ORacle

The parable of the termites

Doug Samuelson
samuelsondoug@yahoo.com

The group of OR/MS analysts had reconvened once again, shortly after the Presidential election. “Well, Ben,” Brett said, “it looks as if all the quant experts were wrong about this one, doesn’t it?”

“Not all,” Ben responded. “John Zogby, one of the best pollsters in the country, said when the FBI memo came out that Hillary had already been on a downward trend for a week or more before that. And Allan Lichtman’s ‘13 Keys’ model was calling it close, and eventually predicting a Trump win, when the polls had Hillary with a double-digit lead.

“But you know what else is interesting? Those two don’t call themselves OR/MS analysts! Zogby is a historian – without a non-honorary doctorate – and Allan Lichtman usually describes himself as a quantitative historian. And the fact is, lots of people, not just OR/MS analysts, relied too heavily on trend analyses combined with polls, so when the polls went wrong, so did everything else. That’s how Nate Silver came up, the morning of Election Day, with a probability of 75 percent of Hillary winning. And other ‘experts’ called it as high as 99 percent.”

“So it wasn’t just totally swung by the FBI memos?” Brett asked.

Ben replied, “No! The FBI memo was like a guy kicking a wall in an old house, and having the wall immediately fall down. Of course it wouldn’t have fallen down right then if he hadn’t kicked it. But it wouldn’t have fallen down from one kick if it hadn’t been full of termites. The termites are the main problem, not the kick.

“And, of course,” Ben added, “you know that Hillary and her people are still blaming the FBI director, the media, dirty tactics by the other side – anything except that they just plain missed where some key voting blocs were going! And this after the Democratic Party chairs in those Upper Midwest states had been screaming for weeks that they needed some new ad buys and personal appearances with a strong economic message! And meanwhile Trump thinks the outcome validated everything he did, so he and his people aren’t looking at lessons learned, either. In fact, some of them had read the same polls and were surprised when they won, which helps to explain the chaotic transition.”

“Wow! Impressive,” Jim grunted sarcastically.

“Even worse than that,” Ben grimaced, “is the reaction I’ve gotten from some people in the profession to my own analyses. I pointed out before the election that both campaigns were concentrating on states Hillary had been thought to have locked up – a sure sign that her people suspected she was in trouble there. And the response was, ‘Interesting, but where’s the O.R.?’”

“What?” Brett exclaimed, clearly puzzled.

“I’ll tell you where the O.R. is,” Ben growled. “All too often, the kind of O.R. that insists on higher math and ignores what works ends up in the windowless rooms in the basements, as Russ Ackoff predicted 35 years ago, while other professions take over the actual business of influencing important decisions. And then those over-specialized analysts grind out articles and presentations lamenting the foolishness of the decision-makers who don’t realize that they should value O.R. They ignore evidence, sticking to models that have stopped working, and then blame everyone else for the bad results. Where else have we seen this lately?”

“Ugh,” Jim and Brett chorused. “Sounds like it’s a pandemic. So what do we do about it?”

“Well, I don’t know about anyone else,” Ben said softly, “but I learned a long time ago not to keep trying to sell analysis to people who are too stubborn or stupid or doctrinaire to value it. Let the people who are responsible for electing this administration deal with its problems. Whether or not I leave the country, as some people I know have contemplated, I think it’s time to take a hard look at the OR/MS profession. Nowadays I call myself a decision scientist, or just a proven trustworthy problem-solver, and forget about trying to explain what ‘operations research’ is. O.R. in its heyday was applied problem-solving by interdisciplinary teams. Nobody had an O.R. degree or any preset definition of what methods were useful. The more defined the discipline becomes, the less useful it is – and the more resistant it is to being dragged back to reality. “

“You sound upset,” Jim remarked. “Are you sure about this?”

“Yeah,” Ben affirmed. “Now’s the time when good O.R. could be tremendously helpful to the country. If O.R. is a uniquely productive way of looking at the world, then the future of the profession is bright. But if it’s codifying a set of ways to analyze and then ossifying around those, it might be doomed. Fortunately, there have been a few people like Gene Woolsey and Russ Ackoff and Saul Gass to remind us how to do it, but I fear people like them are scarce and getting scarcer.”

“That sounds serious,” Jim exclaimed.

“I think we’ll survive,” Ben shrugged. “But those exceptional analysts, willing to get immersed in the real problems, tackle them creatively and notice evidence of going wrong – are our hope for the future. Meanwhile, the campaign hotshots, some members of the OR/MS profession, and yes, Donald Trump and his top staff should remember an old proverb: The most important learning you’ll ever do happens shortly after you know it all!”

Doug Samuelson (samuelsondoug@yahoo.com) is president and chief scientist of InfoLogix, Inc., in Annandale, Va. He is a veteran of a number of political campaigns and related analyses in addition to his long career as a federal policy analyst and consultant.