ORACLE

The magicain's parabable

The group of OR/MS analysts had dissected the weekend’s results of football games and were about ready to leave the water cooler and return to work. One of them asked, “Say! Did you see that letter in the latest issue of OR/MS Today, about the guy who got two papers rejected, 40 years apart, because they ‘didn’t advance the theory’ even though they were good practical applications? How do you like that?”

“Stinks,” Andy, another analyst, responded, to nods all around. “But I think I can top his story. The letter writer, James Ignizio, is a tenured professor, I think with an endowed chair. He’s frustrated, with good reason, but his job isn’t threatened. Now, how about a situation where the person’s manager scoffed at a good application because it wasn’t theoretical enough?”

“You’ve gotta be kidding,” one of the other analysts exclaimed.

“Nope,” Andy related, “true story. Many years ago, I was a federal policy analyst in a group that was supposed to be the analytical hotshots, the people everyone else turned to for hard problems. My boss was another O.R. analyst, at least by classification, but a physicist by training. He thought he was smarter than everyone else in the building and was always trying to prove it.

“So,” Andy continued, “a colleague in another office asked me to tackle a forecasting and inventory problem. After digging around in the data for a few days, I had a simple but solid solution and wrote it up for him. He was happy. He told my boss he was happy. But my boss was not happy, and he told me about it, big time!”

Andy continued, “He explained to me, ‘We’re supposed to be the advanced analysis people, the ones who do the tasks too technically complicated for anyone else. But you didn’t apply any of the techniques people expect from someone with a graduate degree in O.R. This is pretty much just simple arithmetic, with maybe a little algebra. Where have you demonstrated our technical strength?’

“I gently pointed out,” Andy said, “that I’d solved the problem the ‘customer’ brought me, and I thought that was how we showed how smart we were. And even if the math was simple, there was some know-how and logic involved in deciding which numbers to add and multiply, right? To put it mildly, he wasn’t buying that line of reasoning at all.”

“So what happened?” one of the other analysts asked.

“He ordered me to start giving him detailed analysis plans for my projects, showing the advanced techniques I would use,” Andy groaned. “Then he’d criticize the plans and propose more complicated methods, whether they were appropriate or not. Eventually he was directing me to use a technique, I think it was some sort of multivariate regression, when the data clearly violated the assumptions for that technique. I checked with the American Statistical Association Committee on Professional Ethics – they had one at the time, the O.R. societies didn’t and don’t – and they confirmed that I shouldn’t do what he demanded. Using an inappropriate method can skew the conclusions badly. Of course, by this time, I was vigorously hunting for another job, and fortunately I found one. But what a pain!”

At this point Jerry, who had been quiet up to then, interjected, “The real problem is that you didn’t let him keep thinking it was magic.” This got him some puzzled looks.

“I’m an amateur magician – or illusionist, to be more exact – some evenings and weekends,” Jerry explained. “One thing anyone in that line of work learns early is that you never tell people, especially kids, how you do the illusions. It’s not just that you want to limit competition. Many people find magic exciting and explanations a letdown. And if someone paid for that party for their kids, the last thing they want is a letdown. I mean, think about this: if you bought an expensive piece of art and were getting attention and admiration for having it, how eager would you be to have it examined publicly by an expert who could tell if it was counterfeit? Two things can happen. One doesn’t help you much, and the other seriously hurts you. Same way, if your boss was engaging you to do what looked like magic to his bosses, how much would he appreciate your making it all look easy and straightforward? Isn’t that the real objection?”

“I think that also explains the behavior of a lot of editors and associate editors of professional journals,” Mark, another of the group, added dryly. “If most people can understand the articles, they must not be advanced enough, right?” They all laughed, albeit a bit hollowly.

“Makes sense to me,” Andy nodded, “although now I’m the one experiencing a letdown. I thought I was hired to do analysis, not illusions. Viewed this way, what they wanted was not just awkward, it’s even more dangerously close than I realized to being professionally unethical, as it’s misrepresenting the whole profession. So then what should I have done?”

“The best magic trick of all,” Jerry grinned, “and you already know some ways to do it. But maybe you could make it more exciting. Wave your magic wand, chant some secret words, and turn yourself into an employee of somebody else!”

Doug Samuelson (samuelsondoug@yahoo.com) is a senior operations research analyst for IBM in Herndon, Va., working on defense-related analyses. He is also president and chief scientist of InfoLogix, Inc., a small R&D and consulting firm in Annandale, Va.