Thursday, January 26, 2017

The Myth of Certainty

"Doubt is not a pleasant state of mind, but certainty is absurd." - Voltaire

Did you see the movie Moneyball starring Brad Pitt? The movie tells the story of the Oakland A's move from relying on intuition to depending on data when evaluating young baseball players. The A's didn't have the payroll to draw the biggest stars, so they needed a way to find great players that other teams overlooked.


The documentary Linsanity tells the story of NBA player Jeremy Lin, the first Asian-American NBA player. His story was similar to the players in Moneyball: a great player who just didn't look like a great player to scouts because they had no framework for an Asian-American NBA player.

The Undoing Project

In his latest book, The Undoing Project, Moneyball author Michael Lewis explores more deeply why our minds miss obvious evidence of the likely outcome in certain situations. In an interview on CBS This Morning, Lewis summed up his book by saying that one of his discoveries is that often the best person for the job is the person who looks least like our stereotypes. In other words, a doctor who doesn't look anything like we expect a doctor to look has probably had to overcome some major obstacles to get where they are and is most likely to be great at what he or she does. This goes for just about any job--accountant, CEO, NBA player and--I would add--pastor.

The reason for our selective blindness is because of our need to make the world more certain than it is. For years, economists believed that people were basically logical in their choices. Any person would choose the $.69 can of green beans over the $.89 can of green beans. People would not buy lottery tickets once they were informed of their odds of winning.

The Beginning of the End

The work that came out of a partnership between an economist and a psychologist pulled the cornerstone out of the foundation of "utility-based economics," and the edifice has slowly begun to crumble.

People actually will pay more for green beans if they feel like the $.89 can is a better deal--not necessarily for any rational reason. And obviously, quite a few people buy lottery tickets. These two professors (Daniel Kahneman and Amos Tversky) saw this happening in the world around them and wanted to explore why people make illogical choices and how far their illogical their choices would get.

They devised a series of brilliant scenarios to begin gathering quantitative data to support their theories. I highly recommend the book to read about their theories, experiments, and results, but I want to skip to the theory that prompted the book's title: the human tendency towards "undoing" tragedies.

Scenario: Mr. Blue drives the same route to work everyday. One day he takes a different route and is hit by a truck and killed.

A typical first response: "If only he would have taken his normal route! He may still be alive today!"

Our minds want to find ways to "undo" tragedies. We want to find sources of certainty in an uncertain world, and so we imagine scenarios that might restore certainty. But our imaginations are not limitless. They tend to operate within a finite set of parameters. We might even say, "If only he would have left just a few minutes later!"

But we're much less likely to say:
 - "If only he would have discovered that his gas tank was empty and stopped to fill up on the way."
 - "If only he would have discovered that his car was stolen overnight and not been able to drive."
 - "If only his house would have burned down and he wouldn't have gone to work."
And on and on.

In other words, there are an infinite number of "if onlys," but we focus in on a few possibilities in our attempts to "undo" tragedies.

The Temptation of Intuition

One of the temptations is to think that human intuition can somehow influence the outcome of events, as in "if only, I'd followed my intuition and called Mr. Blue that morning."

To test human intuition, Kahneman and Tversky did an experiment where they showed doctors x-rays of a stomach ulcer and asked them to make a guess as to whether it was cancerous. The doctors identified seven factors that they would consider to make their decision. I can't remember all of them (and I already returned the book to the library), but they were things like size, shape, location, etc.

Kahneman and Tversky developed a simple algorithm based on these seven factors to compare with the doctors' guesses. They theorized that their algorithm may need some tweaking that maybe weighted some factors as more significant than others or took other information into account to compete with the accuracy of the doctors.

They made two sets of the x-rays and without the doctors' knowledge actually showed them duplicates.

They were shocked and dismayed to find that not only did the simple algorithm giving each factor equal weight outperform the doctors, but the doctors didn't even make the same diagnosis on the duplicate x-rays! The intuitive opinion of experts was terrifyingly inaccurate.

This so-called intuition that convinces us that we can take our prior experience and hazard a guess as to future outcomes is shockingly wrong. The A's found some great players that other scouts said would never make it. Jeremy Lin is in his seventh year in the NBA (although apparently out right now due to injury).

Often our intuition is right--often enough that we find it mostly works to make decisions based on our gut feelings about a person or a situation.

But it's not always right, and that's what's on my mind.

Uncertainty in the Church

I've recently been involved in two attempts by the church to outguess the uncertainty of our world.

One is the process of licensing new pastors. The other is training pastors to plant churches.

Both are efforts to find certain answers to very uncertain questions. Will this pastor be a good pastor? A lifelong pastor? Will this pastor succumb to moral failure? Or burnout quickly? Will this church plant succeed? Will it survive only as long as its founding pastor is leading? Will it prove to have deep roots?

Image credit: Tibco Blog
The work of Kahneman and Tversky suggests that an algorithm based on a variety of factors common to successful pastors and successful church planters might actually be more accurate than a system based on biased humans evaluating candidates.

And in fact, I've seen aspiring pastors, missionaries, and church planters who were initially rejected prove to be quite successful.

We've attempted to combine data with intuition to help overcome bias, but what if despite our best efforts we actually can't know for sure which pastors and which church plants will succeed? What if we can't ever get our margin of error down to zero? And when we're talking about people who stand to be deeply wounded by our judgments, we really want that margin of error to be zero!

A Third Way

I'd like to offer a third way. I'm going to call it Presence.

There is a temptation to believe that a group of wise and experienced people can enter briefly into the life of a young pastor or church plant and based on their wisdom and experience know with certainty what the future will hold. But our world is terrifyingly uncertain.

I don't think the answer is to follow the path of Moneyball and quantify vast amounts of characteristics and make charts and graphs and data points.

Photo Credit: Olivia Shelton
I think the answer is for those with wisdom and experience to begin walking side by side those on the path. Not only at some fixed point in time (say an interview), but prior to that point when discernment is needed. And after that point when the decision has been made but an uncertain world leads to unexpected circumstances, actions, and emotions.

In a few weeks, I will sit in a room with four or five pastors and they will ask me lots of personal questions in an attempt to determine my future in ministry. They will think they have a good feel for the kind of person I am, what I'm passionate about, the kind of pastor I will be, etc.

But I won't tell them my doubts and fears. I won't tell them my questions and hesitations. I won't tell them anything that will endanger my hope of passing the test.

And--what's probably even more significant--after leaving that room, if I encounter obstacles that make my success uncertain, I won't call them. I won't ask them for help.

Some candidates will have friends, colleagues, and family members who they will turn to for help. Others will try to face their challenges alone. But I doubt that anyone will find this process of evaluating a candidate as a source of aid in the coming days.

There is a temptation to think that if they would have asked better questions, deeper questions, more penetrating questions, they could have somehow predicted the future with more certainty. But I think we need to stop trying to find the answer to the question, "Is this person more likely to succeed or to fail?" because the world is a lot less certain than we like to think it is.

Instead, I think we need to start asking, "How can my presence in this person's life make their success--however likely or unlikely it was to begin with--more probable going forward?"

A Few Suggestions

Some practical ways to implement this theory:

1. Find ways to educate pastors within the context of their current congregation, rather than relocating them. Vineyard Leadership Institute is a good example of this model. They combine video courses taught by excellent professors with discussion and application in local congregations. In other words, don't take pastors away from the people who know their story and background and are trusted advisors. I recognize, of course, that many pastors don't have that kind of support in their local congregations, but it's still worth considering ways to make this a higher priority.

2. Offer connections. I recognize that not every person being evaluated will connect with every evaluator, but candidates might benefit from a list of pastors/church planters who have said, "Call me, email me, text me, ask me questions, ask for prayer. I'm available."

3. Let data determine probability of success, while people provide wholehearted support. There is a part of me that cringes at this idea, but Kahneman and Tversky's work has shown the power of data in all kinds of fields. The cost of a professional athlete makes gathering this data cost-effective, but I doubt that the resources would be available to gather and analyze the data to determine the viability of pastoral candidates. Without that data, our time is probably more well-spent simply doing everything we can to support those coming after us.

A New Paradigm

What I am suggesting is essentially an entirely new paradigm for approaching the way we evaluate potential pastors and church planters. That's kind of a big deal, I know.

But our current paradigm is a weird hodgepodge of data and intuition that has left many candidates feeling attacked, defensive, and more uncertain of their chances of success after going through the process than when they began. On the other hand, it's resulted in others having an elevated opinion of their qualities, which has led to arrogance and a lack of compassion towards those they have been called to serve.

Kahneman and Tversky's work was dismissed or criticized for a long time because of the paradigm shift that it meant for economists, political leaders, business leaders, and anyone else who tried to build a system based on people's decisions, but it slowly began to enter into the public consciousness.

I think the same conversation needs to happen in the church. How certain are we? And if we can't be certain, what can we do instead?