How Hinge's algorithm decides who you date
gizmodo.com
gizmodo.com
I don't see how that's too mysterious, and, for the most part, I don't hold it against Hinge, which has to try to make money.
edit: also, the other thing I forgot to say, is that I have also just keep deleting and recreating my account for every dating app. I don't do that to "game the algorithms" for these apps, though that may be a secondary benefit. I do it because I get tired of them, find them toxic/unhealthy, etc. I kind of just assumed most people did that, and I bet if the author's friend just shrugged and did that in the first instance instead of alleging foul play and conspiracy, they would have had a better time.
IIRC, Hinge always starts off by showing a good number of very attractive people in the main feed regardless, which I always assumed was partly to impress a new user with how worthwhile the app is, as well as trying to figure out how attractive the user is by seeing if attractive people respond to them.
Pretty much every dating app makes its money off (thirsty) men. Average women get hundreds of matches a day; ten a week is a very good rate for the average man.
I'm too old to have used dating apps, but IIUC the conspiracy theory is not just that you have to pay a premium to message the attractive girls, but they're actually paid by the dating app itself to interact with you. (I'm too old for dating apps, but I'm not too old for certain types of "clubs" whose business model was the same).
In Asia for all the popular sites like Tinder, Badoo and OK Cupid -- I've never had a girl not show up where there wasn't mutual interest. Continued after yes or no -- that's just a human problem. But the girls are real for sure...
Like playing the slots: you get occasional, small payouts, and you hear about other people winning jackpots (oops, they slipped up), so you keep putting money in.
A believable service needs to align its interests with yours, in some believable way.
1- it is possible to completely tap the dating market and long term customers are necessary
2- customers all have a similar threshold at which they are just-successful enough to stay engaged long term
Neither of these things are true. The more people "win the jackpot", per your analogy, the more people will use the platform. The market is big enough they don't need to fool you into paying to be unsuccessful.
Don't they just create fake bot accounts to draw you back in when they detect you've abandoned the app? I'd imagine the app is tuned to maximum scam based on the individual user not using a single set of values.
The problem is hard. It can’t be deterministically solved. Any likely statistical improvement is unprovable in a marketing sense (a competitor that is as good as possible at finding matches cannot prove it to the market that they are actually better at finding matches than a method that strings people along).
Your points are only true if there isn’t an information asymmetry, but there is.
I don’t think they are necessarily stringing people along on purpose, but they don’t have incentives to tell people they can’t help them, and they don’t have incentives to improve the product for those who aren’t successful. Those who are successful are probably going to be successful regardless of the platform they are using, and those who are unsuccessful aren’t going to be successful anywhere. And fixing that (rather than just milking money from the unsuccessful) isn’t really worth anything to them.
Don't they? I mean, maybe not intentionally, but it certainly evolved to be that way.
The more they can get you to pay for a once-off party, the less money you have entering your marriage. The less money the man pays, the more encouragement is bride is given to be bitter. This results in entering a marriage with less financial stability than you'd otherwise have.
They sell the message constantly that it's okay to forgo things like home ownership as long as you get the "perfect" (i.e. most expensive) marriage.
There's ever more useless things added to the ritual, resulting in an ever more stressful time due to more friction and more bitterness afterwards. They have spoken rules about how much you should spend based on a multiple (not a fraction) of what your income is.
They don't care, and often even encourage, allowing the future in-laws to make a loan so that the couple can get the best marriage. That certainly looks like optimising for divorce to me.
They sell it as a once in a lifetime experience, but they already know from their own experience that it's 50/50, if that.
To me it looks like they are optimising for revenue, which indirectly means optimising for divorces.
Like, needing to go down to the "Yum! Brands" to get take-out, no not the pizza one, the chicken one. Other chicken one. Reverse-mergers would be a nightmare too.
These days you mostly see it voluntarily done when a company is afraid to sunset an ancient brand with good history but also wants their brand noticed and associated (things like "Columbia Pictures, a Sony Corporation"), but some of those sorts of disclosures in the past were required for certain forms of advertisement (even if just in fine print) and maybe it is time to go back to enforcing that sort of thing.
There is the "Lenovo strategy", where some of the companies do sunset the original brands once licenses become too expensive and they've established their own brand. Lenovo went from "IBM ThinkPad by Lenovo" to "Lenovo ThinkPad" over a relatively few years. I'm still curious if, among others, "GE Appliances, a Haier Company" is a short-term play of this sort with the end goal to drop the "GE Appliances" name and logo in a few years or if Haier is less confident in their own brand and will stick with the mouthful (and GE licenses) long term.
(The other example of Sony would probably be too much for American consumers if they killed Columbia and/or TriStar brands. There's a deep-seated stickiness to the old Hollywood studio brands. Amazon seems to have learned this lesson real quick post-MGM acquisition and backpedaled quite fast on dropping MGM branded stuff for the Amazon Studios brand. They've even backpedaled so hard that they resurrected the Orion brand decades after MGM mostly let it retire. It weirdly blew my mind seeing an Orion Pictures logo on a 2023 trailer for an upcoming film.)
(Though today with tools like Wikipedia more information of that sort is "easily" researchable than ever before. But truth in advertising laws have always said it is better to save the advertising reader/viewer the research work than to trust people will do the research.)
It also allows them to capture the money of disgruntled customers. Someone who has a terrible experience with (say) Tinder may swear it up and start using a "competitor" like ... OkCupid.
IIRC, this is also a strategy in home appliance space. My understanding is most home appliance brands are owned by just two companies: Whirlpool and Electrolux. They've really cheapened their designs, but if you get pissed off your washer breaks after 3 years and decide to go with a competitor, there's a good chance you may end up buying literally the same model, from the same company, with different branding and styling.
At least twice investigators thought they found all of the models and brands in play and then discovered even more. Plus, an older/related recall also still exists covering even more brands and models between 2005 and 2014.
One supplier, hundreds of models across dozens of brands. "Competition" at its finest.
https://abcnews.go.com/US/wireStory/15-million-dehumidifiers...
I have no data to suggest this will happen aside from the fact that there's a large segment of the population that is having more and more trouble finding partners this way due to algorithmic changes forcing them out of eligibility to talk to people they find attractive. I've had at least one friend get permanently banned for violating a rule she didn't violate on Hinge, and their policy was essentially "tough luck, nobody gets unbanned for any reason".
If such a shift did occur, it would be an interesting reversal from the early days of online dating (eharmony, match.com) where the perception was that it was a fallback for people who were unable to meet someone in person.
It's been a while since I've been on Hinge, but Hinge is not like Tinder in terms of edgy/risque conversation, and I played every conversation with a match safe. I woke up one morning and was banned. They say you can appeal, but this seems like mostly something to make you feel like there's a possible recourse.
What I gathered from Reddit is that it's actually rather easy to get banned by simply not talking to people. There were a few times where I wouldn't message someone right away from not having a good ice breaker, I was busy, or the conversation just went nowhere and people would unmatch. When you unmatch with someone it asks if you'd like to report them as well. Since Hinge is meant to be a more serious dating platform, not talking to people can be seen as not taking the platform seriously and trying to hook up or something and if this happens enough times it seems Hinge will flag your account.
This fact also goes into the calculation when grading user interactions.
Next time around I tried Tinder et. al. was a couple of years later(2018/2019), and I immediately noticed less matches, less convos, and the app being more pushy.
I recently, as in this year, tried some of the apps again...it's a total shit show, to be honest.
I've never tried any apps that applied so many dark patterns. It all feels like one big funnel where the goal is to extract money off you through gold memberships, boosting, extra swipes, and what not. I noticed "fake" notices, spammy notices, just a constant barrage of notices and messages NOT from actual people.
After a couple of days I concluded that most (all?) of these dating apps are pay-to-play. You start off with some initial amount of matches, which drop off sharply - then nothing. I asked a friend in the same room if she could see me on Tinder, but no dice, I didn't even show up on her feed - though she came up on mine.
These apps used to be easy and fun - now they just feel like the most pushy and parasitic apps out there.
They also filter you based on age. When I turned 40+, everything changed in terms of presentation and matches literally the next day.
I built a super popular dating app right before Tinder came onto the market... our search was intentionally pretty fuzzy in this regard.
Because you want to ignore that and nitpick that I mentioned other apps doing the same thing, as if that prevents Tinder from doing it, I went out of my way to get a screenshot of my own Tinder settings. What do you know? It's right here. "Only show people in this range" https://imgur.com/a/yt34L2b
Upon further reflection, I'm just going with "they were actually correct".
So yea... maybe they are actually correct, now.
They are scum.
Having worked there I could game their algorithms quite well.
Hate the fact that third spaces have evaporated and you need a dating app to meaningfully participate in the dating marketplace, not that you have to take liberties with facts to increase your opportunities and potential to close into a relationship.
Tangentially, love is grown, not found. Find someone you can grow to love (ie trust and mutual respect, shared values and belief systems).
But, I certainly know that if I had lied initially, it would have never worked from the start.
Your attitude on this matter ("play to win" and "work on your sales skills") is what makes the whole scene really horrible. I feel for your partner, I wonder if they realize that is your train of thought.
After saying what you did, I find your advice at the end contradicts what you said above. "play to win"... but grow love with trust after lying about my age? Weird.
I agree the dating marketplace is terrible, but you can’t change it, you can only change how you operate within in to arrive at a desired outcome. Would you own a business but do no sales or marketing because you find it icky? That’s a quick way to bankruptcy.
That is to say that most relationship success is timing, luck, and emotionally healthy enough people (have to love yourself first before you can love someone else in a healthy way). I wish I had more actionable advice, but I don’t. Be a decent person, sell yourself, and get in front of as many (emotionally healthy) potential partners as possible. That’s it. I agree I am in the minority. I recognize luck for what it is, but try to encourage everyone to get more dice rolls in (see: this sub thread). I wish you luck and happiness.
I really don't know the solution to this problem but until someone solves it all dating apps will devolve into dark patterns.
Explain the game to the user and make a package where the company reserves a large-ish sum on the credit card and they get to cash only when the user stops dating after a successful match.
The dating app doesn't make any money until a successful long-term. The dating app could also require feedback after dates, to both improve matching and verify that one user is not cheating.
If a couple colludes after they get i long-term relationship, they could continue dating... but it increases risks of cheating.
This aligns the interest of the daters and the dating company (less infrastructure expenses and also less risks from reserved but unpaid transactions).
There are open relationships, the app can learn to screen for people with profiles that are prone to that and avoid making the escrow offer.
Also, it means zero monetization for people looking for single encounters and/or flings, which I would bet is a meaningful portion of Tinder's income.
I am shocked at some of the prices I saw people paying? $100+ dollars/month to get a reasonable match? How much more monetization do you need there? A $60/year Strava subscription has done more for my love life than this.
Basically, they monetize by setting people up on dates at specific places they have a deal with, driving customers to these places.
And truly successful in the sense 'never ever need the service again' probably has enough obstacles after the initial matching and dating phase that there could be a fair number of returning customers nonetheless.
Having enough users that could justifiably be called "customer" at all may be a catch of a different kind though. Can't say I know for sure, but I strongly suspect the network effects make it very difficult to attract useful amounts of people to match with to a purely paid service initially, even if it would be readily worth the fees to everyone involved if the system was already established with a large enough network.
Maybe it would have been neat if Musk had bought, and tried to reform, that major online dating service instead of the little bird company...
If you think it's eroding your self image then you can be fairly sure they've tuned something to do that to you because they think it will encourage you to part money for the perception of romantic advantage, or because they think it'll get your eyeballs on some ads, or for some other reason that funnels it money. The best move both romantically and for mental health 90% of users can make is to go touch grass, stop pushing for a date so hard and wait for serendipity to do its thing.
It's definitely most and certainly feels like nearly all, especially once you start to notice just how many of these are all run by the same scummy corporation, Match Group [1]. Most of them are just thin skins over the same underlying terrible monetization platform.
Most of the dating apps that aren't from Match Group generally either have their best features stolen by Match Group and monetized and then out-competed to death or live long enough to see themselves become the villain and reaching for the same monetization dark patterns or even just exiting directly into Match Group.
(Among others, OKCupid was explicitly founded to be the hero against Match Group and eventually succumbed and was bought by Match Group. Tinder was built by whole cloth by Match Group to steal good ideas for features from smaller apps including but not limited to Bumble and was built to be an intentional slow boil-the-frog as it slowly added in other Match Group monetization features and then slowly started passing on its own features to the overall group "platform" as the features themselves stopped being the novelty. Now nearly all of Match Group's apps act like Tinder and swipe left/right.)
[1] https://en.wikipedia.org/wiki/Match_Group#Dating_services_ow...
During the first year of COVID in San Francisco I had a date with a girl who I found completely annoying, she was rude to the waiting staff and insisted on finding “sugar daddy.” I couldn’t believe I was listening to the delusional beliefs this woman had, especially regarding she lied to me about her career after I sat down with her.
I told her exactly how I felt about her and that she really needs to fix her attitude towards people and men, and that I’m done with the date. I left, and never contacted her again.
Coincidentally the next day my tinder account got flagged and banned. I knew it was something to do with her and she was spiteful (though I couldn’t prove it), but the timing was impeccable. Someone had filed a report against me and the automation system at Tinder wouldn’t allow me to speak to anyone.
I couldn’t find a partner and every bar / club / restaurant / location in the city was closed for MONTHS! Tinder was literally exiling me to my room and closing off dating matches (I realize there are other apps but the pool is huge on tinder and I always found the most matches there and best outcomes for my lifestyle).
I think this should be against some policy or something. If folks just understood the isolation I felt those months, going from meeting women every week to 0 was very bad on my mental state.
I don't know how that situation is allowed to persist. There should be some serious antitrust action is this space. What Match Group is going is 100 times more blatant that anything Facebook's done. At a minimum they should be forbidden from acquiring any more competitors.
And sure, restricted users will make new accounts, but that's a problem as old as the internet and something companies just have to fucking deal with.
Of course, we would have to season every sentence with some statistical weasel words ("tends to," "more likely to," "on the average," and so on) and we would have to risk bumping up against someone's carefully crafted worldview in an injurious manner. So be it.
Recall the now-vanished (Thanks, Match.com!) OKCupid blog posts regarding the different statistical spreads between the sexes. The more we try to obscure these differences, the more warped the strategic arms race between The Algo and the Lonesome Losers* will become, as The Algo will try to hide any unfairness and keep you playing.
However, there are a lot of different parties who are quite invested in avoiding bald statements about the reality of the situation. I don't see this changing.
* Lonesome Losers, named after the Little River Band song, strike out but they still keep on trying. They will, however, try to adapt.
(As a side note, "shadowbanning" is just gaslighting by moderators and/or The Algorithm. It's gaslighting, pure and simple.)
I'm not sure what the correct term is for gently nudging people to get off the platform by making their experience worse.
Instead it allows the people who have all day/financial incentives to learn the rules.
I see you have never worked in anti-abuse.
> Rule breakers often make new accounts after an outright ban, so the subtle yet frustrating shadowban can be a better way to curb misbehavior.
The 'Algorithm' applies outside of one dating app.
When tinder came out my single friends all jumped on it. I probably went hoarse with all the tinfoil hat warnings I doled out.
Eventually I shrugged my shoulders and admitted maybe corps aren't as evil as I imagined.
Looking at this article I realise I was a fool to try and judge them during the onboarding phase. They are worse than I'd imagined.
This section is full of mistakes. Especially:
> To use an overly simplistic example, if a lot of 25-year-old women who like Star Wars keep tapping “like” on one man’s profile, the algorithm might show him to more sci-fi fans in that age range.
The example here shows the author fundamentally misunderstands how collaborative filtering works. I am not saying it couldn't technically happen, just as many things are technically possible in a variety of circumstances, but that this is indicative of a Dunning Kruger esque mistake and jeopardizes the rest of his analysis.
> Can you hack the dating app algorithm? These guys are trying.
The simplest and most validly performance-improving hack is to create a new account, which isn't discussed here. The recommendations otherwise gathered are incorrect. In the context of the author's poor understanding of the underlying facts, it makes sense why this was printed though.
If you want to learn accurate things about online dating algorithms, take a look at the resources here: https://monstermatch.hiddenswitch.com/algorithms
This is truly a nightmare. How did we allow such a crucial human function, finding a mate, to become monopolized? Both Democrats and Republicans need to date (though maybe not each other), so the issue should be bipartisan.
I feel like ever since Microsoft got a slap on the wrist back in 2001 (a few days before 9/11), antitrust law has more or less ceased to exist.
don’t talk about yourself
don’t be reactive.
What do you mean by "reactive" here?
That's not a comprehensive list of relevant societal changes, just some factors off the top of my head.
The very existence of the dating apps makes them difficult to avoid, because if that's where a significant % of the population are already looking, then you're ruling yourself out from that population by avoiding them.
I'm personally not hopeful. The world is going to shit, and the only option I have available to me to safeguard my kids future is to acquire as much wealth as possible for them.
If that doesn't work, that means a giant war or similar catastrophe has descended upon us and there is nothing I could have done anyways.
I’ve always wondered if this is because of globalisation. If you want your country’s companies to dominate the world market, then a strong home base would help.
OTOH, like all relatively simple apps, this ought to be trivially easy to create a competitor for, so it must not be economically feasible to do so (just like Facebook, Twitter and Reddit never got any competition no matter how user hostile they became). Why isn't there a free, non-evil Hinge competitor? Because it costs too much to run, and they have to do this to stay in business.
I remember it was always a constant "which app should we be using now" conversation with acquaintances
I totally agree with you, and I would go even further. I'm actually furious that an essential part of human social life has become commoditised by a for-profit industry led by publicly-traded corporations. It being almost a monopoly, hidden to "consumers" [0] under an illusion of choice, is only the poop cherry on top of the whole steaming pile of crap.
One of my fantasies to become a millionaire (from a starting point of billionaire) is to establish an open source, transparent, non-profit and community-based online platform and app whose only mission is to help people meet someone with whom they are happy to meet and viceversa. Imagine a dating app but:
- Nothing encouraging "power users", even discouraging match accumulation.
- No gamification, other than configurable by user choice.
- No secret algorithms for rating. Instead, a choice of open community-developed algorithms for the swipe queue (if swiping is your thing, you can also filter and sort by query).
- No falsifying the potential dating pool size to keep users hooked. Do not show the profile of user A to user B, even if A fits B's filters, if B does not fit A's filters such that A would not see them.
- User engagement KPI in the denominator, not the numerator—the numerator being user reported success and personal satisfaction. The less time spent for good results, the better.
- No "pay to win". Instead of that, please donate if you win.
My dream would be to create a dating(?) platform, nay, a human relationship platform, with no perverse economic incentives that conflict with users matching. I would hope it could be So Damn Good that anyone looking for a date, or a shag, or a FWB, or to meet people to see where it goes, or whatever would feel like a chump to fall for Tinder, Bumble, OkCupid, etc. Let swiping and notification addicts continue to use those for-profit games. Meanwhile, people wanting a human connection could stick to the platform that cares to make it as likely as possible.
Unfortunately, the only thing stopping me from doing this is that I have no plan in motion to become a billionaire such that I can finance this idea ¯\_(ツ)_/¯
.
.
[0] No, damnit, searching for a relationship, a fling, a connection, is not an act of consumption, it is an act of basic fucking humanity!
I think the reason we don't see competitors like that is that dating without aggressive monetization won't be much better for the average user.
https://today.yougov.com/topics/technology/articles-reports/...
I'm not arguing that the dating apps are particular good or successful. I wouldn't be complaining about Match Group if they were. In fact, there's been a rising share of people who are unpartnered: https://www.pewresearch.org/social-trends/2021/10/05/rising-...
In another comment, I talked about societal changes that have resulted in more social isolation: https://news.ycombinator.com/item?id=37213625 I think it's a problem is that we've become increasingly dependent on dating apps. Nonetheless, I don't think "just don't use dating apps" is really a solution, as if they were "strictly optional", because we can't magically go back to the good old days where everyone met their partners in person. The world has changed, and the old world doesn't exist anymore. In many cases, the alternative to avoiding the apps is singleness and loneliness.
We might all be better off if we collectively decided to abandon the dating apps. Short of that, however, it's kind of like "unilateral disarmament".
While the amount of users of that group are high, most of them are still ultimately going to meet their future partner through other avenues. It's worth highlighting that by definition most of them are unsuccessful at that point. It might be that actually one of the apps' main functions for some people is simply to give people some common ground as they exchange grievances about the experience...
On the amount of people in relationships, the data only seems to account for cohabiting and married couples, everyone else seems to be grouped into 'unpartnered'. As such this could also be reflective of changes in living circumstances - more young people still living with parents for example which may reduce opportunities for cohabiting. https://www.pewresearch.org/short-reads/2020/09/04/a-majorit...
In any case I agree with you that something is going on there but I still don't think that apps represent a good solution for most people.
My personal experience was that online dating was a disaster that (temporarily, fortunately) ruined my self-esteem and my best luck always came when I wasn't 'trying' to date at all.
The raw percentages are somewhat difficult to interpret for several reasons. 1) Many people on the apps are not looking for a partner, they're just looking for dates or hookups. In a sense, neither success nor failure of the apps are fully or accurately represented in the cited percentages. 2) The dating apps themselves are only a few decades old and have increased in usage over time, so anyone who has been partnered for a long time, before the existence and rise of the apps, needs to be ruled out beforehand from the data.
> I still don't think that apps represent a good solution for most people.
I didn't say they were. ;-)
> my best luck always came when I wasn't 'trying' to date at all.
Alas, luck is just that, something that can't be relied on, and only blesses the lucky.
The open secret is that people use then to meet sex workers, who in turn sometimes troll people who they perceive as "harassing" them by reporting unwanted spam.
Imagine being lonely, going into a space to find a date, and having someone make your mental health worse because you wouldn't pay them for sex, when the whole reason you joined such a site is because you want an emotional connection and everything else aside... just don't want that.
I haven't used hinge, but I ended up deleting Tinder and OkCupid -- having coauthored a whole ass paper on the latter.
People need to be better about not using a PhD as a way to lift yourself out of economic precarity if you are unlucky enough to have affluent and absusive parents.