascinating tidbit from it. (Yes, I could buy a copy, but I am cheap.)

There are many aspects to how numbers and polls can be used very effectively to sway people. When anti-abortion groups conduct a poll, guess what? Most people oppose abortions, even when the mother’s health is at risk or rape is suspected. Even the name they go by, “Pro-Life” instead of “Anti-Abortion,” is a well-thought-out example of proofiness. Unless you are one of them, in which case you probably feel the terminology is correct. That is the seduction of proofiness.

Now before any Pro-Lifers out there start hurling flames, please remember that I state here for the record that the “Pro-Choice” people use many of the same tactics. This space is too limited to do justice to all aspects of any controversy. Read the book.

But today I want to dwell a bit on what Seife calls “randumbness.” Part of the reason we have been so successful as a species is that we have evolved a wonderful knack for pulling patterns out of data. Even to this day, an observant human can often out-perform a computer in finding patterns in what appears to be random data (think chess). This is both a blessing and a curse. We can quickly decide if a lion in the bush is crouching for attack, but we can also see patterns when there is no reason for them. That is “randumbness.” We see it all the time, but most often in gambling and sports. So-and-so playing for the Somethings has developed a “hot hand.” Is that so? Could it just be a statistical fluctuation? These dice are hot, bet it all.

Humans (me and you included) have a real difficult time with random data. We do our best to impose order on it. We see animals and gods in the night sky when the sprinkling of stars in our vicinity is essentially random. Astronomers might question that since upon deeper analysis than what can be done with naked eye observation and paper and pencil analysis, some types of order can be seen. That does not change the statement that constellations as we know them are order artificially imposed on random orientations to help humans organize the sky. Historically this was extremely important for navigation and establishing a calendar. Unfortunately it also enabled a higher order of randumbness in astrology in which the random patterns of life are the subject.

What is the difference between weather and climate? Do major blizzards crossing the country constitute a refutation of global warning? How much of weather and climate is a pattern and how much is truly random? Why does prediction work? Why does it not work well?

Have you even thought about what is means to say 20% chance of rain tomorrow as opposed to saying there is a 1/6 chance a die throw will come up 3? And how do those examples differ from the far deeper randomness of radioactive decay? That last question bothered a lot a very bright people for a long time.