Published in the San Diego Union-Tribune, June 27, 2016
The two people I most want to have dinner with are Sigmund Freud and Daniel Kahneman. I want to ask Siggy when is a cigar just a cigar. And I want to ask Danny boy why people make stupid decisions. Freud is not around anymore, but Kahneman is – and he has some thoughts on that subject.
“I look at large organizations that are supposed to be optimal, rational, and the amount of folly in the way these places are run, the stupid procedures that they have, the really, really poor thinking you see all around is actually fairly troubling,” so saith the guru.
As most of you know, Kahneman is a Nobel laureate, and is considered to be the father of behavioral economics – aka why do we make the stupid decisions we make. Here are some nuggets from a recent lecture.
One theme is that we make decisions in a way that is not “commensurate with the complexity and importance of the stakes.” I see this in my own little companies. Decisions get made off the cuff, by the seat of the pants, or by gut instinct. To combat this, I will often stand up and say “this decision really matters” to call attention to its significance. Wave your arms, light a small fire – make sure the team knows that this one has big implications. In other words, slow down the rush to judgment. But that still doesn’t mean you will get to the right decision.
Kahneman says that human beings are hamstrung by “overconfidence, limited attention and cognitive biases.” In other words, we tend to jump to conclusions. We want to tie up the loose ends and complete the narrative before we get halfway through the book.
And he says we are highly susceptible to “noise.” His use of that word implies that our decisions are random and unpredictable. He did a study with radiologists and found that about 20 percent of the time, with the same X-ray, they get a completely different diagnosis.
Kahneman also studied financial institutions (now that is a rich vein indeed for stupid decision making) and found that decisions involving hundreds of thousands of dollars frequently hinged on the opinion of one individual. And when Kahneman presented the exact same set of facts to different groups of financial decision makers, their conclusions varied between 40 to 60 percent. In other words, half the time, at least half of the group had their head – in a different cloud. Wow, tell that to Mr. Dodd and Mr. Frank.
And the topper is this one. Novice financial decision makers had as much variation as experienced professionals. In other words, your loan approval is a toss-up. Now multiply this into the hundreds of millions and apply it to the world of investment bankers – and you understand why 40 percent of M&A deals get unwound within 36 months of making them. (You gotta love those bankers, they come right back in and flip it again for some new fees.)
So the solution is – algorithms. Kahneman is clear, “the indications are unequivocal …. when it comes to decision making, algorithms are superior to people.” Data dictates. And here is the best part – “it does not require a massive amount of data … make a list of five or six dimensions, give them equal weight and you will typically do better than a very sophisticated statistical algorithm.” (Who needs big data? Small seems to work better).
One of my companies recently had a really big decision, a life-death decision that would impact the entire company and its future. Three of us sat in a room with a white board for 3½ hours. We listed four options, and within those options, we reduced each to only two variables. When Kahneman talks about an algorithm, he is really talking about a mathematical decision process – a weighted set of “what ifs” coupled with a decision tree of outcomes A or B or C for each. (And it appears that we got to the right decision – the company is still breathing.)
In the end, for each decision process, he also adds a “global rating.” He calls this his concession to human intuition – the gut thing – and it turns out that the global rating was highly accurate, more than any of the individual assessments alone. But, that correlation could only occur after “the exercise of systematically and independently evaluating the constituents of the problem.”
Rule No. 471
Love the white board.