Lying for Jesus, Mathematically

I've previously commented about Marvin Bittinger's book, The Faith Equation: One Mathematician's Journey in Christianity. I called it a "combination of ignorance and intellectual dishonesty". Now that I've had a chance to read it more carefully, I find I was too kind. It is pure and utter dreck. Actually, "dreck" is far too kind. I find it hard to convey the self-satisfied stupidity that is found on nearly every page.

Instead of giving a detailed critique, in the spirit of the Carnival of Mathematics, I'll focus on some of the questionable mathematics that Bittinger uses.

Christian apologists have long been fascinated by the power of mathematics. My colleague Wesley Elsberry has taken apart an argument from 1925 here, where the author claims that the current rate of growth in human populations implies a young earth. The writer of that bogus argument claimed that "Figures will not lie, and mathematics will not lie even at the demand of liars." Unfortunately, the reverse is true: it's easy to lie for Jesus, mathematically. And probability theory is one of the easiest tools to abuse.

In The Faith Equation, tiny probabilities are assigned, often with little or no justification, and probabilities are multiplied together with no evidence of independence. These tactics are particularly evident in Chapter 4, "The Probability of Prophecy". In this chapter, Bittinger concludes that prophecies in the Bible constitute an event of probability 10^-76, which is a miracle that proves the accuracy of the Bible and the existence of God.

As I've already pointed out, Bittinger ignores significant criticism about his claimed prophecies. Tim Callahan's Bible Prophecy: Failure or Fulfillment?, Farrell Till's Prophecies: Imaginary and Unfilfilled and Jim Lippard's Fabulous Prophecies of the Messiah all take issue with many of the prophecies claimed by Christians. I see no sign that Bittinger has read these critiques; he certainly hasn't cited them in his reference list.

I'm not going to get into the accuracy of individual prophecies here; instead, I want to comment on one tool that Bittinger uses to justify his small probabilities. On page 93, we read:

"There is a concept from probability that we use often in these arguments. Suppose an assertion, such as God promising never again to flood the earth after the time of Noah's Ark, occurred t years ago, and to date the prophecy either has not been fulfilled or was just fulfilled. Statisticians would then estimate the probability of the event to be approximately one over twice the number of years: 1/(2t). We refer to this as the time principle and use it extensively."


There are two problems here: first, the "time principle" is completely nonsensical and second, it is not used by "statisticians" as Bittinger claims.

The "time principle" is nonsensical for several reasons. First, it is based on years, an entirely arbitrary way to measure time. We can get any probability we like from the formula 1/(2t) simply by changing the unit of measurement. If we measure time in centuries instead of years, the probability increases by a factor of 100. If we measure time in seconds, the probability decreases by a factor of about 31,000,000. Second, a well-established principle of probability is that if a space is partitioned into events, the sum of all the probabilities must be 1. But the sum of 1/(2t) for t from 1 to n can never be 1, since it is .91666... for n = 3, and 1.041666... for n = 4. Third, it doesn't take into account the character of the assertion. If I asserted in 1975 that "people will write the year 2000 on their checks", this would clearly not be fulfilled until 25 years later. Yet it would occur with probability 1 (or at least close to 1), not 1/50 as the "time principle" suggests.

Is the "time principle" used by statisticians, as Bittinger claims? I used MathSciNet, the online version of Mathematical Reviews, a review journal that attempts to review every noteworthy mathematical publication. I found no references to this principle anywhere in the literature. I then consulted a statistician down the hall at my university, who had never heard of this principle and agreed it was nonsensical.

So Bittinger's "time principle" is pseudomathematics, and is not used by genuine mathematicians. I asked Bittinger where he got it from, and he replied, "Your point is well-taken and I must admit that in some ways the time principle is a stretch. I did "develop" it on my own, and had it corroborated by a top-notch statistician in my department - mathematicians do that you know. I should have said something to this effect, and not "from probability."" I am glad that Bittinger admits that his "time principle" is bogus, and I hope to see a forthright admission to this effect on the website for his book.

Chapter 6 of The Faith Equation discusses the power of prayer. He begins by discussing a controversial study by Randolph C. Byrd that appeared in Southern Medical Journal 81 (7) (July 1988), 826-829, which claimed to show that heart patients showed a statistically signficant benefit from intercessory prayer. Bittinger does not acknowledge any criticism of the Byrd study; indeed, he says, "To a statistician, Byrd's study proved intercessory prayer was effective." But Tessman and Tessman (Skeptical Inquirer (March/April 2000), 31-33 pointed that Byrd's study is bogus for three reasons: the analysis of the results was conducted in a non-blinded fashion by Byrd, the criteria used for evaluating the outcomes were created after the data had been collected, and the study's co-ordinator was non-blinded. Bittinger does not cite the work of Tessman and Tessman, nor other critiques by Sloan, Bagiella, and Powell (Lancet 353 (1999), 664-667) and Posner (Scientific Review of Alternative Medicine 4 (1) (Spring/Summer 2000). By refusing to acknowledge informed criticism of these prayer studies, Bittinger abdicates his responsibility as a professor and an academic.

These two examples should suffice to show how the case in The Faith Equation is so transparently weak that even non-mathematicians should be able to spot the flaws.