Thursday, June 30, 2005

The Scarlet Algorithm

In writing a short-story comedy about people sticking messages in library books and the like, I created a book about an algorithm or procedure for finding out who your best romantic partner is. I called this fictional book "The Scarlet Algorithm". The idea is that such an algorithm would help people find conjugal partners. This idea is something that has been around a long time. In 1966, my college was introduced to something called Operation Match. The idea is that people going to a certain dance would fill out a questionnaire about themselves and who they want as a partner, and the computer would find for each such person a list of persons of the opposite sex that are the most compatible with them. This did not always turn out right, partly because people put down on paper who they think they want rather than who they want. Nowadays this practice continues in the plethora of dating services on the Internet. To me this is not the best way to meet someone. The best way is through joining clubs or organizations that represent your interests and finding people there.

But if one were going to organize an Operation Match dance, what would be the algorithm? Probably some variant of the Hungarian algorithm, which matches elements of one set with elements of the other to optimize total worth, or of the Gale-Shapley algorithm, which eliminates the chance that "affairs" will form; i.e., two people who prefer each other to the people that they have been assigned to. Recently I used the Hungarian method to pair major league baseball teams with their AAA affiliates, and found that major league baseball could cut major-league-team-AAA-affiliate costs by 40%.

Other mathematical schemes describe love itself. John Alan Lee describes something called the "Colors of Love", in which he describes love as a two-dimensional continuum, specifically the triangle described by x + y + z = 1, with x, y, z >= 0. The three variables in this are "primary colors", which Mr. Lee describes as eros, ludus, and storge. He then describes three secondary colors, with mania between eros and ludus, pragma between ludus and storge, and agape between eros and storge. He then describes tertiary colors, but only for combinations involving mania, including manic storge, which would be a mix of a secondary color with its opposite complement. I therefore conclude that Mr. Lee is really dealing with a four-dimensional model, with four primary colors of eros, ludus, storge, and mania, however, it is hard to picture colors in four dimensions.

Then there are rating schemes. One is the familiar "10" or "perfect 10". This envisions that a man, for example, will rate women on a scale from 1-10, and that what he really wants is a perfect 10. There is even a movie with the name "10", which describes such a desirable woman. First of all, it should be a scale of 0-10. This is the most natural way of rating a scale on two extremes, with 0 at one end and 1 at the other; just multiply this scale by 10.

But another problem is that one will find that there are too many 10s out there. The scale suggests that each number from 0 to 10 will have the same number of people in them; since we meet hundreds of people there will be tens of 10s out there. A better idea is to use a logarithmic scale; see Logarithms Keep Dr. Brown in Perspective". In this scale, a 1 is one out of 10, and so is the counterpart of the "perfect 10". A 2 is one in a hundred, a 3 is one out of a thousand, out of 1,000 with 3 zeroes after the 1. A 6 would have six zeroes after the one, and would be one in a million.

So even though the two subjects of mathematics and romance don't seem to have much in common, there are ways of applying the mathematics to help out with finding a suitable partner. These ways I call "The Scarlet Algorithm".

2 comments:

Wolfy said...

In 1969, in Tampa, FL, I was attending USF and filled out an Operation Match questionaire. I met my future wife, and we married that year. We have been together ever since. Thanks to such an innovative program.

Jim said...

I also used Operation Match, in 1966-1967 and dated about five "ideal mates". But it did not produce any results. I eventually met my future wife in 1967 at a singles club. Operation Match was innovative in its time, but I thought the methods it used were superficial - answering arbitrary questions and trying to match up interests. But glad to see it worked in your case.