How Linear Models Are Designed to Trick You Into Making Bad Decisions
Oversimplifications of complex topics are used to manipulate and deceive people
Whether you recognize it or not, you’ve been trained to think in a way that makes you easy to control. That training severely inhibits your ability to make an impact on the world.
We often use models to create a simplified representation of reality that is easy to understand. However, our world has become so complex that the basic models humanity has relied on for generations are on the verge of becoming irrelevant.
One of the most basic models is a simple straight line. In this article I will show you how easy it is to capitalize on the limitations of linear thinking in order to trick a person into making terrible choices.
The human race has reached a point in our evolution where linear models are no longer capable of representing the complexity of the decisions we regularly face.
I believe our overreliance on linear models is one of the reasons our society is so divided. We are easily tricked by oversimplifications that bear little resemblance to the realities of the world. It’s imperative that we come to understand that our models have to become more complex if they are to provide us with actionable information.
Why do we make bad decisions?
When I launched a business back in 2011, it became immediately apparent how easy it is to be seduced by colossally bad strategies. I was stunned by how often awful choices initially presented themselves as the most appealing.
I realized I needed a mechanism to help guide my decision-making. I’ve since discovered that the model I’ve adopted is applicable in every facet of my life.
Sometimes you do the wrong thing because it’s “safe,” or it’s “expected.” Sometimes the strategy that will bring the biggest gains in the future, open you up to ridicule in the present.
If you are going to succeed, you need to learn how to tune out baseless criticisms and be confident in your decisions.
Eventually, I realized that part of the reason I was susceptible to making poor decisions is because society trains us to have an excessive and unwarranted confidence in linear models.
Models make decisions easier, but you must be careful…
Whenever you are dealing with a complex problem, it’s useful to create a simplified model.
Some models are more useful than others. For example, here’s an example of a zero-dimensional model:
It’s a point. That doesn’t help you very much, but we’re just getting started.
Models can be very useful. However, it can be easy to overlook the fact that models are representations and not the “thing” itself. Again, this is basic, but when you start trusting models, it can be easy to forget that they are, by definition, incomplete. In a little while, I’ll show you how people make unfounded assumptions all the time.
A related concept is that of significant figures. When you type 7.26 X 3.87 into a calculator, it spits out 28.0962. Your calculator doesn’t magically create accuracy up to four decimal places. But, we trust our calculators and sometimes we jot those numbers down without even thinking about it.
It can be tempting to develop a sense of confidence in a model that has historically provided a good return. However, people get into trouble when they start relying on the model to the extent that they disregard data points that contradict their oversimplified representation.
The more complex the model, the more difficult it is to use, but the more accurately that model will represent your circumstances.
The limitations of linear thinking
Plotting data points along a straight line has effectively lifted the human race out of the stone age. However, this model has severe limitations, and if we are going to take the next step in our evolution, we need to add greater complexity to our decision-making.
Furthermore, you need to recognize how oversimplifications can be used to deceive you.
It’s really easy to manipulate people by taking advantage of their confidence in oversimplified models.
We’re surrounded by this, and I’ll provide some examples in a moment.
Simple models are easy to comprehend, but they lack accuracy. It’s dangerous and irresponsible for a complex society to use simple models as the basis of its decision-making. But we do it all the time:
Linear representation of good versus evil
Linear representation of liberal versus conservative
Linear representation of profit versus loss
Linear representation of male versus female
Linear representation of moral versus immoral
I would contend that any time you’re evaluating anything in the context of a straight line, you’re prone to making unwarranted assumptions. This can come in the form of predictions that are implied by the model, but which will not be observed in real life.
Which CEO is the best?
Let’s start with a basic example.
Imagine you have a straight line. It’s a typical X axis. There’s a zero in the middle, positive numbers to the right, and negative numbers to the left. This is a one dimensional representation (and that should be enough to tell you something).
Let’s say the line represents corporate profit by percentage. This is one of the most basic linear representations. Profit equals good. Loss equals bad.
Now, let’s put a CEO’s performance on the line. Let’s say the year end review shows that the business lost 20% of its value (bad, bad, bad).
What can we conclude?
A. The CEO did an awful job.
B. You can’t conclude anything.
Tick-tock, tick-tock
Got an answer yet?
The answer is B.
Although you might be tempted to assume a 20 percent loss indicates the CEO did a bad job, you really don’t have enough information to be certain. If you make the decision to fire the CEO without further evaluation, you run a high risk of making a huge mistake.
Let’s add a few more data points
Even when we add a few more data points, this is still a linear representation. We really won’t get anywhere until we add another dimension (but that will come later).
Let’s add in the data points of our competitors.
As you can see, the competitors are all clustered in the range of -70% to -100% compared to the previous year.
Oh, all of a sudden that CEO many of you wanted to fire a moment ago is starting to look pretty good.
So, let’s ask the question again, what can we conclude from this model?
A. The CEO did a good job.
B. You can’t conclude anything.
Tick-tock, tick-tock
Got an answer yet?
That’s right! The answer is STILL B! If we want to do a good job and make a good decision, we can’t have a knee-jerk reaction to everything. It’s time to roll up our sleeves and get to work!
Nothing is simple. When politicians and media personalities claim things are simple, they’re lying in order to trick you in almost all instances.
Now let’s add another dimension
As your model gets more complex, it becomes more difficult to interpret, but your interpretations will be more actionable.
Simply adding one more dimension, in this case a Y axis, makes your model exponentially more complex. The point I want to hammer home is that, in the modern world, you need to ALWAYS base your decision making on two dimensional models.
Always, always, always, always, always!
The difference between one dimension and two dimensions is the difference between aspiring to fly and walking on the moon. It’s that significant (I’m hammering on this point because as you sit there reading this, you have no awareness of how much of your thinking is based on flawed, one-dimensional concepts).
Consider the global economy
Getting back to our CEO example, let’s make the Y axis signify “economic growth in country of origin.” This allows us to compare how the CEO’s business did compared to her competitors, and compared to the economic situation of the country where the business is based.
Let’s say we get this:
From this graph we can see that our CEO managed to guide our company to a minimal loss even though she faced the greatest economic downturn.
The yellow and green CEOs had huge losses despite favorable economic conditions.
The blue and orange CEOs had significant losses, but also faced tough economic conditions.
Again, this is a basic example, but see how quickly it became complicated? If you want to make good decisions, your process needs to follow this path.
Oversimplifications are not your friend
This model is now complex enough that we can make a compelling argument to defend the performance of a CEO during a year when her company suffered a significant drop in profits.
The single point that shows a loss of 20% makes us interpret her performance as bad.
The comparison to competitors that showed catastrophic losses makes us interpret her performance as acceptable.
The context of national economy shows she managed only a small corporate loss during a massive recession. This allows us to interpret her performance as phenomenal.
But it’s important to note that each of these interpretations becomes progressively more complicated to defend.
A person who stands up in a room and says, “Our CEO led us to a 20% loss! She’s TERRIBLE!” is going to be very convincing.
The person who says that is going to be even MORE convincing if he’s speaking to a crowd of people who have been conditioned to put unreasonable faith in linear models.
People get fooled all the time
All you have to do is look to the news to see a dozen examples of how oversimplified models are deliberately deployed to trick people into making bad decisions.
This applies to your business. This applies to the decisions you make as a consumer. This applies to your romantic choices. This applies to every aspect of your life.
We aren’t in the stone age anymore. You can’t just draw a line between two points and presume you’ve covered every possible scenario.
Your life decisions are too important to give undue credibility to one dimensional models… but we do it all the time!
Yes/no, good/evil, right/left
Our whole society is built on flawed models designed to confuse us.
Yesterday I watched a heated exchange between Ted Cruz and Merrick Garland. Cruz asked Garland a loaded question and Garland insisted on providing an elaborate answer rather than a simple “yes” or “no.”
Think about that for a moment.
Consider how commonplace it is for lawyers to demand “yes/no” answers in court. That’s a linear model. In the complex modern world, virtually NOTHING is accurately represented by “yes” or “no.”
Why do we allow for such loaded questions when such important decisions rest in the balance? It’s because we’re conditioned to believe a straight line, linear model is an acceptable way to think about complex concepts.
IT ISN’T!
Our whole society needs to develop to the point where everyone understands one dimensional models don’t provide us with actionable information. Lawyers already understand it, that’s why Garland refused to give a response.
The political spectrum is also a flawed model
People think nothing of thinking of our political spectrum as a straight line labeled left, center, and right. The problem comes when people make absurd assumptions based on that oversimplified representation.
The trouble is, it is really, really complicated to explain our true political landscape with any accuracy. I recently tried to do it here.
The “left/right” linear model leads to unfair assumptions.
One example of this is how any policy that lands away from the center point of the political spectrum is labeled as “extreme” by default.
Is that a fair thing to say?
This is a case of the model artificially dictating how we interpret real world concepts. Labeling certain policies as “extreme” is a consequence of the oversimplification and not the policies themselves. This is like a value judgement that’s based on meaningless significant figures.
Is it really “extreme” to suggest children should have access to affordable healthcare? That isn’t an assessment that’s widely supported on the global scale. However, many Americans view affordable healthcare as an “extreme” position because that’s where it happens to land on our absurd model.
The concept of affordable healthcare isn’t flawed, our linear political model is.
Don’t draw real world conclusions based on oversimplified models.
Society has evolved beyond linear models
I believe that the human race is at a transition point. The modern era is exponentially more complex than it was only a few decades ago. However, linear models still dictate much of our thinking.
The problem is that a linear model can be easily manipulated to fabricate evidence that sends you in the wrong direction.
I believe this is why we have the phrase, “Trust your gut.” It’s already part of the zeitgeist that bold individuals know they must sometimes do the exact opposite of whatever “conventional wisdom” dictates.
We already know to be skeptical of model based evidence. What we haven’t figured out is how to remove all the flawed models from positions of influence.
Some decisions are inconsequential and don’t require a complex model. However, any time you are making an important decision, you need to take a moment to determine whether the model you are basing your decision on is oversimplified.
Unless you deliberately think about these concepts, you will continue to make bad decisions based on flawed models.
I’ve come to believe that misplaced faith in oversimplified representations of complex concepts is one of the main things holding back the progress of the human race. The next time you are confronted with a big decision, either in business or in life, take a moment to step back from your model and see how it looks when you add in another dimension.
It takes longer to interpret the data on a two dimensional model, but you increase your odds of discovering the most profitable course of action.
Perhaps a thousand years from now, we’ll add a 3rd dimension, and in a hundred thousand years, we’ll consider a 4th. For now, going from one dimension to two is more than most people can handle.
This publication is reader sponsored. If you have the means, please consider sponsoring at whatever level is comfortable for you!
My CoSchedule referral link
Here’s my referral link to my preferred headline analyzer tool. If you sign up through this, it’s another way to support this newsletter (thank you).
Excellent teaching essay. This piece is science, this tidbit, is something most people wouldn't consider unless they learned it recently, or recently reminded that graphs and their accuracy levels, often make a big difference. All we have to do is begin to read them and interpret them. We can learn this! 👏
This is why I (internally) scream and tear my hair when someone says "Teaching algebra is useless, no one will ever use it." and here you are applying the systems of basic algebra to your life in a wonderfully interesting way!