Stop Forecasting And Start Dreaming



By @autobusiness

Forecasts are rational, useful, and necessary. Right?

Whether it’s 2013 US car sales forecasts made by Edmunds, Audi’s ambitious sales forecasts, population forecasts made by the UN, or inflation forecasts made by the Federal Reserve – everyone’s in on the forecasting game. But not all forecasts are created equal.

Listening to the weatherman for a week and comparing his predictions with climatological outcomes would show you how tough the forecasting game really is. And weather systems are far simpler than most of the systems we try to make date-specific predictions about. Is it really any wonder that everyone from Audi to the United Nations revises their forecasts until prediction and predicted merge seamlessly, like an outfielder catching a fly ball?

Forecasting of complex systems is nearly, if not completely, impossible. So why do we keep believing them?

We believe such forecasts for four primary reasons. 1) We believe they’re possible. 2) “Forecasts are better than nothing.” 3) They’re very hard to ignore.  4) It makes for a good tweet.

Let’s take a closer look at each of these.

1) The “Forecasts Are Possible” Argument:

Most of us forecast in our personal lives. We’re expected to. We’ve been asked “What do you want to do when you grow up?” since we were knee high to a grasshopper. When we entered our twenties and thirties, the question became “When are you getting married?” and so on and so forth.

Popular fodder for this argument are the scores of self-improvement books that encourage you to develop a 5-year plan. Do you recall what your life was like in January 2008? The chances that you would’ve accurately predicted your current situation is awfully close to zero, and this example was chosen for its relative simplicity! Imagine trying to pull off that trick for something far more complex with many more variables, such as a business, industry or global economy.

economist_slow_downWe’ve been trained from an early age to forecast and we believe in it intuitively, but this ignores that we can scarcely predict the outcomes of basic systems like weather and elections. So what makes us think that we can guess, much less deduce from piddling equations relying on trivial data, how many cars will be sold in the coming year or what the price of gold will be in 12 months from now? It’s a fool’s errand, yet we base our investments, our mortgages, our car loans, and our government pensions in the hands of the highly deluded and equally influential people who play this very game: that of date-dependent prediction. These influential people are easy to spot because they often call themselves “economists” to increase their credibility.

But “economists” aren’t the only ones who forecast, we all do it.

2) The “Forecasts Are Better Than Nothing” Argument:


Secondly, there’s the “forecasts are better than nothing” argument. Good evidence to the contrary is the case of US banks in the mid-2000s. They bet our money on the assumption that house prices in the US would rise indefinitely. They even used fancy models like the Black-Scholes to legitimize their assumptions. Then *BOOM*. “Our models never predicted this” was often heard in the aftermath. Their modeled forecasts not only hid risks and fooled investors, but they were untouchable scapegoats when the sh*t hits the fan.

Doing “nothing”, often seen as the alternative to forecasting, could only be worse if forecasting had neither negative nor unintended consequences. Since it frequently has both when used in complex systems (see Stalin, “economists”, etc), then not forecasting must be preferable. And as we’ll soon see, not forecasting isn’t the only alternative to forecasting.

3) The “Forecasts Are Very Hard To Ignore” Argument:

Thirdly, also known as the “Everyone Is Doing It, Even The Weatherman” argument, we address the argument that forecasts are very difficult to ignore. Between daily newspapers, TV, social media, cell phones, radio ads in the car, and billboards on the road, we rarely have a moment to ourselves that isn’t externally influenced. There’s always a stimulus for us to react to and forecasts are no exception. As Alain de Botton put it, this inundation of information makes us feel that “at any point, somewhere on the globe, something may occur to sweep away old certainties.” This is certainly the case with forecasts that prognosticate global doom and ingenuity-powered salvation in the same breath.

But many things are hard to ignore, like someone else’s crying children or the panhandler on the street, and we seem to manage just fine.

4) The “Forecasts Make For A Good Tweet” Argument:

Western Culture, in the Islamic tradition, favours using numbers to “prove” things. We don’t want fluffy ideas, we want data!, preferably in a graph or pie chart presentation. This inclination is also ideally suited to our 140-character thoughts, or what are often called “tweets”. As such, numbers go viral very, very easily. Numbers, particularly when translated graphically, tell a story. Unfortunately, graphs and data are far too easily manipulated. Facts end up conforming to theories, rather than vice versa.

Y’know what else makes for good tweet? What you ate for lunch. Since no one ever bet the family farm on the number of patties in your burger, you should probably stick with the mundane. Everyone else is doing it.


1. Ignore Forecasts

Don’t confuse data, graphs, and forecasts for scientific absolutes – they’re not the same thing. If you’re skeptical of the weatherman, apply that skepticism to all date-specific forecasts, particularly when the systems are highly complex.

Also, don’t think that you’re better at forecasting than anyone else. You suck at it too.

2. Seek Balance

Data and graphs can be very useful in some domains, like determining vehicle crash safety or aerodynamics, but it can also be very harmful in other domains, like predicting EV adoption rates or predicting when you’re going to get married (particularly if you’re single) only to become disheartened when the “deadline” comes and goes.

But don’t let that prevent you from learning about your world and searching for love. As the 13th century Buddhist Zen teacher Dōgen wrote:

Knowledge is illusion; not knowing is ignorance.

US electric car sales forecast

3. Replace Forecasts With Bucket Lists

Before we can achieve our goals, first we must survive (obvious, but frequently skimmed over). Within the framework of our lifespan, we’re free to achieve whatever we can squeeze in. And just as we can’t know when our last breath will leave us, we can’t know when our goals will ultimately be accomplished.

Ditch the 5-year plan. Make a bucket list. And if your bucket list includes “Become the President of the United States”, don’t quit just because you’re 30 and not there yet.

Dream big. And go after those dreams every single day.

“Stay hungry, stay foolish” and stop forecasting.



[Image credits: Burkley Duffield, The Big Picture, author, UC Berkley via Altzar]