How Can We Make Our Predictions Foolproof?

Nowadays, the pace at which life changes seems faster than ever. Trends come and go, technology changes (almost daily), and don't even get me started on politics!

Sometimes it feels as though you need to be able to predict the future just to keep up with all these new changes. You could go on your gut instinct to guide you, but really this is just guesswork, and you're beginning to think that some of the wisdom you've gained through experience no longer applies to this brave new world.

So who can you rely on for accurate predictions on what the future may bring?

Expert Predictions Versus Laymen Predictions

In 1907, Charles Darwin's cousin, statistician Sir Francis Galton, made an interesting discovery at a county fair. Visitors were asked to guess the weight of an ox. Very few guesses were close, even among cattle experts. However, the average of all the wrong guesses was right on the mark!

A 20-year study at the Wharton School asked both intelligence experts and everyday folks to estimate the likelihood of some global future event such as, "Will there be a significant attack on Israeli territory before May 10, 2014?" Against expectations, the everyday folk's answers turned out to be 30 percent more accurate than the experts'.

Many of you experienced leaders may be offended by my questioning your ability to predict future events. I don't question your knowledge and insights, but simply point out that they don't necessarily translate into accurate future-casting.

The Wharton study, for instance, found that experts spend much of their time with fellow specialists deep inside their field. And while great minds do tend to think alike, spending so much time with similar people can result in groupthink.

Often when problem solving, experts tend to focus on the "knowns" within their field, rather than the external forces at work that are changing it. Their confident knowledge leads to quick decisions. But, if and when they ultimately change their minds, it can be a slow process.

Be Flexible to Avoid Over-Confidence

Changing circumstances can result in cognitive dissonance. Human instinct works to eliminate shades of gray – we come to a decision, we move on.

The Wharton study allowed participants the flexibility to change their predictions of future events over a three-month period. Those with changeable attitudes updated their beliefs more often, and their predictions were found to be far more accurate than those that had fixed beliefs.

Daniel Kahneman, who won a Nobel Prize for his work in behavioral economics, agrees that over-confidence and cognitive bias can lead to errors. He found that much of human error cannot even be put down to a systematic cause. Instead, a lot of errors made by people are simply "… noise, in the sense that it's random, unpredictable [and] cannot be explained."

Individual experts can worryingly become confused and error-prone because of this over-confidence. As Kahneman explains, "You put the same x-ray in front of radiologists and, about 20 percent of the time, in some experiments they don’t reach the same diagnosis."

So, why not have a group of experts settle on a decision, you might ask? Well, there are pitfalls here as well. As he goes on to clarify, when a group of people discuss a case, there is often "huge conformity pressures," which can lead people to underestimate the level of disagreement among them.

Kahneman suggests using algorithms to "temper human judgment," because they are noise-free – something that humans are not. However, although this sounds to me like it could work well for short-term judgments, long-term predictions surely require more thought, and developing algorithms for them would be a challenge.

So... what's the answer?

Polling and the Wisdom of the Crowd

Well, telephone polling is one way to gauge predictions accurately and is used by many organizations to do just that.

Polling is intended to tap into the wisdom of the crowd. However, only a generation ago these polls were found to be far off the mark. This was primarily because phone books omitted cell phone users, which meant that polls ignored much of the younger population.

The science behind poll predictions has, thankfully, improved a lot since then. Polling expert Nate Silver, for instance, whose website FiveThirtyEight.com uses demographic analysis as well as a range of political polls (each of which is weighted on timing and past accuracy), has earned great acclaim in recent years for his ability to accurately predict the future. Although, even he gave Trump less than a 30 percent chance of winning!

If you do decide to rely on a poll, it's important to think about who you want to target. Try using the method that I borrowed from James Surowiecki's book, "The Wisdom of Crowds." Create your very own "wise crowd" by making sure that it fulfills his four criteria:

1.   Each person should have private information because even this diversity is just an eccentric interpretation of the known facts.
2.   Know that people's opinions are independent and not determined by those around them.
3.   Your sample group should be decentralized, with different specialties, and can draw upon local knowledge.
4.   Have a mechanism for aggregating your group's private judgments into a collective decision.

If these points feel too generic, find more practical help developing your own survey by visiting Survey Monkey.

Surveying a full range of people may not always possible (to say nothing of being impractical and time-consuming). If this is the case, the Delphi Method is an excellent way to gain expert advice on future predictions, while reducing the problems described above. Tips and explanations on the Delphi Method are available to Mind Tools Club members but, if you don't have club membership and want to learn more, leave a comment and I'll dive into it in a future blog.