We've paused intake of new clients for 2021. Please check back in 2022. Thank you.

Assortative matching

Published by Priyanka on

Be it math or match-making, I have ideas that I’d rather not explain. My methods are usually simple, non-foolproof and lazy. It’s some sort of an intuition that works for me, but might not for someone else. There’s no science to it and hence, I don’t bother crystallising it into structured thoughts or sharing it with others. So, when people ask how I make matches, I usually give a blanket response such a hunch (which is partly true) but the answer really is as trivial as assortative matching.

When people approach me (or not), I judge them. As a matchmaker, being judgemental is an essential skill. It gives you a base to start learning about someone. It gives me a sense of who someone is good (or good enough) for from the pool of people I already have within my radar. The overarching metric of good for each other is driven by a combination of aspects like – general attitude, educational background, upbringing, family values, looks, general interests (or the lack of it as well), etc.

Depending on how rational the person is, they either find my match on par/ below their aspirations. If I were to tell you that you’re no Angelina Jolie, but you’re still very much entitled to want a Brad Pitt but you don’t have to land one, you would not want me as your match-maker. So, if I were to put this in a manner that is more palatable…

There’s an algorithm known as the Gale Shapley’s algorithm that is popularly used in matching markets and is found to be very closely applicable even to online dating websites. In essence, this algorithm states that if all parties in a matching market had a prioritized list of their preferred matches, then a stable match is possible within that market such that no one is cheating on the partners finally assigned to them through the matching algorithm.

In the real world however, making such priority lists is very costly. Firstly, you don’t have complete visibility of all potential matches at time zero and hence, you can’t possibly have your preference list ready. Secondly, it takes an incredibly long time to build a priority list since the transaction cost of assessment is quite high.

Imagine you had a binder full of women, you’d pick your top 5 women and go down your order of preference one by one. Let’s say Woman No.4 is your ideal match at time zero but you wouldn’t know that till you came to Woman No.4. So, you’d start at Woman No.1, put random whatsapp texting, maybe a meet or two and do some casual horoscope matching (in random order depending on how orthodox your family is), only to get rejected by her.

You end up repeating this process for Woman No.2 and 3 and finally arrive at Woman No.4. But thanks to the high transaction cost, Woman No.4 is already gone or has just started the process unlike you and hence, rejects you since she’s still ambitious and so now you have no choice but to only pick from Woman No.5 and below.

If you’re young, you are assured a fresh new pool of women and hence, there might be a greater chance of finding women who are far superior in the new batch. But, very often, the pool only decreases with age. More so for women than men. If you don’t believe me, hopefully Christian Rudder’s Dataclysm could convince you better?

Recently, I made an assortative match and voila, the couple’s getting married later this year. When I got the details of the girl, I immediately knew a boy in my Marriage Broker Auntie database who’d be good for her and I didn’t waste a second trying to set them up. It took nearly 2 months to just organise the introductions (as both of them had been busy on Bharatmatrimony and what not) but since I had very strongly and explicitly anchored the match as being perfect, the 2 month transaction cost didn’t matter much. 

This is a great anecdote to show that clearing house models take forever to find you matches unless you have a well-wishing judgemental marriage broker auntie working exclusively for you.  Ahem.