Selection Bias and How it Can Be Used to Gain Wider Choices

I've read Brian Greene's book The Hidden Reality, which is a fascinating read on the current work of physicists and mathematicians in discovering string theory in an attempt to prove that the world is in fact part of the multiverse which contains many galaxies, planets, and stars just like ours.

Note: This article is based on the work of the author and founder of the website "riskology", Tyler Tervooren, in which he tells us about his experience in changing his life.

They test every possible intuition in order to fit their theory into the paradigm of science as we understand it today. This is how new scientific theories and laws are accepted by what is currently understood and approved.

However, the problem that repeatedly comes up is that of selection bias, which is about tests that seem to work, but in reality, it is not since the samples tested were not diverse enough. When they were dissected again with a larger set of data and subjects, some of their new discoveries collapsed.

Selection bias disrupts science and objective study, but can it really be beneficial to people who are trying to change their lives for the better by ignoring the odds and facing unreasonable challenges?

My hypothesis says: "Yes", it can be used usefully, but it all depends on how you apply it. So, let's look at these two examples:

First: Doomed business

Think about this scenario for a moment:

You have a business idea and it is really expensive to start, or you may use up all your savings, and you are not sure if people will actually like what you are selling. Thus, you decide to take some first sensible steps and ask 20 of your closest friends if they will buy your product when it launches. So, you turn to your mom, dad, grandparents, aunts, uncles, and all your cousins and best friends.

Every one of them says, "Yeah! I'll definitely buy it. You've got to start your business." However, surely there will be someone out there who thinks the internet is a fad.

Based on your findings, you will be thrilled and 95% confident that you have a very valuable idea. You go to the bank, withdraw all your money, and spend it lavishly on developing your product. Let's say your product is a new anti-hair loss formula for big cats, and no one in your family has a pet.

You release the product, sell exactly 19 tubes to friends and family who said they'd buy it because they felt guilty, and give one to the person who is not convinced about your work, just to prove them wrong.

You wasted all your savings on selection bias and got a fake result, because you asked the wrong questions to the wrong people. Of course, this is a strange scenario and you are smart enough to avoid such a disaster, but it does explain the concept well.

Second: A talented violinist

Think about this scenario:

You want to be a great violinist, but you've never held a violin and that scares you. Also, you are not sure if you'd be able to do that.

However,  you judge by the fact that you don't actually know anyone who plays the violin and you saw a five-year-old on YouTube taking part in a lighting ceremony Olympic torch, so you think it can't be that hard, and it doesn't seem like much of a challenge, and you start learning it.

As you get deeper into it and get to know the instrument, the music, and the culture, you start to realize how naive you were when you started, but it won't matter anymore.

You are now very good at this new skill, you can have musical performances whenever you want, you have just had an audition with a traveling orchestra, you have reached the top, and your self-esteem is at the highest level.

You have learned a completely new skill that you have always wanted to learn, and your work has been very hard, but you did it anyway because you loved it.

If you had known the scale of the challenge you would face when you started, would you have taken the first step? Would you even allow yourself to dream about it? How would you react?

In this case, your selection bias has helped you embark on an insurmountable challenge. Sometimes, you need to do thorough research before putting yourself in harm's way.

Other times, it's better not to know the possibilities in the first place because it is not important even if you have never carried a violin or used it before, as long as you learned something new that made you happy, and in my opinion this is a success.

The result

Selection bias can make or break you, but it all depends on what your goal is. If you're trying to start a business you're not sure about with all your money, or you want to find a way to convince the world that the universe is actually very different from what we currently understand, selection bias is your enemy and you must confront it.

On the other hand, if your goals are internal, personal, and you just need to convince yourself to start something important to you, then by all means ignore statements that say you don't have good chances, because it doesn't matter. Enjoy your selective ignoring.

Additional case study

When I started Riskology, I used the same strategy to create a website design even though I didn't realize I was doing it at the time.

Although deep down I knew paying a web designer was worth the money, because what they do is hard work that no amateur can do, I kept telling myself, "I can do it."

I was incredibly naive, but the truth is, it wasn't really important. The only hard cost I had to design the site myself was $65 and a lot of time - and I had a lot of it - spouting horrible phrases while sitting infront of my computer.

I did not know how much work I would need or what I should learn, but I set about it anyway, and met with many failures even though I pressed on. I really had nothing to lose in this absurdity, rather I was learning many things.

Instead of worrying about looking and searching every well-designed website online - which is an impossible task - I picked five great-designed websites, contacted them, and spent a lot of time trying to understand why they looked so good, and I wanted to know what caught my eye about design as deemed effective.

Then I took what I learned from those designs and applied it to work on my own, borrowing ideas liberally and arranging them in a way I thought made sense for the audience I wanted to attract, and then I got into fierce competition.

The end result was a design that was too weak to appeal to 99% of the people online, but it is very good at getting the attention of the people I want to come to this site.

It is too bad for a large scale site that's trying to market their content at scale, but for the people I want to talk to, it's pretty good, and I think that's been a big factor in the growth I've seen at Riskology.

Some lessons to make this kind of experiment works better

1. Avoid imitation

Freely draw inspiration from others, but don't steal. The direct imitation approach always leads to limited success, and you cannot go beyond the original because you are a phony.

So, find out what inspires you and apply it instead. Benefit from the core of the idea, not from the model itself, as this opens room for unlimited success.

2. Study the reason for success

It is easy to find friends, relatives, and other close people who have an opinion on how something should be done, but unless they have already done it themselves it is not very helpful.

Taking advice from someone who has not achieved what you are trying to do is usually a fast road to failure, but it is tempting because it is easy to get advice. So, go the extra mile and find good case studies that you can actually learn something valuable from.