Just one more big football game coming up, and this week we undertake that riskiest of all editorial stunts by unequivocally predicting the Super Bowl winner for you.
This is an article from the Jan. 14, 1974 issue
Well, that is what we thought we were going to do until late last week. Actually, we now present a pair of predictions, each from a highly qualified source but each arriving at a different bottom line. Tex Maule, the dean of our pro football writers, finally switched to the Vikings for reasons he explains on page 50. Bud Goode, a Los Angeles statistician who believes mightily in computers (and understands them, moreover), tells us, beginning on page 42, why the Dolphins will win, and by how much. Or at least his computer does.
Goode, reports Joe Marshall, who wrote the story about him, is fundamentally opposed to gambling and worries that because he predicts the outcome people will take him for simply another line-maker. But he has spent years refining his techniques with the computer (the jargon runs to phrases like "factor analysis," "correlation matrix" and "multiple regression analysis") and he could no more resist making a prediction on the Super Bowl than Carl Eller could resist sacking Bob Griese if the opportunity arose.
Interviewing Goode, says Marshall, is a little like being in the company of an on-line, turned-on computer. "He's a nonstop talker to begin with and he's always saying computer sorts of things like, 'Statistically, you should never follow a truck. When I drive the freeways, I always think in terms of the numbers.' He uses the word 'correlate' a lot, and once I remarked that everything in his world seemed to correlate. 'Oh, yes,' he said. 'Even boys and girls correlate, but that is a dichotomous variable.' " Once Goode had fed in all the relevant facts, it took the computer less than one second to decide that Miami would win next Sunday, and by how much.
Being human, Maule took a little longer to come up with his own regressive but correlated prediction. And, humanly, he was able to abandon one set of conclusions in favor of another as his analysis progressed last week. Maule doesn't mind computers, forgiving them even their vocabulary—reasonable enough, since he has himself contributed to the esoteric language of pro football. For example, he introduced ST readers to "stunting," the now-fashionable term for certain tricky defensive maneuvers, after the Detroit-Green Bay Thanksgiving Day game of 1962. Nevertheless, Maule believes there are elements in the game that no machine can accommodate; the inventiveness of a coach's mind, for instance, can nullify the variables on which a game plan (or a computer prediction) is based. Maule also believes that balls do take funny bounces and that players can unpredictably rise (or fall) far from their normal levels. Putting it all together, Maule tossed out his first notion, which was that Miami would win, and voted for the Vikings. His choice for Super Bowl VIII was thus consistent and predictable; the Maule Invariable, as he himself recognizes, was at work. He has always picked the NFL (or, since 1969, the NFC) representative. And three-sevenths of the time he has been right.