Category: Decision Making

Do Algorithms Beat Us at Complex Decision Making?

Decision-making algorithms are undoubtedly controversial. If a decision is being made that will have a major influence on your life, most people would prefer a human make it. But what if algorithms really can make better decisions?

***

Algorithms are all the rage these days. AI researchers are taking more and more ground from humans in areas like rules-based games, visual recognition, and medical diagnosis. However, the idea that algorithms make better predictive decisions than humans in many fields is a very old one.

In 1954, the psychologist Paul Meehl published a controversial book with a boring sounding name: Clinical vs. Statistical Prediction: A Theoretical Analysis and a Review of the Evidence.

The controversy? After reviewing the data, Meehl claimed that mechanical, data-driven algorithms could better predict human behavior than trained clinical psychologists — and with much simpler criteria. He was right.

The passing of time has not been friendly to humans in this game: Studies continue to show that the algorithms do a better job than experts in a range of fields. In Daniel Kahneman’s Thinking Fast and Slow, he details a selection of fields which have demonstrated inferior human judgment compared to algorithms:

The range of predicted outcomes has expanded to cover medical variables such as the longevity of cancer patients, the length of hospital stays, the diagnosis of cardiac disease, and the susceptibility of babies to sudden infant death syndrome; economic measures such as the prospects of success for new businesses, the evaluation of credit risks by banks, and the future career satisfaction of workers; questions of interest to government agencies, including assessments of the suitability of foster parents, the odds of recidivism among juvenile offenders, and the likelihood of other forms of violent behavior; and miscellaneous outcomes such as the evaluation of scientific presentations, the winners of football games, and the future prices of Bordeaux wine.

The connection between them? Says Kahneman: “Each of these domains entails a significant degree of uncertainty and unpredictability.” He called them “low-validity environments”, and in those environments, simple algorithms matched or outplayed humans and their “complex” decision making criteria, essentially every time.

***

A typical case is described in Michael Lewis’ book on the relationship between Daniel Kahneman and Amos Tversky, The Undoing Project. He writes of work done at the Oregon Research Institute on radiologists and their x-ray diagnoses:

The Oregon researchers began by creating, as a starting point, a very simple algorithm, in which the likelihood that an ulcer was malignant depended on the seven factors doctors had mentioned, equally weighted. The researchers then asked the doctors to judge the probability of cancer in ninety-six different individual stomach ulcers, on a seven-point scale from “definitely malignant” to “definitely benign.” Without telling the doctors what they were up to, they showed them each ulcer twice, mixing up the duplicates randomly in the pile so the doctors wouldn’t notice they were being asked to diagnose the exact same ulcer they had already diagnosed. […] The researchers’ goal was to see if they could create an algorithm that would mimic the decision making of doctors.

This simple first attempt, [Lewis] Goldberg assumed, was just a starting point. The algorithm would need to become more complex; it would require more advanced mathematics. It would need to account for the subtleties of the doctors’ thinking about the cues. For instance, if an ulcer was particularly big, it might lead them to reconsider the meaning of the other six cues.

But then UCLA sent back the analyzed data, and the story became unsettling. (Goldberg described the results as “generally terrifying”.) In the first place, the simple model that the researchers had created as their starting point for understanding how doctors rendered their diagnoses proved to be extremely good at predicting the doctors’ diagnoses. The doctors might want to believe that their thought processes were subtle and complicated, but a simple model captured these perfectly well. That did not mean that their thinking was necessarily simple, only that it could be captured by a simple model.

More surprisingly, the doctors’ diagnoses were all over the map: The experts didn’t agree with each other. Even more surprisingly, when presented with duplicates of the same ulcer, every doctor had contradicted himself and rendered more than one diagnosis: These doctors apparently could not even agree with themselves.

[…]

If you wanted to know whether you had cancer or not, you were better off using the algorithm that the researchers had created than you were asking the radiologist to study the X-ray. The simple algorithm had outperformed not merely the group of doctors; it had outperformed even the single best doctor.

The fact that doctors (and psychiatrists, and wine experts, and so forth) cannot even agree with themselves is a problem called decision making “noise”: Given the same set of data twice, we make two different decisions. Noise. Internal contradiction.

Algorithms win, at least partly, because they don’t do this: The same inputs generate the same outputs every single time. They don’t get distracted, they don’t get bored, they don’t get mad, they don’t get annoyed. Basically, they don’t have off days. And they don’t fall prey to the litany of biases that humans do, like the representativeness heuristic.

The algorithm doesn’t even have to be a complex one. As demonstrated above with radiology, simple rules work just as well as complex ones. Kahneman himself addresses this in Thinking, Fast and Slow when discussing Robyn Dawes’s research on the superiority of simple algorithms using a few equally-weighted predictive variables:

The surprising success of equal-weighting schemes has an important practical implication: it is possible to develop useful algorithms without prior statistical research. Simple equally weight formulas based on existing statistics or on common sense are often very good predictors of significant outcomes. In a memorable example, Dawes showed that marital stability is well predicted by a formula: Frequency of lovemaking minus frequency of quarrels.

You don’t want your result to be a negative number.

The important conclusion from this research is that an algorithm that is constructed on the back of an envelope is often good enough to compete with an optimally weighted formula, and certainly good enough to outdo expert judgment. This logic can be applied in many domains, ranging from the selection of stocks by portfolio managers to the choices of medical treatments by doctors or patients.

Stock selection, certainly a “low validity environment”, is an excellent example of the phenomenon.

As John Bogle pointed out to the world in the 1970’s, a point which has only strengthened with time, the vast majority of human stock-pickers cannot outperform a simple S&P 500 index fund, an investment fund that operates on strict algorithmic rules about which companies to buy and sell and in what quantities. The rules of the index aren’t complex, and many people have tried to improve on them with less success than might be imagined.

***

Another interesting area where this holds is interviewing and hiring, a notoriously difficult “low-validity” environment. Even elite firms often don’t do it that well, as has been well documented.

Fortunately, if we take heed of the advice of the psychologists, operating in a low-validity environment has rules that can work very well. In Thinking Fast and Slow, Kahneman recommends fixing your hiring process by doing the following (or some close variant), in order to replicate the success of the algorithms:

Suppose you need to hire a sales representative for your firm. If you are serious about hiring the best possible person for the job, this is what you should do. First, select a few traits that are prerequisites for success in this position (technical proficiency, engaging personality, reliability, and so on). Don’t overdo it — six dimensions is a good number. The traits you choose should be as independent as possible from each other, and you should feel that you can assess them reliably by asking a few factual questions. Next, make a list of questions for each trait and think about how you will score it, say on a 1-5 scale. You should have an idea of what you will call “very weak” or “very strong.”

These preparations should take you half an hour or so, a small investment that can make a significant difference in the quality of the people you hire. To avoid halo effects, you must collect the information one at a time, scoring each before you move on to the next one. Do not skip around. To evaluate each candidate, add up the six scores. […] Firmly resolve that you will hire the candidate whose final score is the highest, even if there is another one whom you like better–try to resit your wish to invent broken legs to change the ranking. A vast amount of research offers a promise: you are much more likely to find the best candidate if you use this procedure than if you do what people normally do in such situations, which is to go into the interview unprepared and to make choices by an overall intuitive judgment such as “I looked into his eyes and liked what I saw.”

In the battle of man vs algorithm, unfortunately, man often loses. The promise of Artificial Intelligence is just that. So if we’re going to be smart humans, we must learn to be humble in situations where our intuitive judgment simply is not as good as a set of simple rules.

Blog Posts, Book Reviews, and Abstracts: On Shallowness

We’re quite glad that you read Farnam Street, and we hope we’re always offering you a massive amount of value. (If not, email us and tell us what we can do more effectively.)

But there’s a message all of our readers should appreciate: Blog posts are not enough to generate the deep fluency you need to truly understand or get better at something. We offer a starting point, not an end point.

This goes just as well for book reviews, abstracts, cliff’s notes, and a good deal of short-form journalism.

This is a hard message for some who want a shortcut. They want the “gist” and the “high level takeaways”, without doing the work or eating any of the broccoli. They think that’s all it takes: Check out a 5-minute read, and instantly their decision making and understanding of the world will improve right-quick. Most blogs, of course, encourage this kind of shallowness. Because it makes you feel that the whole thing is pretty easy.

Here’s the problem: The world is more complex than that. It doesn’t actually work this way. The nuanced detail behind every “high level takeaway” gives you the context needed to use it in the real world. The exceptions, the edge cases, and the contradictions.

Let me give you an example.

A high-level takeaway from reading Kahneman’s Thinking Fast, and Slow would be that we are subject to something he and Amos Tversky call the Representativeness Heuristic. We create models of things in our head, and then fit our real-world experiences to the model, often over-fitting drastically. A very useful idea.

However, that’s not enough. There are so many follow-up questions. Where do we make the most mistakes? Why does our mind create these models? Where is this generally useful? What are the nuanced examples of where this tendency fails us? And so on. Just knowing about the Heuristic, knowing that it exists, won’t perform any work for you.

Or take the rise of human species as laid out by Yuval Harari. It’s great to post on his theory; how myths laid the foundation for our success, how “natural” is probably a useless concept the way it’s typically used, and how biology is the great enabler.

But Harari’s book itself contains the relevant detail that fleshes all of this out. And further, his bibliography is full of resources that demand your attention to get even more backup. How did he develop that idea? You have to look to find out.

Why do all this? Because without the massive, relevant detail, your mind is built on a house of cards.

What Farnam Street and a lot of other great resources give you is something like a brief map of the territory.

Welcome to Colonial Williamsburg! Check out the re-enactors, the museum, and the theatre. Over there is the Revolutionary City. Gettysburg is 4 hours north. Washington D.C. is closer to 2.5 hours.

Great – now you have a lay of the land. Time to dig in and actually learn about the American Revolution. (This book is awesome, if you actually want to do that.)

Going back to Kahneman, one of his and Tversky’s great findings was the concept of the Availability Heuristic. Basically, the mind operates on what it has close at hand.

As Kahneman puts it, “An essential design feature of the associative machine is that it represents only activated ideas. Information that is not retrieved (even unconsciously) from memory might as well not exist. System 1 excels at constructing the best possible story that incorporates ideas currently activated, but it does not (cannot) allow for information it does not have.”

That means that in the moment of decision making, when you’re thinking hard on some complex problem you face, it’s unlikely that your mind is working all that successfully without the details. It doesn’t have anything to draw on. It’d be like a chess player who read a book about great chess players, but who hadn’t actually studied all of their moves. Not very effective.

The great difficulty, of course, is that we lack the time to dig deep into everything. Opportunity costs and trade-offs are quite real.

That’s why you must develop excellent filters. What’s worth learning this deeply? We think it’s the first-principle style mental models. The great ideas from physical systems, biological systems, and human systems. The new-new thing you’re studying is probably either A. Wrong or B. Built on one of those great ideas anyways. Farnam Street, in a way, is just a giant filtering mechanism to get you started down the hill.

But don’t stop there. Don’t stop at the starting line. Resolve to increase your depth and stop thinking you can have it all in 5 minutes or less. Use our stuff, and whoever else’s stuff you like, as an entrée to the real thing.

Breaking the Rules: Moneyball Edition

Most of the book Simple Rules by Donald Sull and Kathleen Eisenhardt talks about identifying a problem area (or an area ripe for “simple rules”) and then walks you through creating your own set of rules. It’s a useful mental process.

An ideal situation for simple rules is something repetitive, giving you constant feedback so you can course correct as you go. But what if your rules stop working and you need to start over completely?

Simple Rules recounts the well-known Moneyball tale in its examination of this process:

The story begins with Sandy Alderson. Alderson, a former Marine with no baseball background became the A’s general manager in 1983. Unlike baseball traditionalists, Alderson saw scoring runs as a process, not an outcome, and imagined baseball as a factory with a flow of players moving along the bases. This view led Alderson and later his protege and replacement, Billy Beane, to the insight that most teams overvalue batting average (hits only) and miss the relevance of on-base percentage (walks plus hits) to keeping the runners moving. Like many insightful rules, this boundary rule of picking players with a high on base percentage has subtle second – and third-order effects. Hitters with a high on-base percentage are highly disciplined (i.e., patient, with a good eye for strikes). This means they get more walks, and their reputation for discipline encourages pitchers to throw strikes, which are easier to hit. They tire out pitchers by making them throw more pitches overall, and disciplined hitting does not erode much with age. These and other insights are at the heart of what author Michael Lewis famously described as moneyball.

The Oakland A’s did everything right, they had examined the issues, they tried to figure out those areas which would most benefit from a set of simple rules and they had implemented them. The problem was, they were easy rules to copy. 

They were operating in a Red Queen Effect world where everyone around them was co-evolving, where running fast was just enough to get ahead temporarily, but not permanently. The Red Sox were the first and most successful club to copy the A’s:

By 2004, a free-spending team, the Boston Red Sox, co-opted the A’s principles and won the World Series for the first time since 1918. In contrast, the A’s went into decline, and by 2007 the were losing more games than they were winning Moneyball had struck out.

What can we do when the rules stop working? 

We must break them.

***

When the A’s had brought in Sandy Alderson, he was an outsider with no baseball background who could look at the problem in a different and new light. So how could that be replicated?

The team decided to bring in Farhan Zaidi as director of baseball operations in 2009. Zaidi spent most of his life with a pretty healthy obsession for baseball but he had a unique background: a PhD in behavioral economics.

He started on the job of breaking the old rules and crafting new ones. Like Andy Grove did once upon a time with Intel, Zaidi helped the team turn and face a new reality. Sull and Eisenhardt consider this as a key trait:

To respond effectively to major change, it is essential to investigate the new situation actively, and create a reimagined vision that utilizes radically different rules.

The right choice is often to move to the new rules as quickly as possible. Performance will typically decline in the short run, but the transition to the new reality will be faster and more complete in the long run. In contrast, changing slowly often results in an awkward combination of the past and the future with neither fitting the other or working well.

Beane and Zaidi first did some house cleaning: They fired the team’s manager. Then, they began breaking the old Moneyball rules, things like avoiding drafting high-school players. They also decided to pay more attention to physical skills like speed and throwing.

In the short term, the team performed quite poorly as fan attendance showed a steady decline. Yet, once again, against all odds, the A’s finished first in their division in 2012. Their change worked. 

With a new set of Simple Rules, they became a dominant force in their division once again. 

Reflecting their formidable analytic skills, the A’s brass had a new mindset that portrayed baseball as a financial market rife with arbitrage possibilities and simple rules to match.

One was a how-to rule that dictated exploiting players with splits. Simply put, players with splits have substantially different performances in two seemingly similar situations. A common split is when a player hits very well against right-handed pitchers and poorly against left-handed pitchers, or vice versa. Players with spits are mediocre when they play every game, and are low paid. In contrast, most superstars play well regardless of the situation, and are paid handsomely for their versatility. The A’s insight was that when a team has a player who can perform one side of the split well and a different player who excels at the opposite split, the two positives can create a cheap composite player. So the A’s started using a boundary rule to pick players with splits and how-to rule to exploit those splits with platooning – putting different players at the same position to take advantage of their splits against right – or left-handed pitching.

If you’re reading this as a baseball fan, you’re probably thinking that exploiting splits isn’t anything new. So why did it have such an effect on their season? Well, no one had pushed it this hard before, which had some nuanced effects that might not have been immediately apparent.

For example, exploiting these splits keeps players healthier during the long 162-game season because they don’t play every day. The rule keeps everyone motivated because everyone has a role and plays often. It provides versatility when players are injured since players can fill in for each other.

They didn’t stop there. Zaidi and Beane looked at the data and kept rolling out new simple rules that broke with their highly successful Moneyball past.

In 2013 they added a new boundary rule to the player-selection activity: pick fly-ball hitters, meaning hitters who tend to hit the ball in the air and out of the infield (in contrast with ground-ball hitters). Sixty percent of the A’s at-bat were by fly-ball hitters in 2013, the highest percentage in major-league baseball in almost a decade, and the A’s had the highest ratio of fly ball to ground balls, by far. Why fly-ball hitters?

Since one of ten fly balls is a home run, fly-ball hitters hit more home runs: an important factor in winning games. Fly-ball hitters also avoid ground-ball double plays, a rally killer if ever there as one. They are particularly effective against ground-ball pitches because they tend to swing underneath the ball, taking way the advantage of those pitchers. In fact, the A’s fly-ball hitters batted an all-star caliber .302 against ground-ball pitchers in 2013 on their way to their second consecutive division title despite having the fourth-lowest payroll in major-league baseball.

Unfortunately, the new rules had a short-lived effectiveness: In 2014 the A’s fell to 2nd place and have been struggling the last two seasons. Two Cinderella stories is a great achievement, but it’s hard to maintain that edge. 

This wonderful demonstration of the Red Queen Effect in sports can be described as an “arms race.’” As everyone tries to get ahead, a strange equilibrium is created by the simultaneous continual improvement, and those with more limited resources must work even harder as the pack moves ahead one at a time.

Even though they have adapted and created some wonderful “Simple Rules” in the past, the A’s (and all of their competitors) must stay in the race in order to return to the top: No “rule” will allow them to rest on their laurels. Second-Order Thinking and a little real-world experience show this to be true: Those that prosper consistently will think deeply, reevaluate, adapt, and continually evolve. That is the nature of a competitive world. 

Simple Rules for Business Strategy

The book Simple Rules by Donald Sull and Kathleen Eisenhardt has a very interesting chapter on strategy, which tries to answer the following question: How do you translate your broad objectives into a strategy that can provide guidelines for your employees from day to day?

It’s the last bit there which is particularly important — getting everyone on the same page. 

Companies don’t seem to have a problem creating broad objectives (which isn’t really a strategy). Your company might not call them that, they might call them “mission statements” or simply “corporate goals.”  They sound all well and good, but very little thought is given to how we will actually implement these lofty goals.

As Sull and Eisenhardt put it: 

Developing a strategy and implementing it are often viewed as two distinct activities — first you come up with the perfect plan and then you worry about how to make it happen. This approach, common through it is, creates a disconnect between what a company is trying to accomplish and what employees do on a day-to-day basis.

The authors argue that companies can bridge this gap between strategic intent and actual implementation by following three steps:

  1. Figure out what will move the needles.
  2. Choose a bottleneck.
  3. Craft the rules.

1. Moving the Needles

The authors use a dual needle metaphor to visualize corporate profits. They see it as two parallel needles: an upper needle which represents revenues and a lower needle which represents costs. The first critical step is to identify which actions will drive a wedge between the needles causing an increase in profits, a decrease in costs, and sustain this over time.

In other words, as simple as it sounds, we need an actual set of steps to get from figure a. to figure b.

screen-shot-2016-10-17-at-1-36-10-pm

What action will become the wedge that will move the needles?

The authors believe the best way to answer this is to sit down with your management team and ask them to work as a group to answer the following three questions:

  1. Who will we target as customers?
  2. What product or service will we offer?
  3. How will we provide this product at a profit?

When you are trying to massage out these answers remember to use inversion as well. 

Equally important are the choices on who not to serve and what not to offer.

Steve Jobs once pointed out that Apple was defined as much by what it didn’t do as by what it did.

2. Bottlenecks

Speaking of inversion, in order to complete our goal we must also figure out what’s holding us back from moving the needles — the bottlenecks standing in our way.

When it comes to implementing a strategy of simple rules, pinpointing the precise decision or activity where rules will have the most impact is half the battle. We use the term bottleneck to describe a specific activity or decision that hinders a company from moving the needles.

You may be surprised at the amount of bottlenecks you come across, so you’ll have to practice some “triage” of your issues, sorting what’s important from what’s really important.

The authors believe that the best bottlenecks to focus your attention on share three characteristics:

  1. They have a direct and significant impact on value creation.
  2. They should represent recurrent decisions (as opposed to ‘one off’ choices).
  3. They should be obstacles that arise when opportunities exceed available resources.

Once we’ve established what the bottlenecks are, it’s time to craft the rules which will provide you a framework in which to remove them.

3. Craft the Rules

Developing rules from the top down is a big mistake. When leaders rely on their gut instincts, they overemphasize recent events, build in their personal biases, and ignore data that doesn’t fit with their preconceived notions. It is much better to involve a team, typically ranging in size from four to eight members, and use a structured process to harness members’ diverse insights and points of view. When drafting the dream team to develop simple rules, it is critical to include some of the people who will be using them on a day-to-day basis.

This probably seems like common sense but we’re guessing you have worked at least one place where all information and new initiatives came from above, and much of it seemingly came out of nowhere because you weren’t likely involved.

In these situations it’s very hard to get buy-in from the employees — yet they are the ones doing the work, implementing the rules. So we need to think about their involvement from the beginning.

Having users make the rules confers several advantages. First, they are closest to the facts on the ground and best positioned to codify experience into usable rules. Because they will make decisions based on the rules, they can strike the right balance between guidance and discretion, avoiding rules that are overly vague or restrictive. User can also phrase the rules in language that resonates for them, rather than relying on business jargon. By actively participating in the process, users are more likely to buy into the final rules and therefore apply them in practice. Firsthand knowledge also makes it easier to explain the rules, and their underlying rationale, to colleagues who did not participate in the process.

It’s important to note here that this is a process, a process in which you are never done – there is no real finish line. You must always plan to learn and to iterate as you learn — keep changing the plan as new information comes in. Rigidity to a plan is not a virtue; learning and adapting are virtues

***

There’s nothing wrong with strategy. In fact, without a strategy, it’s hard to figure out what to do; some strategy or another must guide your actions as an organization. But it’s simply not enough: Detailed execution, at the employee level, is what gets things done. That’s what the Simple Rules are all about.

Strategy, in our view, lives in the simple rules that guide an organization’s most important activities. They allow employees to make on-the-spot decisions and seize unexpected opportunities without losing sight of the big picture.

The process you use to develop simple rules matters as much as the rules themselves. Involving a broad cross-section of employees, for example, injects more points of view into the discussion, produces a shared understanding of what matters for value creation, and increases buy-in to the simple rules. Investing the time up front to clarify what will move the needles dramatically increases the odds that simple rules will be applied where they can have the greatest impact.

***

Still Interested? Read the book, or check out our other post where we cover the details of creating your simple rules.

Choosing your Choice Architect(ure)

“Nothing will ever be attempted
if all possible objections must first be overcome.”

— Samuel Johnson

***

In the book Nudge by Richard Thaler and Cass Sunstein they coin the terms ‘Choice Architecture’ and ‘Choice Architect’. For them, if you have an ability to influence the choices other people make, you are a choice architect.

Considering the number of interactions we have everyday, it would be quite easy to argue that we are all Choice Architects at some point. But this also makes the inverse true; we are also wandering around someone else’s Choice Architecture.

Let’s take a look at a few of the principles of good choice architecture, so we can get a better idea of when someone is trying to nudge us.

This information can then be used/weighed when making decisions.  

Defaults

Thaler and Sunstein start with a discussion on “defaults” that are commonly offered to us:

For reasons we have discussed, many people will take whatever option requires the least effort, or the path of least resistance. Recall the discussion of inertia, status quo bias, and the ‘yeah, whatever’ heuristic. All these forces imply that if, for a given choice, there is a default option — an option that will obtain if the chooser does nothing — then we can expect a large number of people to end up with that option, whether or not it is good for them. And as we have also stressed, these behavioral tendencies toward doing nothing will be reinforced if the default option comes with some implicit or explicit suggestion that it represents the normal or even the recommended course of action.

When making decisions people will often take the option that requires the least effort or the path of least resistance. This makes sense: It’s not just a matter of laziness, we also only have so many hours in a day. Unless you feel particularly strongly about it, if putting little to no effort towards something leads you forward (or at least doesn’t noticeably kick you backwards) this is what you are likely to do. Loss aversion plays a role as well. If we feel like the consequences of making a poor choice are high, we will simply decide to do nothing. 

Inertia is another reason: If the ship is currently sailing forward, it can often take a lot of time and effort just to slightly change course.

You have likely seen many examples of inertia at play in your work environment and this isn’t necessarily a bad thing.

Sometimes we need that ship to just steadily move forward. The important bit is to realize when this is factoring into your decisions, or more specifically, when this knowledge is being used to nudge you into making specific choices.

Let’s think about some of your monthly recurring bills. While you might not be reading that magazine or going to the gym, you’re still paying for the ability to use that good or service. If you weren’t being auto-renewed monthly, what is the chance that you would put the effort into renewing that subscription or membership? Much lower, right? Publishers and gym owners know this, and they know you don’t want to go through the hassle of cancelling either, so they make that difficult, too. (They understand well our tendency to want to travel the path of least resistance and avoid conflict.)

This is also where they will imply that the default option is the recommended course of action. It sounds like this:

“We’re sorry to hear you no longer want the magazine Mr. Smith. You know, more than half of the fortune 500 companies have a monthly subscription to magazine X, but we understand if it’s not something you’d like to do at the moment.”

or

“Mr. Smith we are sorry to hear that you want to cancel your membership at GymX. We understand if you can’t make your health a priority at this point but we’d love to see you back sometime soon. We see this all the time, these days everyone is so busy. But I’m happy to say we are noticing a shift where people are starting to make time for themselves, especially in your demographic…”

(Just cancel them. You’ll feel better. We promise.)

The Structure of Complex Choices

We live in a world of reviews. Product reviews, corporate reviews, movie reviews… When was the last time you bought a phone or a car before checking the reviews? When was the last time that you hired an employee without checking out their references? 

Thaler and Sunstein call this Collaborative Filtering and explain it as follows:

You use the judgements of other people who share your tastes to filter through the vast number of books or movies available in order to increase the likelihood of picking one you like. Collaborative filtering is an effort to solve a problem of choice architecture. If you know what people like you tend to like, you might well be comfortable in selecting products you don’t know, because people like you tend to like them. For many of us, collaborative filtering is making difficult choices easier.

While collaborative filtering does a great job of making difficult choices easier we have to remember that companies also know that you will use this tool and will try to manipulate it. We just have to look at the information critically, compare multiple sources and take some time to review the reviewers.

These techniques can be useful for decisions of a certain scale and complexity: when the alternatives are understood and in small enough numbers. However, once we reach a certain size we require additional tools to make the right decision.

One strategy to use is what Amos Tversky (1972) called ‘elimination by aspects.’ Someone using this strategy first decides what aspect is most important (say, commuting distance), establishes a cutoff level (say, no more than a thirty-minute commute), then eliminates all the alternatives that do not come up to this standard. The process is repeated, attribute by attribute (no more than $1,500 per month; at least two bedrooms; dogs permitted), until either a choice is made or the set is narrowed down enough to switch over to a compensatory evaluation of the ‘finalists.’”

This is a very useful tool if you have a good idea of which attributes are of most value to you.

When using these techniques, we have to be mindful of the fact that the companies trying to sell us goods have spent a lot of time and money figuring out what attributes are important to you as well.

For example, if you were to shop for an SUV you would notice that there are a specific number of variables they all seem to have in common now (engine options, towing options, seating options, storage options). They are trying to nudge you not to eliminate them from your list. This forces you to do the tertiary research or better yet, this forces you to walk into dealerships where they will try to inflate the importance of those attributes (which they do best).

They also try to call things new names as a means to differentiate themselves and get onto your list. What do you mean our competitors don’t have FLEXfuel?

Incentives

Incentives are so ubiquitous in our lives that it’s very easy to overlook them. Unfortunately, this can influence us to make poor decisions.

Thaler and Sunstein believe this is tied into how salient the incentive is.

The most important modification that must be made to a standard analysis of incentives is salience. Do the choosers actually notice the incentives they face? In free markets, the answer is usually yes, but in important cases the answer is no.

Consider the example of members of an urban family deciding whether to buy a car. Suppose their choices are to take taxis and public transportation or to spend ten thousand dollars to buy a used car, which they can park on the street in front of their home. The only salient costs of owning this car will be the weekly stops at the gas station, occasional repair bills, and a yearly insurance bill. The opportunity cost of the ten thousand dollars is likely to be neglected. (In other words, once they purchase the car, they tend to forget about the ten thousand dollars and stop treating it as money that could have been spent on something else.) In contrast, every time the family uses a taxi the cost will be in their face, with the meter clicking every few blocks. So behavioral analysis of the incentives of car ownership will predict that people will underweight the opportunity costs of car ownership, and possibly other less salient aspects such as depreciation, and may overweight the very salient costs of using a taxi.

The problems here are relatable and easily solved: If the family above had written down all the numbers related to either taxi, public transportation, or car ownership, it would have been a lot more difficult for them to undervalue the salient aspects of any of their choices. (At least if the highest value attribute is cost).

***

This isn’t an exhaustive list of all the daily nudges we face but it’s a good start and some important, translatable, themes emerge.

  • Realize when you are wandering around someone’s choice architecture.
  • Do your homework
  • Develop strategies to help you make decisions when you are being nudged.

 

Still Interested? Buy, and most importantly read, the whole book. Also, check out our other post on some of the Biases and Blunders covered in Nudge.

Luck Meets Perseverance: The Creation of IBM’s Competitive Advantage

On Monday October 28, 1929, the stock market took one of the worst single-day tumbles anyone alive might have seen, with the Dow Jones averages falling about 13%. The next day, October 29th, the market dropped yet again, a decline of 12%. By the end of the year, the Dow Jones average was down more than 45% from its high of 381. Market historians are familiar with the rest of the story: The sickening slide would not stop at 45%, but continue until 1932 to reach a low of 41 on the Dow, a decline of about 90% from peak to trough.

American business was in a major Depression. But at least one businessman would decide that, like General Erwin Rommel would say years later, the path was not out, but through.

***

International Business Machines, better known as IBM, was created from the ashes of the Computing-Tabulating-Recording Company (C-T-R) in 1917 by Thomas J. Watson, who’d learned his business skills at the National Cash Register Company (NCR). Before Watson’s reorganization of C-T-R, the company was basically in three businesses: computing scales (to weigh and compute the cost of a product being weighed), time clocks (to calculate and record wages), and tabulating machines (which used punch cards to add up figures and sort them). Watson’s first act of genius was to recognize that the future of IBM was not going to be time cards or scales, but in helping businesses do their work more effectively and with a lot less labor. He set out to do just that with his tabulating machines.

The problem was, IBM’s products weren’t yet all that different from its competitors’, and the company was small. IBM’s main tabulating product was the Hollerith machine, created by Herman Hollerith in Washington D.C. in 1890 to improve the Census tabulating process, of all things. (It sounds mundane, but he saved the government $5 million and did the work in about 1/8th of the time.) By the late 1910s, the Hollerith machine had a major competitor in the Powers Accounting Company, which had a similar product that was easier to use and more advanced than the Hollerith.

Hollerith_Punched_Card
Hollerith Punch Card

 

HollerithMachine.CHM
Hollerith Machine

 

Watson knew he had to push the research and development of his best product, and he did, hiring bright engineers like Fred Carroll from NCR, who would go on to be known for his Carroll Press, which allowed IBM to mass-produce the punch cards which did the “tabulating” in the pre-electronic days. By the mid-1920s, IBM had the lead. The plan was set in late 1927.

Watson then pointed to where he wanted IBM to go. ”There isn’t any limit for the tabulating business for many years to come,” he said. “We have just scratched the surface in this division. I expect the bulk of increased business to come from the tabulating end, because the potentialities are greater, and we have done so little in the way of developing our machines in this field.”

Underneath that statement lay a number of reasons—other than the thrill of new technology—why Watson zeroed in on the punch card business. When seen together, the reasons clicked like a formula for total domination. IBM would never be able to make sure it was the world leader in scales or time clocks, but it could be certain that it was the absolute lord of data processing.

[…]

Watson had no epiphanies. No voice spoke to him about the future of data processing. He didn’t have a grand vision for turning IBM into a punch card company. He got there little by little, one observation after another, over a period of 10 to 12 years.

(Source: The Maverick and his Machine)

Watson’s logical, one-foot-at-a-time approach was reminiscent of Sir William Osler’s dictum: Our main business is not to see what lies dimly at a distance, but to do what lies clearly at hand. And with a strategy of patenting its proprietary punch-cards, making them exclusively usable with IBM tabulators and sorters, IBM was one of the market darlings in the lead-up to 1929. Between 1927 and 1929 alone, IBM rose about four-fold on the back of 20-30% annual growth in its profits.

But it was still a small company with a lot of competition, and the punch card system was notoriously unreliable at times. He had a great system to hook in his customers, but the data processing market was still young — many businesses wouldn’t adopt it. And then came the fall.

***

As the stock market dropped by the day and the Depression got on, the economy itself began to shrink in 1930. GDP went down 8% that year, and then another 7% the following year. Thousands of banks failed and unemployment would eventually test 30%, a figure that itself was misleading; the modern concept of “underemployment” hadn’t been codified, but if it had, it probably would have dwarfed 30%. An architect working as a lowly draftsman had a job, but he’d still fallen on hard times. Everyone had.

Tom Watson’s people wondered what was to become of IBM. If businesses didn’t have money, how could they purchase tabulators and punch cards? Even if it would save them money in the long run, too many businesses had cut their capital spending to the bone. The market for office spending was down 50% in 1930.

Watson’s response was to push. Hard. So hard that he’d take IBM right up to the brink.

IBM could beat the Depression, Watson believed. He reasoned that only 5 percent of business accounting functions were mechanized, leaving a huge market untapped. Surely there was room to keep selling machines, even in difficult times. Watson also reasoned that the need for IBM machines was so great, if businesses put off buying them now, certainly they’d buy them later, when the economy picked up. His logic told him that the pent-up demand would explode when companies decided to buy again. He wanted IBM to be ready to take advantage of that demand.

He’d keep the factories building machines and parts, stockpiling the products in warehouses. In fact, between 1929 and 1932, he increased IBM’s production capacity by one-third.

Watson’s greatest risk was running out of time. If IBM’s revenue dropped off or flattened because of the Depression, the company would still have enough money to keep operating for two years, maybe three. If IBM’s revenue continued to falter past 1933, the burden of running the factories and inventory would threaten IBM’s financial stability.

[…]

Watson’s logic led him to make what looked to outsiders like another insane wager. On January 12, 1932, Watson announced that IBM would spend $1 million—nearly 6 percent of its total annual revenue— to build one of the first corporate research labs. The colonial-style brick structure in Endicott would house all of IBM’s inventors and engineers. Watson played up the symbolism for all it was worth. He would create instead of destroy, despite the economic plague.

(Source: The Maverick and his Machine)

Most companies pulled back, and for good reason. Demand was rapidly shrinking, and IBM’s decision to spend money expanding productive capacity, research, and employment would be suicide if demand didn’t return soon. All of that unused capacity was costly and would go to waste. Watson took an enormous risk, but he also had faith that the American economy would recover its dynamism. If it did, IBM would come out on the other side untouchable.

Somehow, Watson had to stimulate demand. He had to come up with products that companies couldn’t resist, whatever the economic conditions. Again, thanks to Charles Kettering’s influence, Watson believed that R&D would drive sales. (ed: Kettering was chief engineer at General Motors.) So Watson decided to build a lab, pull engineers together, and get them charged up to push the technology forward.

Throughout the 1930s, IBM cranked out new products and innovation, finally getting its technology ahead of Remington Rand or any other potential competitors.

[…]

Within a few years, Watson’s gamble of manufacturing looked disastrous. As IBM pumped increasing amounts of money into operations and growth, revenue from 1929 to 1934 stalled, wavering between $17 million and $19 million a year. IBM edged toward insolvency. In 1932, IBM’s stock price fell to 1921 levels and stayed there—11 years of gains wiped out.

(Source: The Maverick and his Machine)

By 1935, IBM was still stagnating. Watson made the smart move to get out of the money-losing scale business and use the money to keep the remaining businesses afloat, but he was drowning in excess capacity, inventions be damned.

Then IBM got a stroke of luck that it would ride for almost 50 years.

After all of his pushing and all of his investment, after the impossible decision to push IBM to the brink, Tom Watson was rewarded with The Social Security Act of 1935, part of FDR’s New Deal. It was perfect.

No single flourish of a pen had ever created such a gigantic information processing problem. The act established Social Security in America—a national insurance system that required workers to pay into a fund while employed so they could draw payments out of it once they retired, or if a wage-earning spouse died. To make the system work, every business had to track every employee’s hours, wages, and the amount that must be paid to Social Security. The business then had to put those figures in a form that could be reported to the federal government. Then the government had to process all those millions of reports, track the money, and send checks to those who should get them.

Overnight, demand for accounting machines soared. Every business that had them needed them more. An officer for the store chain Woolworth told IBM that keeping records for Social Security was going to cost the company $250,000 a year. Businesses that didn’t have the machines wanted them. The government needed them by the boatload.

Only one company could meet the demand: IBM. It had warehouses full of machines and parts and accessories, and it could immediately make more because its factories were running, finely tuned, and fully staffed. Moreover, IBM had been funding research and introducing new products, so it had better, faster, more reliable machines than Remington Rand or any other company. IBM won the contract to do all of the New Deal’s accounting—the biggest project to date to automate the government…

This period of time became IBM’s slingshot. Revenue jumped from $19 million in 1934 to $21 million in 1935. From there it kept going up: $25 million in 1936, $31 million in 1937. It would climb unabated for the next 45 years. From that moment until the 1980s, IBM would utterly dominate the data processing industry—a record of leadership that was unmatched by any industrial company in history.

(Source: The Maverick and his Machine)

By combining aggressive opportunism and a great deal of luck, IBM was forged in the depths of the Great Depression. Like John D. Rockefeller before him, who bought up refineries during periods of depression in the oil industry, and Warren Buffett after him, who scooped up loads of cheap stocks when the stock market was crumbling in the 1970s, Watson decided that pushing ahead was the only way out.

History certainly didn’t have to go his way — FDR might not have been elected or might not have been able to enact Social Security. Even if he’d done it two years later, IBM still might never have made it.

But Watson’s courage and leadership did open the possibility of serendipitous fortune for IBM if the world didn’t end. Like oxygen combining with fuel to create internal combustion, those elements forged a monstrous competitive advantage when the match was finally lit.

Still Interested? Check out the excellent The Maverick and his Machine by Kevin Maney, where the excerpts above come from.