Better decisions through blindness

If you knew you could be happier and make better decisions by blinding yourself, would you do it?

Ok, not complete blindness, but just to those factors you decided in advance were irrelevant.

In the science fiction short story “Liking what you see”, people can have a small, reversible operation which causes ‘calliagnosia’, or the inability to detect human beauty or ugliness. You can still recognize people, but the part of your brain responsible for discerning beauty, disgust or lust is disconnected.

Adopters go through life freed from constantly, involuntarily noticing how superficially attractive people are. Attraction, friendship and love still happen – but based on actions, not aesthetics. 

This idea – intentionally blind yourself to irrelevant information – has serious merit. There is convincing evidence that we are strongly influenced by factors that should be completely irrelevant to the decision we’re making. From the book’s afterword:

Psychologists once conducted an experiment where they repeatedly left a fake college application in an airport, supposedly forgotten by a traveler. The answers on the application were always the same, but each time they included a different photo of the fictitious applicant. It turned out people were more likely to mail in the application if the applicant was attractive. This is perhaps not surprising, but it illustrates just how thoroughly we’re influenced by appearances; we favor attractive people even in a situation where we’ll never meet them.

Here are a few more:

  • The acceptance rates for female authored papers at Behavioral Ecology went up by 8% after they implemented a double blind review process.
  • Adding in a blinding screen during auditions increased the percentage of women in orchestras by between 25% and 46%.
  • Including information about how much peers were savings on a retirement plan document decreased savings rates.

Attractiveness, gender, or other irrelevant information can be more than a harmless distraction – it may be actively leading us to make worse decisions. I see two pathways for this effect:

  1. We have to spend brain power and time processing useless information. This crowds out useful information, or analysis we could have done.
  2. The irrelevant information may unconsciously bias us as we evaluate the wrong question (‘is she attractive’ vs ‘is she a good manager’).  

Which brings us to this: when irrelevant information ‘on’ by default and harmful, it may be worth actively blinding yourself to it.

Please don’t mis-understand: I think blinding is smart for purely selfish reasons. I’m not encouraging it to make the world more fair or just (though that’s not a bad side effect). If I want to make great decisions, hire the best fit people, etc I need to set up the best process. If that means removing distracting or misleading information, I want to do it.

Decide what to consider (and avoid considering) ahead of time

There are many ‘defaults’, set by nature, for what information we perceive and pay attention to. They’re usually not bad, but we also usually accept them uncritically. 

An anecdote:

We posted an opening for a quantitative portfolio analyst, and received hundreds of CVs. A brief reflection made it clear we were spending lots of time reviewing these CVs haphazardly, and they had a lot of noisy signal. So what exactly were we looking for?

  1. Someone who could put their ideas into high quality code quickly
  2. Someone who grok’d investing & portfolios both practically and theoretically

We modified the application so that in order to submit, applicants had to write a very short code snippet, and take a view on a portfolio optimization issue.

The results were impressive: the number of completed applications went down by about 90%, and the quality of the candidates went up. There was less subjectivity about the quality, because we were judging an ecologically valid example of their work, not who they were. We were spending less of our time on each application, but focused on the highest quality signals.

Less overall information, less time spent reviewing it, and higher quality results.

A distraction-less personal finance dashboard

Suppose you could design the interface for your own investing software. The interface where you will make all your investing decisions. You’ve got limited space, and you’ve got to decide what you always see when you log in. What would you put on that first landing page?

For mine, I’d intentionally not show most of what shows up on DIY brokerage sites today:

  • The shares/price/value of each position I hold
  • Whether those positions are in a gain or a loss
  • The historical performance
  • Market news related to my holdings

These aren’t just useless for making forward-looking decisions – they’re actively harmful. 

And I know what I would want: 

  • My remaining safe-to-spend amount in my checking account
  • For each of my future liabilities (goals)
    • whether or not I’m on track
    • a prospective shortfall if I’m off track
    • How much an additional $100 invested today would help me towards my goal
  • My risk relative to the target level
  • My portfolio drift, and how much I’d need to deposit to rebalance tax free
  • Upcoming events I need to plan for (e.g. taxes, end-of-year stuff)
  • Potential liabilities (goals) I might not have thought of, but are common for people like me.

Such an interface probably wouldn’t be exciting. I might not be motivated to log-in as often. But it would empower me to be a better investor.

Some ‘yeah buts’

There are, of course, valid objections to intentional blinding:

  • Beauty is pleasurable.
  • Knowing you’re ‘up’ feels good.
  • Sometimes, we don’t know what information is irrelevant till after we’ve seen it.

All true. We all have to balance small pleasures against a virtuous long-term life. 

But we could probably all benefit from being more intentional in separating where we go for fun, or beauty, or profits. 

2 thoughts on “Better decisions through blindness

  1. Hi Dan,

    Would you mind sharing what language the code snipit was in for the application question? I’m going to start learning to code and I am in process of figuring out which language to learn.

    Thank you!

    1. Hi Casey,

      We didn’t demand any specific language, but it really helps if it’s in R, Julia, Python or some other open-source language. This is just so that licenses aren’t a barrier to running the code.

      You can read my guidance here: http://www.dpegan.com/652-2/

Leave a Reply

Your email address will not be published.