The instruction you have been giving yourself — stop thinking about them — is the one thing known not to work. Deliberate thought suppression produces a rebound: the suppressed thought returns more often than it would have, and it arrives with the effort attached. So the target has to change. Not fewer thoughts. Less fuel.
What is usually going on
You are checking, and each check re-supplies the system. The profile, the last-seen, the mutual friend's photos. Every look is a small dose delivered to a craving circuit that was halfway through extinguishing, and extinction restarts from close to zero. This is why people who check twice a day genuinely are not getting better, and conclude they must be broken rather than that they are dosing.
You are replaying, and calling it processing. The conversation reconstructed with a better line in it, the argument re-run, the search for the moment it turned. Replaying feels like work because it is effortful and it is about something important. The test is whether it arrives anywhere: processing produces a sentence you did not have before, while rumination produces the same three scenes at a slightly higher resolution.
The evenings have nothing in them. The thoughts are not evenly distributed — they concentrate in unstructured time, which after a breakup is most of it, because a large part of your week used to be a person. Empty time is the substrate.
You are still narrating your life to them. A funny thing happens and there is a half-second where it is going to be told to them. This is the habit outliving the relationship, and it fades on its own timeline. It is not a sign, and it is not a decision you are making.
You are keeping the door open on purpose, quietly. Some of the thinking is maintenance of a possibility — rehearsing what you would say, keeping the account warm, imagining the reconciliation. If that is the case, the thinking will not reduce while it has a job. That question is better handled where it lives: whether to text them, and why you keep going back.
What works instead of stopping
Make checking expensive. Not willpower — friction. Mute, unfollow, archive, log out, hand the password to someone for a fortnight. Every serious account of this recovery has the same unglamorous centre, and the people who do it find the first four days awful and the second week noticeably quieter.
Give the replaying a container. If the loop will run, let it run at a set time: fifteen minutes, written down, same hour each day. Postponing a thought to a scheduled slot is a well-tested handle on rumination and it works better than banning it, largely because the loop turns out to be much less interesting when it has permission.
Change what the thought is followed by. You cannot stop the thought arriving. You can stop the sequence that runs after it — get up, go outside, put something on. The goal is that the thought stops being a doorway into the next ninety minutes.
Fill the specific empty slots. Not your whole life. The two or three recurring hours that are worst — Sunday afternoon, the commute, the hour after dinner. Structure targeted at the actual gaps does more than a general resolution to keep busy.
Do the boring identity work. The thoughts persist partly because your sense of yourself is still partly stored in them. The people you knew before, the thing you used to do, the sentence about yourself that has nothing to do with the relationship — that is the repair, and it is slower and duller than any insight.
Write one account and stop revising it. Much of the looping is an unfinished attempt to explain what happened — the mind keeps opening the file because there is no settled version. Writing a plain paragraph that says what the relationship was, why it ended and what you would do differently tends to close more of it than a hundred replays, and the reason is that a replay never produces a conclusion while a paragraph has to. It does not need to be fair to either of you. It needs to exist.
One warning about the friction stage: the first days after cutting the checking are usually worse, not better, which catches people out and sends them back. That spike is what extinction looks like from the inside, and it is the strongest evidence that the checking was doing something.
What to do tonight
- Mute and archive, tonight, before you talk yourself out of it. Not a statement to them — just distance between you and the dose.
- Pick your fifteen-minute slot for tomorrow and write the loop down when it comes. On paper it is shorter than it is in your head.
- Name the worst hour of your week and put something in it for the next three weeks.
- If the thought arrives at two a.m., get out of bed. The bed is a bad place to argue with your own head, and it teaches the night the wrong thing.
What it depends on
The loop is not equally available to everyone. Some people's systems settle into a plain sadness that runs out. Others get a monitoring process that treats the absence as an unresolved emergency, and an unresolved emergency does not stop being reviewed. That is the anxious pattern, and it has nothing to do with willpower. The looping half is measurable on its own terms too — overthinking behaves like a pattern with triggers rather than like a flaw in your character, which is why it answers to changed conditions and not to resolve.
Knowing which one you are running changes the priority. If the loop is the problem, the work is the container and the friction. If it is the absence itself, the work is regulation and people, and the thoughts quieten as a side effect.