Day: July 28, 2011

The ascribed difference is pretty straightforward:

• Match percentage indicates a statistical likelihood of reciprocal romantic interest between a given pair of users
• Friend percentage indicates a statistical likelihood of reciprocal friendship compatibility between a given pair of users

While it can be somewhat confusing to see both numbers when one is using OKC exclusively for dating purposes, some users indicate an affinity for the friendship number as a "counterbalance" to the flames of mutual lust that may arise… That is, a high Match percentage could indicate a strong romantic/initial interest, whereas the Friend percentage shows potential for a longer term,less sexually-focused relationship.

As Di Huang says, however, the validity of either percentage "on the ground" in real world settings is questionable, and until OKCupid/IAC release hard data on date outcomes/feedback/success rates tied back to these match percentages, it will be impossible to validate how accurate the underlying algorithms are.

What is the difference between a match and friend score on OKCupid?

No.

If that were the case, one could also make the argument that any of the physical statistical profile fields were also a form of Eugenics – specifically, things like body type, race, height, eye color, etc.

Unless you have access to an amazing new site (tell me, I'll sign up!), there are no dating sites currently which use their matching algorithms to force a mating that produces offspring. This would seem to be a pragmatic requirement of any successful Eugenics program.

It is possible that, if enough credence were given to a given dating site's algorithm as being actually effective in producing accurate and successful longterm matches that would lead to offspring, and enough people worldwide joined as a result (numbering in the billions), that this would create a "defacto" Eugenics application – assuming that the site used at least some genetically-influenced characteristics for its matching, and further favored certain types of matches across races/cultures/etc worldwide. Data from 23andMe could then also be used to accelerate the speed of eugenic effects after the first few generations of offspring.

As someone working on an online dating site, I find the prospect to be a highly unlikely outcome within our lifetimes, either intentionally or accidentally.

Would a dating site that used 23andMe data be considered Eugenics?

Let's step back from the details of the question for a sec…

What you can learn from someone checking out your profile is simply that their account is active, nothing more. If nothing else, this should say to you that they are worth writing to, if you are interested.

Don't overanalyze things. If you write and get no reply from someone who did check your profile, you have exactly the same amount of data about their interest level as someone who hasn't seen your profile and doesn't reply to your message.

Given all the other reasons they may not have messaged you (many sites only let paying members send messages, for example), it's in your best interest to pursue everyone you take a fancy to when you're searching online, so long as you take your time, and make a proper effort.

To be perfectly clear: I don't think that "who's viewed me" is a useless metric, but I think that it shouldn't carry much. I also don't think that "never viewed me" users are more likely to be interested.

In the end, I'm just concerned that the querent might spend so much time trying to find someone "likely to be interested", that they forget to BE INTERESTING, and to PURSUE PEOPLE THEY ARE INTERESTED IN. 🙂

If you know that someone has visited your dating profile and not sent you a message, is it better to contact them or to explore other matches?

Off the top of my head:

• Differentiation between paying and non-paying members
• Differentiation between real people and bots/scammers
• Differentiation between active and inactive members
• Reliable behavior-based matching systems
• Iterative feedback on member dates/interactions to improve their matches over time

What is missing on existing dating sites?

This depends on what you mean by "online dating".

If you mean the "modern http-based web (aka InterTubes™)", that will take a bit of research to pin down in terms of "verifiably chronological ancestor to all dating websites"… Definite early front-runners as far as mass adoption would be Match.com and YourOneAndOnly.com, which Match later acquired and replaced its own offering with (1993-1994 timeframe).

Prior to that, however, numerous "sites" existed for online dating via USENET, Gopher, Telnet-based BBSes and MUDs, etc.

Even further back, proprietary dialup services like AOL, Compuserve, Prodigy and the like, as well as local dialup offerings in major metro areas running PC-based BBS software (only one I can personally recall is "Date-a-Del", running a Citadel variant in Minneapolis, MN, certainly active at least as far back as 1988) all offered various solutions to post and respond to personals advertisements, or even rely on computer matching algorithms to select potential mates.

And none of these includes all the sites/services that were used extensively for dating purposes as a by-product of some other function or intended mode of communication.

Note: I included dialup candidates in this answer specifically because of the origins of the word "online" itself in the context of computing – "a central computer establishing a connection to a peripheral or remote device".

What was the first dating site?