Tuesday, November 28, 2006

Wise men say...

A wise man once said "That was not a pause, that was a boring gap!"

Actually it was Max Headroom.

If there's one think I like, it's refuting things; and I refute my last post.

Internet dating is not dead. It had just gone for a lie down, some tea and a biscuit. I found this rather good online matching (I guess; it's complex) site that works on a few simple premises: firstly, it's free (!) and secondly, there seems to be some thought put into how it works.

It's made by some Harvard-based mathematicians, who I suspect might be likely to become quite well-off once the Googlorg or Microsoft snap it up. Instead of psychology as a basis (or random monkeys, which is what most of the others seems to use) they use basic statistics to match people. They ask you a lot (I mean it. A lot. Think 500+) questions, most quite personal, and also ask what response you think a good match might give. They also apparently track everything - where you go on the site, the kinds of profiles you click on, etc. As an architect, I can see what this is doing right away - some pretty major set-theory analysis - probably using similar software to the kind of stuff used in designing airports and transport hubs. (Also the same stuff that WETA used to do all the big army scenes in LoTR, but I digress).

And it works.

The matches it suggests are actually closer matches for the type of person I am interested in than other sites that ask you to define a very limited series of responses to some fairly plain questions.

It reminds me of flat-hunting back in 1997, when the flatshare agency I went to asked me a whole lot of worthwhile questions like:

"Do you like wild parties? y/n"
"Do you get drunk every weekend? y/n"
"Do you enjoy taking drugs? y/n"

The woman who ran it looked at my survey and said "I don't think we'll find any matches for you in Mosman" (where I was living & looking at the time). As things turned out, she did have another group who were looking for a flatmate about 10 years older than me, and she gave me their address; and the rest became history. I did get a flat (and a whole bunch of really good friends) out of it; but I could have saved the $50 and found that bunch anyway, just by asking friends-of-friends. It seemed to me at the time the psychology of matching people with only 20 or so questions was a bit of an imprecise science.

OKCupid goes to quite some lengths to show you the science & math behind their system, and while my eyes usually glaze over with statistics, the graphs are pretty and look right.

One of my mottoes in life is If it looks right, it is right.

If there was a way of telling if it smelled right, I'd be even more sure; I think smelnet is going to be a part of Web 3.0. That's the rumour, anyway.

So far, I am one happy camper with OkCupid's system. Quite happy. I even suspect it smells right, too.

4 Comments:

Blogger alyshajane said...

I think OKCupid works quite well, too. :)

5:00 pm  
Blogger syn said...

I protest about this suggestion that psychology is the basis for most online dating sites!

and by the way, super big congratulations! you know we all want to meet her now. :) Whenever you can stop floating and get yourself back to the rest of us dreary lot. She sounds wonderful!

8:58 am  
Blogger Corinoco said...

You're absolutely right, Syn, I think the random monkey theory is the correct one.

11:06 am  
Blogger syn said...

You must mean random kookaburras. Bear says there are no monkeys in australia, even though I can hear them! He calls them kookaburras.

4:01 pm  

Post a Comment

<< Home