- Image by Extra Medium via Flickr
I approximated that each of us had likely already met the second-best girl we would see in our entire young-adult lives.
This was a claim made one night in college by my friend Bryan Hernandez. “Is this true?”, we said. For some reason, there was something depressing about this (for many of us.) I think largely because it meant, to some degree, our future was already decided for us.
Decisions are tough–especially when you have to choose an option, hope that it’s right, and give up surveying all the other options. And choosing who to marry is arguably one of the most important decisions we’ll make in our lives. So naturally, one should come up with a mathematical strategy to make the best decision! (well maybe…..)
So why did Bryan make this claim? Here’s what he says in his post on the topic, Modeling the Marriage Decision:
I’ll agree it’s a pretty bold claim, but I had reason. That week in my Probability and Random Variables class, one of my all-time favorites, we had been studying an area of probability known as Sequential Decision Problems. This class of problems concerns itself with applying classical probabi
lity theory to questions regarding how to make probabilistically optimal decisions in a sequence of random events.
Here’s the problem Bryan had studied that is analagous if you make some assumptions:
“You are in need of a secretary and only have time to interview N. You are able to rank the relative quality of each after interviewing her. At the end of the interview you must either reject the applicant without the ability of accepting him later, or accept him without the ability to continue interviewing the remaining applicants. What is the strategy that maximizes the probability of selecting the best candidate out of all N?”
Image via Wikipedia
Of course, this isn’t completely how it works. But go with it.
To make the optimal decision, there is an algorithm you can follow to get the best possible algorithm. First, estimate how many potential partners you will potentially meet (“N”) during your decision period and apply the following algorithm:
“Interview the first N/e applicants and reject them all. Select the next applicant that is better than the best out of that first group.”
For a more in depth coverage of this issue, check out Bryan’s full post on the topic: Modeling the Marriage Decision.
Nerds are funny.
-Kevin
8.13.2010
