The second book I’ve started this year is Algorithms to Live By: The Computer Science of Human Decisions by Brian Christian and Tom Griffiths. Here are the links for you to purchase it:
This is another book that I’m listening too while I walk Maggie. I only listen to 1 chapter per day and often repeat chapters to make sure the lessons sink in. This book goes through specific examples of human behavior that are best solved using specific algorithms. It’s interesting to learn how simple an algorithm is to apply, but also obvious how difficult it may be to disregard human inputs such as emotion.
Ideal Stopping Point
The lesson deals with the ideal stopping point in reference to ‘The Secretary’ problem. This algorithm is the answer to questions such as:
How long should I look for the ideal candidate before selection?
How do I achieve the greatest probability of selecting the best candidate?
Stop Looking at 37%
The first example in this section deals with apartment hunting in San Francisco. This area is known for its excessive demand and requires a renters ability write a check faster than the competition. Apartments are often only available for a single day before they are acquired. If you plan on hunting for apartments for a month, then you should hunt without your checkbook for 37% of the month (about 11 days).
After that point you should be ready to right a check as soon as you find an apartment better than all of those you have already seen. At first this sounds incredibly aggressive. It also sounds too easy to implement. The issue is fighting the urge that you may miss out on an even better apartment. You have to fight the impulse that what you don’t know is better than what you do.
This principle applies for hiring the ideal candidate as well. If you know the duration of your search you should treat it in a similar fashion as the apartment example. This algorithm also works when dealing with a set number of candidates. You should look at the first 37% without making any commitments. I had a REALLY tough time hearing this because I am in the process of applying to positions. For companies that I track on a regular basis, I may reach out and apply on the same day they post a new position. That means I’m in the 37%. As I’m learning more about this algorithm it makes me think I should wait 2 weeks before applying to any position. That keeps me out of the 37% assuming the company plans on hunting for around a month. This also assumes the company is treating the situation the way the algorithm would. Interesting little conundrum isn’t it?
Finding a Wife
The funny example deals with relationships. If you plan on ‘hunting’ for a partner between the ages of 18 and 40, then you should date freely until you are 26. At that point you should propose to the first woman (in my case) that meets all of your qualifications (terrible word, I know). You should also propose early and often. As soon as you know she meets your standards you should propose. Don’t dilly dally. Now that you’re past the age of 26 the odds of finding the best match are only getting smaller. I laugh because I’m 31 right now. I guess that means I better go buy a ring? Don’t worry. I’m too much of a romantic to follow this procedure.
This concept of treating human situations as algorithms is extremely interesting to me. We often clog up the system with emotions and human error. Now that I understand the stopping rule (I need to learn more) I will definitely try to apply it to more questions I face. I’m currently hunting for a new residence in Boulder, CO so that will be my first chance to test the concept.