Hacker News Reader

Comment by abeppu | Hacker News Reader

So on the one hand it's nice to see when a cool algorithm or framework is used IRL.

On the other hand, actually applying a 'stable marriage algorithm' to people makes a bunch of assumptions which are worth interrogating:

- Do people actually know what they want even in the short term?

- Do people's preference change meaningfully over longer periods of time?

- Does your preference _upon reading people's profiles_ have any strong relationship to your preference after spending any amount of time with them?

The property that right after being matched, no one could have matched with someone whose profile they preferred more does not mean that this will create stable relationships.

View on HN

Least they are using an algorithm that has the intended purpose of matching people. For profit dating apps are more interested in keeping you coming back then they are in creating successful relationships

This is an extremely common sentiment I see online - that there is some conspiracy to keep people single to reduce subscription churn. I have run two dating apps (my own apps, not working for a company), and what I found through trial and error was that no matter how many levers you pull and tweaks you make, a lot of it is simply user-behavior problems, not a tech or product design problem. There is no reliable way for a dating app to only serve users non-committal matches for example. The reason someone's matches aren't committal is because they don't want to commit.

The thought experiment I usually pose is to ask what you think a free, open source, and community supported dating app would change. The answer is "not much".

You ran not one, but two, dating apps that were your own apps, and were not part of a company?

Which apps were these? What were your user metrics?

Seems somewhat suspect to extrapolate from your personal experience to what the large dating apps are allegedly doing if you are not approaching their scale. You've mentioned your personal experience as some kind of authority a few times now in this discussion, why not name the two apps directly? Apologies if I missed it but I don't see where you gave any indication that this appeal to authority is credible.

They weren't Tinder sized. They were only published in latam. The first (Mil Besos) flopped and my second attempt (MiAmor) got to a several hundred thousand downloads and I sold it for not very much because I was tired of the moderation. Neither exists anymore. This was more than 10 years ago. There used to be A LOT of dating apps around 2014-2018, I mean hundreds of them on the Apple and Play Store. There has been a lot of consolidation because almost the only thing that matters for dating apps is giving users optionality, which means having more user profiles the other ones.

I'm not claiming I'm an expert, but just sharing personal insight. It doesn't take many users before statistical observations of behavior make themselves apparent. Start producing 5-10k matches per day on your platform and you will see what people want and don't want really quickly.

I don't wish to belittle you but, if I were you I would certainly not be in this discussion saying things like "Source: I used to build and run dating apps" (you said this in another comment in this discussion) when making claims about the major dating app companies if these were the two apps in question.

Certainly not in the manner you conveyed your experience. The market dynamics and design problems of these apps are very well known. Given the scale of your experience, I would suggest your invocation of it as some kind of authority was superfluous at best.

Congratulations on your sale though. Have a nice day.

I think that source is fine. Not necessarily because running a small dating app gives you the same experience as running a large one; but because it also means you thought long and hard about the space, too.

Definitely more than your average armchair backseat quarterback.

I'm happy to hear from someone who's built 2 dating apps on the topic of dating app algorithms, they probably know more than I do. I appreciate the contribution, and don't think the gatekeeping is necessary.

I think the claims were about behaviors for 10s of thousands of users of the dating apps, which limits the range of actions a company could take.

Considering that population size, wouldn't millions of users have the same behaviors?

Note that it's possible for dating problems to be user behavior problems in a way that is suits the profit motive of dating sites, without any conscious intent from anyone in the ecosystem.

Model a world where everybody's self-perception of their mating value is just a little bit higher than their actual value - in other words, one not too far off from reality. Some segment of this population will drop their standards in the face of repeated rejection, and then eventually find a partner because of this. Their "lowered standards" become an accurate self-perception of reality. But some segment of the population will be unwilling to do this. They have a problem - they cannot find a partner no matter how hard they look.

A problem creates a market. People who can't find a partner turn to dating apps to find a partner - some may pay, and a lot more will click on ads. There's a decent chance that dating app management isn't consciously optimizing for failure in dating and success in clicking on ads. However, all the apps that are good at finding partners for their customers find that their customers disappear, while the apps that are good at getting people to click on ads generate huge amounts of revenue and take over the market. Which is entirely composed of people who don't actually want to find a partner, anyway, because those people who care more about the outcome of finding a partner will usually drop their standards enough to find one.

The same dynamic applies in many other markets. Evolution and selection bias > intent and intelligent design. The market ends up composed of people who can't achieve the ostensible purpose of the market, largely through their own actions, because people who can drop out of the market. Happy people often go through life with very few transactions: they might change jobs a couple times early in their career before finding a boss and job they're happy with and staying there until retirement, or they date around a little and then find a partner they marry for life, or they might rent a couple apartments before buying a house, renovating once, and then staying in it for 40 years. The transactions are overwhelmingly composed of people who are dissatisfied, because that's why they're buying things.

I am interested in creating my own dating app/website, but the problem I foresee is getting users - a Catch 22 problem.

I would like to create a service that opperated like OkCupid used to before they sold out. Visible compatibility percentages and being able to search profiles. No curation/algorithms or swiping.

I'm interested in helping people, not taking their money.

Getting users to dating apps is probably easier than you think. Because men will go to where the women are so you only need to work on attracting female sign ups. It's actually that simple. The best way to launch is with a waitlist of warm users in 1 large city. I've seen a lot of people pining for the 2010-era OkCupid so you probably would have plenty of people willing to give it a try.

> There is no reliable way for a dating app to only serve users non-committal matches for example

While that certainly sounds difficult, it sounds feasible for e.g. Facebook.

> The thought experiment I usually pose is to ask what you think a free, open source, and community supported dating app would change. The answer is "not much".

That sounds more like a framing problem. If the ultimate goal is promoting long-lasting new romantic relationships, and it's developed without any profit motive, I'd imagine a general social network that primarily organizes in-person social gatherings.

Facebook has a dating app and it's completely free. There are no subscriptions or paywalled features. It works about as good as all the other ones.

Except your dating pool is limited to Facebook users.

My take is that it isn't about serving non-committal matches and withholding promising matches or whatever, it has more to do with whether you feel like there are better matches on the app if you just keep lookin or waiting. I can see how and app using the stable matching algorithm could really make you feel like you're getting the best match you're going to get, which I think could have a real impact on commitment.

"it has more to do with whether you feel like there are better matches on the app if you just keep lookin or waiting."

Yes, the FOMOism in dating app user psychology is extremely real, but this isn't a product problem, it's a human psychology problem. In theory, you could make an app with, say, a lifetime number of swipes to prevent infinite FOMOism, but it would fail to be a viable product model because the #1 thing that I found users wanted -- more than everything else -- was optionality. And if you don't offer it, they'll go to someone who does.

Hinge actually already uses Gale-Shapley and it doesn't seem to be producing meaningfully improved outcomes.

https://blogs.cornell.edu/info2040/2021/09/30/hinge-and-its-...

Right, FOMO is a human psychology problem...

...that many businesses exploit for profit.

But a business model taking advantage of it and causing it are two different things.

Yeah, this seems to be a truism along the lines of "scammers intentionally use poor English to filter for stupid people". It gets repeated a lot because it resonates with people, but there's no evidence it's actually done deliberately. Sure, they're making UI and other decisions to maximize for time on the app, but in order to avoid serving you the best matches for a long-term relationship, they'd have to know what those are, and still serve you only plausible, appealing matches that are somehow bad (if they were too bad, you'd quit using the app). I don't think they have this much insight.

*I came very close to falling for a scam that was written in perfect English, formatted convincingly, and appeared to come from an organization I was expecting to receive a message from at the time. Believing that scams will always be obviously stupid increases your chances of falling for one that isn't.

> but in order to avoid serving you the best matches for a long-term relationship, they'd have to know what those are

No? With modern dating apps they can just serve you people who have already swiped left on you. Even if they made it a random presentation of candidates, they will never be incentivized to matchmake for free. If you're not paying then you'll get shown people in the area and that's it. That's all they need for their ad-revenue.

I think we're talking about different populations. You are talking about someone who gets zero matches at all but we are talking about people that get matches, but those matches never turn into relationships. To be fair, both populations of people have the same complaint that "dating apps don't work", but for extremely different reasons.

I don't think that dating apps are showing users profiles that have already swiped left, and if they do, it's not an attempt to get users to convert to a paid subscription. Consumer psychology doesn't work that way. People are far more willing to pay to upgrade a good experience than they are to pay to try to improve a bad one. Further, an unappealing profile that gets zero likes will continue to get zero likes after they upgrade to a paid subscription. This means that user will churn almost immediately after 1-2 months of paying for something that doesn't work. They typically don't stick around paying for years while getting zero matches all the while.

Dating apps DO derank unappealing profiles, but rather than doing it to try to get them to convert, it's instead done to try to prevent high bounce rates. Imagine a user downloads a dating app and is shown profiles in completely randomized order. This user is far more likely to get frustrated and uninstall after a few minutes than if they were to download the app and be shown profiles that are appealing. This rigged ranking may sound like its reducing relationship formation, but they never would have swiped on the unappealing profiles anyway; the counterfactual is they simply uninstall the app in frustration.

I agree with this. In my latest stint on dating apps as a woman looking for a man, I matched with dozens of profiles from guys average in terms of height and salary (when disclosed), and also in a niche of "attractive to me personally but not in the top 10% of conventionally-attractive dudes". So I know dozens of guys got at least one match, and most of the time they liked me first, I hadn't swiped in either direction when they were shown my profile.

Upwards of 90% of matches did not convert to dates, because many did not respond to messages at all, did not or could not carry on a brief "normal human" conversation, or if they did the first two things, never proposed meeting irl. I think you are correct in your assessment that the median user on a dating app is very non-committal.

I don't know... mid 2010's pre-match.com OkCupid and PlentyOfFish seem to get a lot of love. Not saying they didn't have problems but they did actually match a lot of happy couples and online dating didn't feel like nearly as much of a wasteland as it does today

That's before they were acquired by Match Group.

the stated vs revealed preference problem is very big when it comes to dating, and it's made worse that a lot of people have confusion and even shame about what their preferences actually are. And a lot of preferences are mimetic, which means demand for certain things is always outstripping supply. So it's really not even worth it to figure out someone's preferences if there is literally no one out there to pair them to.

The productive question would be to figure out how far outside someone's preferences they can go while still settling happily.

I think what you should be doing as a dating app — much more so if a government one — is some kind of RL based on length of actual marriage.

At least, you should have some algorithm estimating marriage length / likelihood of divorce based on your (non-app) statistics, and then use that as one factor for your in-app calculations.

Just to call out, you're implicitly assuming that strictly lower divorce rates is the correct end-goal. This may not be what you really want, eg you might find many abusive pairings where the weaker partner is too scared to leave. In the style of "The optimal amount of fraud is non-zero" [1], I'll claim that the optimal amount of divorce is non-zero.

[1] https://www.bitsaboutmoney.com/archive/optimal-amount-of-fra...

The sentiment of the linked post is different than the one you describe. The post is explaining why it's ok to have a non-zero value for a metric seen as objectively bad (like fraud) if there's high confidence that the non-zero value can be traded for a disproportionately larger metric somewhere beneficial (user acquisition).

In this case the undesirable metric is divorce. If the parent comment's feature for measuring marriage lengths and estimating likelihood of divorce were implemented, what disproportionate benefit would allowing more matches with a nonzero likelihood of divorce bring? The app wouldn't be enforcing whether divorces happen, it would just use the estimation of it happening as a matchmaking metric.

> what disproportionate benefit would allowing more matches with a nonzero likelihood of divorce bring

I think one benefit would be the individual participants' self-determination (there may be others). It's ok if you don't find my previous example (abusive relationship) convincing, but in the absence of an objection I'll reuse that. If we could look through all the marriages with a 0% chance of divorce, my bet is that we'd find an abnormally high percentage of abusive relationships. I claim that if a match has _no_ chance of divorce, it probably indicates a lack of self-determination for one of the partners.

Taking a step back, it might be clearer to just argue that length / non-divorce is not what people really want from a marriage. I can't speak for everyone, but I want a happy and fulfilling life, and I seek a partnership to the extent that it supports my desire for a happy and fulfilling life. Yes, I do want the partnership to be long, and I don't want to get divorced. Of course I want my marriage to last forever. But this is not my primary goal in marriage. There are many boring or uninteresting people (apologies to them) who I'm sure I could live with forever, that I'd immediately discard as candidates. Alternatively (making the abuse angle more tangible), I also have a lot of international friends and family that would happily find me an international bride who would never divorce me, and would be happy to marry somebody with a US passport. In fact this is a somewhat common offer, though of course they never say it in those words. You can decide for yourself whether that is abuse, but at least I wouldn't feel comfortable with it.

> I also have a lot of international friends and family that would happily find me an international bride who would never divorce me, and would be happy to marry somebody with a US passport. In fact this is a somewhat common offer, though of course they never say it in those words. You can decide for yourself whether that is abuse, but at least I wouldn't feel comfortable with it.

I'm pretty sure in this specific case, they get an instant green card if abuse happens. So the incentives point in the opposite direction (pro-divorce)

That’s fair, but on the other hand they’re in a new country where they may not speak the language, and know almost nobody. To say nothing of religious/cultural views of divorce

But if the abuser can be obligated to pay legal fees, there could be incentives for lawyers to seek these kinds of cases out

Gale-Shapely provides the best possible stable outcome for the proposing side.

You’re never going to match with people who aren’t actively looking.

You’re only ever going to match with the people who are looking at the same time as you.

Your criticisms are true of real life without apps. That’s why we have this thing called “divorce”.

Comments like these always seem to miss the point. They aren’t trying to solve marriage. They are trying to create social lubricants to get young people more interested in dating. This algorithm tries to ensure that at the very least your core values at the time match up.

The match is whoever wants to start on the journey with the user, at that moment, and the algorithm is there to give the potential couple a better starting point. I guess at any rate that’s better than leaving it to hope and chance.

I don’t think you are legally bound to marry. And if it stops working, divorce before resentment.

Reply on news.ycombinator.com