The pull is real and the explanation everybody repeats is borrowed. Rewards that arrive unpredictably keep a behaviour going more stubbornly than rewards that arrive on schedule, in every setting where that has been studied, and nobody has run that study inside a dating app. Here is what is mechanism, what is allegation, and what is actually evidence.
What a variable schedule is
The idea comes from mid twentieth century learning research and it is simple enough to state in a sentence. If a behaviour is rewarded every time, it is easy to stop when the rewards end. If it is rewarded sometimes, unpredictably, it becomes very hard to stop, because no single absence of reward carries any information. Nothing tells you the run is over.
That is the whole mechanism, and it is why fruit machines, fishing and refreshing your email all have the same feel.
The three places a swipe is unpredictable
Look at what the sequence actually offers you.
**The match.** You do not know which swipe produces one, or whether any will. The interval is genuinely unpredictable, not designed unpredictability necessarily, just the arithmetic of two people independently choosing.
**The reply.** A match is not a conversation. Whether a message gets answered, and how long it takes, is another layer of unpredictability sitting on top of the first.
**The person.** Even a good conversation might be someone who is warm on Tuesday and gone on Thursday. That is a third layer, and it is the one that hurts.
Three stacked layers of uncertainty is unusual. Most things that hold attention only have one.
What you can observe without any theory at all
You do not need a model to notice these. Check them in your own app rather than taking anyone's word, including mine.
Notifications that tell you something has happened without telling you what. A deck with no end, so no session has a natural finishing point. Paid tiers on several major apps that sell you the information the free tier withholds, which is usually who already liked you. Streaks, boosts and timed windows that make waiting expensive.
Those are product features that you can see. Whether they were built to produce compulsion is a different question, and it is not one that has been answered.
The effort is lopsided, and that gets worse as the session runs
There is a second thing going on that needs no theory at all, and for a lot of people it explains more of the evening than the reward schedule does.
A swipe costs you nothing. A real conversation costs a great deal: attention, a first message that is not terrible, the willingness to be ignored after writing it. So the two activities are not competing on level terms. As you get tired, the cheap action stays cheap and the expensive one gets more expensive, which means the ratio shifts steadily towards swiping the longer you stay in.
That is why sessions have a characteristic shape. The first ten minutes contain most of the actual reading, and the last forty contain almost none. By the end you are not evaluating anybody. You are clearing a queue, and a queue with no end is a task you cannot complete.
It also explains why the sessions feel so unrewarding in retrospect. You spent an hour on the version of the activity that produces nothing, because it was the only version you had the energy for.
What is established, and what is not
This is the part almost every article on the subject gets wrong, so it is worth being precise.
**There is no peer-reviewed empirical test of variable-reward design in dating apps.** Not one. Every confident article claiming these apps are engineered to be addictive is extending a mechanism from other settings, which is a reasonable thing to do and not the same thing as evidence.
**There is litigation.** A US class action, *Oksayan et al. v. Match Group*, was [filed in February 2024](https://topclassactions.com/lawsuit-settlements/consumer-products/mobile-apps/tinder-hinge-class-action-claims-apps-are-purposely-addictive/) alleging gamified, compulsion-driving design. The court sent it to arbitration and stayed the case. **No court has found that dating apps are addictively designed.** Those remain allegations.
**Regulation is silent on it.** Ofcom's guidance for dating and social discovery services covers [romance fraud, intimate image abuse, harassment and stalking, and gender-based harms](https://www.ofcom.org.uk/online-safety/illegal-and-harmful-content/dating-and-social-discovery-know-the-online-safety-risks-rules-and-how-to-comply). It says nothing about addictive design or compulsive use, because that is not currently within its duties.
**The academic ground is thin.** A 2025 review of 32 quantitative studies found only three used randomised designs, and concluded that ["the concept of online dating addiction remains debated"](https://pubmed.ncbi.nlm.nih.gov/40608479/).
There is also a warning worth carrying from the wider field. Researchers have noted that ["recent publications have suggested that nearly all daily life activities might lead to a genuine addiction"](https://pubmed.ncbi.nlm.nih.gov/26014667/). A mechanism that explains everything explains nothing in particular.
Why the frame is still worth having
Given all that, why keep it? Because of what it changes about how you read your own behaviour, and because of what it implies practically.
It moves the explanation from character to schedule. "I have no self-control" and "I am doing something that unpredictable rewards make hard to stop" produce very different next steps, and only one of them produces a plan.
It also tells you what to change. If unpredictability is doing the work, then the fix is not to swipe more carefully, it is to make the reward predictable or absent. Which gives you three concrete moves:
- **Turn off the notification that carries the unpredictable signal.** Not important-only. Off. The badge is the delivery mechanism.
- **Batch the checking.** Once a day at a fixed time turns an unpredictable reward into a scheduled one, and a scheduled reward is much easier to stop.
- **Cap the deck.** Fifteen likes, then out. A finite session has an end, and the absence of an end is half the problem. [Putting the app down mid-swipe](/guides/dating_apps/putting-the-app-down-mid-swipe) covers the mechanics.
Where this frame gets misused
Two ways, both common.
It becomes an excuse. "It's a variable reward schedule" is a true sentence that can be used to explain why nothing is your responsibility, and it usually shows up in the hour before a two hour session.
And it becomes a conspiracy. There is a real difference between "this product has properties that make it hard to put down" and "this company engineered your misery deliberately". The first is defensible from what anyone can observe. The second is currently an allegation in a stayed lawsuit. You do not need the second to be true in order to act on the first.
If you want the version that is about you rather than about the product, [the audit](/guides/dating_apps/are-dating-apps-making-things-worse-for-you) is a better use of an evening than any amount of mechanism.
Common questions
**Are dating apps designed to be addictive?** Nobody has demonstrated it. A US class action alleges it and was sent to arbitration and stayed, and no court has made that finding. The design features people point to are real and visible in the apps. The intent behind them has not been established.
**Is swiping the same as a slot machine?** It shares the unpredictable reward structure, which is why the comparison keeps getting made. It has never been tested empirically in dating apps, so treat it as a useful analogy rather than a finding.
**Does knowing the mechanism help me stop?** Not on its own. It helps you pick the right intervention, which is friction and predictability rather than willpower. Knowledge without a changed environment tends to produce better-informed sessions.
**If it is the design, is any of this my fault?** Fault is the wrong frame for both halves. The product has properties, you have an evening, and the only useful question is what you change. [Dating on purpose, if you go back](/guides/dating_apps/dating-on-purpose-if-you-go-back) is where that lands if you are not quitting for good.
Sources
- [Oksayan et al. v. Match Group class action](https://topclassactions.com/lawsuit-settlements/consumer-products/mobile-apps/tinder-hinge-class-action-claims-apps-are-purposely-addictive/). The 2024 filing alleging addictive design, and the arbitration order.
- [Ofcom, dating and social discovery services guidance](https://www.ofcom.org.uk/online-safety/illegal-and-harmful-content/dating-and-social-discovery-know-the-online-safety-risks-rules-and-how-to-comply). The harms UK rules cover, none of them compulsive use.
- [Problematic Online Dating, Journal of Medical Internet Research, 2025](https://pubmed.ncbi.nlm.nih.gov/40608479/). 32 studies, only 3 randomised, and the concept remaining debated.
- [Billieux et al., on the over-application of addiction models](https://pubmed.ncbi.nlm.nih.gov/26014667/). That nearly all daily life activities have been proposed as addictions.
