Hacker News Reader

Comment by goodside | Hacker News Reader

I worked in the online dating industry for five years (OkCupid 2011-2015, Grindr in 2020). I’ve heard this argument many times but it isn’t true. I’ve never heard of anyone in the dating industry applying this logic, because no app is good enough at making permanent matches that it makes sense to worry about.

View on HN

Funny enough some dating sites such-as jdate measure success based on marriage-kid outcomes. With funding raised based on those stats.

That's sounds like a church, lol.

> That's sounds like a church, lol.

Probably more like a synagogue, but point taken…

I would assume for the users of Jdate that marriage-kid outcomes is exactly their reasoning for using a niche service like Jdate. Other dating apps and services....maybe less so. On the one hand that gives jdate a clear target and metric for success and marketing, but also gives them a smaller % of the population to try to appeal to.

I think you're in agreement with the article - the author says that they didn't need to taint results because they were already terrible. And that they're not going to fix things because there's no incentive.

I think the OkCupid questions in that period were great (submit your answer, what you'd like your match to answer, and how important you think the question is). My ex found her now-husband on that site.

I tried it around 2010 and it seemed very focused on personality quizzes and oversharing; besides wondering how all these people had 20 different favorite movies, my 100% match was someone I already knew and wouldn't want to date.

Didn't seem to consider revealed preferences and the appeal of people who aren't exactly like you much.

The hardest part of a dating site is getting new people through the door. The personality quizzes, in a very buzzfeedish way, were very effective at drawing people in. In fact they were a little too effective, because they drew in many people who weren't serious about dating and just wanted to fill out quizzes.

What OKCupid failed at was being a usable dating site. The matching algorithm was awful.

When I was an OkCupid user, those questions seemed like a smart way to design matching at first, but the more profiles and match evaluations I looked at, I realized that the data they gathered was extremely dirty because too many people didn't understand this system. You'd see questions where they had filled out the "answers you'd accept" answer as if they were answering for themselves, or that they didn't realize they were supposed to check all the answers they would accept (thus rejecting people who actually aligned pretty well). A prime example is the question that asks "would you prefer to give a massage or receive one." You'd see all kinds of people saying they'd like to receive a massage, but they wouldn't accept a partner who said they'd prefer to give a massage.

If a dating app is filtering out people who can't comprehend a simple question then that's a bonus

Very Darwinian of you but it's not just a simple question, it's an oddly-presented survey whose directions are a little confusing and unlike any other survey you've ever taken. Proficiency on this task isn't really part of my dating criteria. I ultimately married a pretty smart person who utterly failed to understand it, but I had already learned to discount the "match percentage."

When such a large percentage of people blow it, I tend to blame the UX, not the individuals.

Good surveys are hard. Good online surveys doubly so.

> If a dating app is filtering out people who can't comprehend a simple question then that's a bonus

Conversely, if a question cannot be understood by, say, 25% of the population, it's not a simple question.

Plus, successful couples are great advocates for the website that introduced them. You lose two customers in the immediate term but gain more as time goes on. (Also OkCupid, 2012-2015.)

Do you have any insider knowledge from your time with Grindr on

1) why the app appears to remain a terribly bad buggy mess and never seems to improve? Is there any awareness of this in the company? Is it just a legacy code issue?

2) why the desktop web version was shut down?

3) why unlimited pricing is that outrageous?

1) If everyone on the app finds great matches, everyone will stop using the app, and therein kill the possibility of the company in making a profit.

2) They laid off developers to save money, and desktop apps can be hacked to get around paywalls and user controls a lot more easily than on mobile apps, so getting rid f desktop apps always looks like a win for a software maker.

3) There is so much competition with expensive cloud host overhead that if a company can gaslight users into thinking that their app is worth money, they push it all the way in a bid to maximize their traditionally weak earnings on the deceptively small user bases they have... They also create tons of fake accounts and profiles to also emulate that they are running vibrant communities.

> desktop apps can be hacked to get around paywalls and user controls a lot more easily than on mobile apps

This would imply that paywalls and user controls aren't implemented entirely on the server side, which sounds like an obviously bad idea?

Hmm, well it wouldn't be the first time:

- https://web.archive.org/web/20190223153650/http://www.ydesou...

- https://news.ycombinator.com/item?id=8173197

- https://news.ycombinator.com/item?id=28303536

Reply on news.ycombinator.com