INFORMS annual meeting day zero: Decision Analysis celebrates its 50th birthday

Decision Analysis has reached a certain age. As with most ideas, decision analysis came about over a period of time, but the official methodological dating committee has deemed Ron Howard's 1964 paper, "Decision Analysis: Applied Decision Theory" to be the day the stork arrived. To celebrate, DAS and SDP organized a pre-informs session featuring talks, drinks and dinner, held the day before the start of the INFORMS annual meeting (call it INFORMS day zero).

Has Decision Analysis aged as well?

To get myself in the right frame of mind for the "birthday party", I read several sections of Duncan Luce & Howard Raiffa's "Games and Decisions" (I'd like to say "reread", but in truth I never got around to reading it before now). Published in 1957, the book falls into the "late pre DA" period, but its treatments of utility theory and decisions under uncertainty sound right on today. DA practice has gone through a number of changes over the years, but the foundations have held up well.

One point of difference that did strike me is the classification of games/problems into decision making under 1) certainty, 2) risk and 3) uncertainty. Like most technical terms that overlap common language, the words "risk" and "uncertainty" have plenty of meanings to choose from, but in DA they are often treated as near synonyms, although risk conveys more connotations of a downside. Raiffa's definition of a decision under risk is a situation where the possible outcomes and their probabilities are known, such as the toss of a fair coin, a legitimate gambling game, etc. When the probabilities are themselves unknown, he defines that decision as one under uncertainty. The under uncertainty category also includes less tractable problems such as those featuring what we today call Rumsfeldian "unknown unknowables". By these definitions I do nearly all my work in the "under uncertainty" area.

The birthday party was separated into two parts, history and the future. I am normally a big fan of history, whether it's something like DA that pertains to my work or not, but since I've heard many of the stories the founding fathers of DA like to tell I wasn't expecting to learn anything new. That said, there is something to seeing and hearing the giants of the field all in one place. With Raiffa particularly I'm reminded of the old quip that there's nothing really special about Shakespeare, all he did was knit together a bunch of well known cliches.

Several attendees remarked to me that in aggregate the history presentations were heavily weighted to the first half of those 50 years, and I had to agree. While the field has had many successes over the decades, I hope the next 25 years are more groundbreaking than the last.

For the "future" session, many of the presentations were still stuck in history, but one notable exception was the talk by Katherine Weller, an ADA colleague. Before the session she polled a number of people, including me, about our aspirations for the field 25 years from now, and then gave a very engaging backcasting talk, spinning our goals into a future history talk from 2039.

Happy birthday DA! Onward and upward.