Using math to choose a mate

http://news.cnet.com/8301-17852_3-10309716-71.html

by cwan • 17 years ago

21 6 17 years ago

news.cnet.com

smanek2• 17 years ago

They are talking about: http://en.wikipedia.org/wiki/Secretary_problem

http://en.wikipedia.org/wiki/Stable_marriage_problem is also interesting (Residency programs use it to match up students to hospitals)

Tichy• 17 years ago

Hm, I think the next time Google has a job opening, I'll let some time pass before I send in my application :-)

arohner• 17 years ago

There's a huge problem with this though. From http://en.wikipedia.org/wiki/Secretary_problem : "There are n applicants for the position, and the value of n is known."

That is categorically false when talking about marriage. How do you define "potential life partners"? Is that anyone of the appropriate sex? Anyone who meets your criteria (but potentially you don't meet theirs?).

Further, the number of people you will meet is a function both of time, and of the probability that you will meet a new person on any given day. Then there's the issue that your desirability is non-linear over time, based on age, experience, etc.

smanek2• 17 years ago

I don't think it would be unreasonable to assume I could seriously date 1-3 people a year (other people might go slower/faster, but that seems like about how long it takes for me to get to know someone in that context), with an average of 2. Assuming I start seriously dating at 18, and want to get married before 28, I've got 20 potential mates to evaluate.

So, the 'algorithm' is that I should date as many as I can (each for just enough time to be able to order them ) till I am ~22 (1/e*20/2 ~ 3.7 years), and then between 22-28 only date women long enough to compare her to all the women I dated from 18-22. If she is better, I should marry her. Otherwise, move on ASAP.

Of course, I've already 'failed' miserably at that plan (e.g., a 3 year relationship from 19-22), but that's what the math says you should do ;-)

dhimes• 17 years ago

This is simply an interesting mathematical problem, perhaps only a zeroth-order approximation to real life of course. In real life, what you learn "interviewing" the first n candidates changes the criteria for candidates n+1...N. That is, you are growing too (if you are doing it right).

Perhaps the characteristics you and your future mate require are actually innate, and s/he can be selected early? Possibly, but then you wouldn't know what you should be looking for, and arranged marriages are a better solution.

Oh, yeah, and s/he has to like you back!

chimariko• 17 years ago

I like the approach described in this article, which is, in fact, "Take some time to look around and then choose." Choose too soon -- you don't have a relevant criterion. Wait too long -- you might have already missed the best units.

I wonder if there is an insight into the actual model behind the resulting change_mode_number = total/euler_number formula?

Using math to choose a mate

Related Stories