Filters shrink the pool faster than anybody expects

There is a settled way of talking about search filters. It is worth asking how much of it survives contact with the detail.

Filters multiply, they do not add

If a height filter removes a third of the pool and an age range removes another third, the two together do not remove two thirds and stop there. They compound, because each restriction applies to whatever survived the previous one, and four mild filters can leave a small fraction of the original field.

People generally set them one at a time, judging each as reasonable in isolation, and never once see the combined effect. The visible result is a thin feed and a sense that nobody suitable is on the app, which then gets blamed on the platform. Widening a single filter by a small amount often restores far more of the pool than the size of the change would suggest.

The removals are invisible

A filter is unusual among decisions because you never see its consequences: the profiles it excluded simply never appear in front of you. That means there is no feedback loop, and no moment where you notice you have excluded somebody you would have liked. Compare it to meeting people in person, where you regularly find yourself getting on with somebody who would not have survived your stated criteria.

Here is the part people skip: most people can name a past partner who fell outside at least one of the filters they currently have switched on. Reviewing your filters against your own history is a quick and slightly uncomfortable exercise that usually loosens several of them.

Deal-breakers and preferences are different objects

A deal-breaker is something where the answer decides the outcome regardless of everything else, such as whether somebody wants children. A preference is something you would choose given equal alternatives, and that describes most of what people actually put into filters. Filtering on deal-breakers is efficient because you would have ended the conversation anyway, so removing them early costs you nothing.

Filtering on preferences is expensive, because it removes people you would have happily met had you simply seen them appear. Sorting your criteria into these two piles before touching a slider is usually more effective than any adjustment to the sliders themselves.

Distance and age behave differently from the rest

Distance filters are worth setting honestly, because a relationship that requires two hours of travel each way is a genuinely different proposition. Age filters tend to be inherited from whatever you set years ago and are rarely revisited even as your own position keeps changing.

A common pattern is an age range that would exclude somebody a year older than your last partner, set without any thought behind it. Both filters interact with population density, which is why identical settings produce a full feed in a city and an empty one outside it.

If your feed has run dry, distance and age are the two levers that move it most for the least loss of what you actually care about.

What filters cannot do at all

No filter captures pace, humour, generosity, curiosity or whether somebody is kind under pressure, which is most of what makes a relationship work. Stated attributes are the easiest thing to index and among the weakest predictors of whether two people will get on in a room. This is the central limitation of the format: a platform can only sort on what can be typed into a field.

Research on attraction has generally struggled to predict compatibility from stated preferences, and people frequently end up with partners who do not match their own criteria. Treating filters as a rough gate rather than as a specification keeps them useful without letting them do work they cannot do.

If any of this feels coercive rather than awkward, that is a different conversation.

A practical way to reset them

Turn everything off except genuine deal-breakers, spend a week with the wider feed, and notice who appears that would not have before. If the wider feed is unmanageable, tighten one filter at a time rather than several at once, so you can see what each one costs.

Keep a note of the criteria you relaxed and whether anything came of it, since memory tends to confirm whatever you already believed. Pay attention to whether a filter is protecting your time or protecting you from having to make a judgement about a person. The goal is not the widest possible pool; it is a pool where every exclusion is one you would defend out loud.

The takeaway

Filters compound. Keep the ones you would defend out loud and switch the rest off for a week.

The right person will not need you to be strategic about it.