Self-sabotage is a repeated pattern of behaviour that works against something you say you want, in a way that looks deliberate from outside and feels nothing like a decision from inside: leaving the essay until the last night, picking a fight the week after a good holiday, not applying for the job you are qualified for. The word itself carries a theory, that some hidden part of you is working against the rest, and that theory is both wrong and unusually cruel.
There is a better account, and it is more useful than the first one. Every one of these behaviours is doing a job, and the job is almost always protecting you from something that feels worse than the failure it causes.
What self-sabotage means
It means a pattern rather than an event. One missed deadline is a missed deadline; the fourth in a row, always on the work you care most about, is the thing the word describes, and the repetition is what separates it from ordinary bad luck.
Three things it is not. It is not a decision, because nobody experiences it as choosing to lose. It is not laziness, which explains none of the cases where the behaviour appears only on high-stakes work and never on the trivial kind. And it is not a secret wish to fail, an idea with a long popular life and no research behind it.
That last one is where the damage is. If the explanation is that part of you wants you to lose, only two responses are available: try harder, and think worse of yourself. The first has usually been tried. The second reliably makes the behaviour more frequent, because shame is one of the feelings these strategies exist to keep away, so the interpretation feeds the thing it is trying to explain. This is the same loop we describe in why people-pleasing is not the same as kindness, running on a different behaviour.
What follows are four mechanisms, in descending order of how well each one is evidenced. They overlap, and most people recognise more than one.
Self-handicapping: giving failure a cause that is not your ability
Self-handicapping is creating an obstacle in advance, so that if you fail there is something other than your ability to point at. Leave the essay to the last night and a poor mark means no time; start three weeks out, work carefully and still get a poor mark, and there is nothing standing between you and the result.
It was demonstrated in an experiment in 1978. Participants worked on problems and were told they had done well, but for half of them the problems had been unsolvable, so the praise arrived without any sense of why. Facing a second round, those participants were the ones who chose a drug they had been told would impair their performance. [berglas-jones-1978] They were not protecting the result. They were protecting an explanation, and they were willing to damage the result to keep it.
Note what that study did not establish. It was a small laboratory experiment with undergraduates, the effect appeared in the male participants rather than across the sample, and a drug choice is not an essay deadline. The wider evidence comes from field research: a meta-analysis pooling 36 studies and about 25,550 students found academic self-handicapping associated with modestly worse achievement, a correlation of around -0.23. [schwinger-2014] That is a real relationship rather than a large one, and it is correlational, so it does not settle which way the arrow points.
The everyday forms are recognisable. Starting far too late. Going out the night before something that matters. Taking on so much that nothing can be done properly. Announcing loudly that you have not prepared, which handicaps the impression rather than the performance and works just as well. Choosing the option you are less likely to be judged on.
The uncomfortable part is that it succeeds. The belief about your ability survives intact, and the price is the outcome.
Procrastination: repairing the mood now, paying later
Procrastination is short-term mood repair rather than poor scheduling. The task produces a feeling, usually boredom, dread or self-doubt, and avoiding the task removes that feeling immediately, while the consequence lands on a version of you who does not exist yet. [sirois-pychyl-2013]
That is why time-management fixes so often fail. A better calendar addresses the scheduling problem of somebody who did not have one, and does nothing about a feeling. We cover this at length in why procrastination is not a time-management problem, including what does help; here it matters mainly as one mechanism among several, and as the one most likely to be running alongside the others.
Self-sabotaging a relationship: why it gets worse when things go well
Withdrawing from a relationship that is going well is a protective strategy rather than a contradiction, because leaving first hurts considerably less than being left. It intensifies at exactly the moments an outsider would call good news, since closeness is what raises the stake.
Attachment research describes this as a deactivating strategy: a set of moves that reduce how much you need somebody, which develop when needing people has previously gone badly. [mikulincer-shaver-2016] They are not conscious tactics. They feel like accurate perception, which is what makes them so hard to argue with from the inside.
The concrete forms:
- Picking a fight in the days after a good weekend, an “I love you”, a holiday or moving in together.
- Going quiet rather than escalating, so the distance arrives without an argument to point at.
- Finding a flaw that was entirely tolerable a month ago and is suddenly disqualifying.
- Keeping an exit visible, whether that is a conversation with somebody else, an unmade plan, or a private list of reasons this was never going to work.
- Testing, usually by withdrawing and watching whether the other person follows.
The internal logic is clean, which is why it survives so long. If I end this, I chose it. If they end it, it happened to me.
Two things this does not mean. It does not mean the relationship is wrong, and the strategy is at its most confusing precisely when the relationship is good. And it is not a character defect: avoidance of this kind usually developed somewhere it was the sensible response, and it has simply outlived the situation that produced it.
The most useful single move is to name it to your partner while things are calm, rather than during the fight. Unexplained withdrawal is far more damaging than explained withdrawal, because the person on the receiving end has to invent a reason, and the reason they invent is almost always about themselves.
The ceiling problem: when success raises the bar you then have to clear
The fourth pattern is about what a success costs you next. Do well once and you have set a level you are now expected to hold, so failing at a lower level can feel safer than succeeding and then failing from a height.
It shows up as the promotion that becomes next year’s baseline, the good draft that every later draft is measured against, the strong first term that turns into an expectation. Nobody decides this. It runs as a quiet arithmetic about exposure, and it is often mistaken for modesty.
Be careful with this one. It is the least well evidenced of the four: it is described consistently in clinical settings, but the experimental literature on “fear of success” from the 1970s did not replicate well and the idea is no longer treated as a stable trait. Treat it as a description that many people recognise rather than as a finding, and if it does not fit you, it does not fit you.
Which pattern do you recognise?
Tick anything true. This is a reflection prompt rather than a test, and it produces no diagnosis.
0 of 6 ticked
When several of these run together, the common thread is usually a single thing you are protecting: most often what a real failure would say about you, once there is nothing left to blame it on. That is the thing worth working on, rather than any of the behaviours individually. The last item matters most, because knowing and continuing is the clearest sign the behaviour is doing a job, not that you lack willpower.
Two of these together usually share a trigger, and the fastest way to find it is to ask what each one would have exposed if you had not done it. Write the answer as a plain sentence rather than as a category, so "if I try properly and it is average, I find out I am average" rather than "fear of failure". The specific version is the one you can actually test.
Nothing here matched, which is worth knowing before you take on a label that does not fit. If somebody has described you this way and you do not recognise it, ask them for the specific occasions rather than the summary.
Why trying to stop does not work
Because the behaviour is load-bearing. Remove it and you have not removed the threat, you have removed the protection, which is why the pattern tends to return or to be swapped for a different one within weeks.
A schematic of the trade-off described in this article, not measured data.
The point of putting those side by side is that the two columns are roughly the same size. If the protection were small, willpower would be enough and most people would have solved this already.
This is what the person who forces themselves to start three weeks early discovers. They sit down, the work is now genuinely exposed, the anxiety they had been avoiding arrives on schedule, and they stop. Nothing went wrong with their discipline. They removed the buffer while the thing it was buffering was still there.
The question that actually helps
Not “how do I stop”, but “what is this keeping me safe from”. That reframing does most of the work, because it turns an accusation into an investigation, and because it points at the part that can change.
A workable sequence:
- Write the fear as a specific sentence. Not “fear of failure” but “if I prepare properly and it is still average, I find out I am average.” Categories cannot be tested; sentences can.
- Ask what would actually follow if that sentence were true. Usually the honest answer is smaller and more bearable than the version you have been avoiding, and occasionally it is not, which is also worth knowing.
- Make the protection unnecessary rather than forbidden. Lower the stakes deliberately on one occasion: a draft explicitly labelled rough, an application you tell nobody about, a conversation with your partner about the withdrawal before it starts.
- Test it once, small. This is a behavioural experiment, the central technique in cognitive behavioural therapy, and it is worth asking for by name.
For the relationship pattern specifically, ask for attachment-informed therapy or emotionally focused therapy, both of which work on the strategy rather than on the arguments it produces.
Two related things sit underneath most of this. Our guide to where self-worth comes from covers why worth that rides on performance is the version that does the damage, and what the evidence on imposter syndrome actually shows covers the case where the fear is specifically about being found out. If what you recognise is mostly about what a failure would say about you as a person, the free self-esteem screener uses the Rosenberg scale, takes about two minutes and scores in your browser.
When to seek help
Speak to a doctor or a therapist if the pattern has cost you something you cannot easily recover, such as a job, a course or a relationship, or if it has been running for years across several areas of life rather than showing up in one. Describe the specific occasions rather than using the word self-sabotage, because the occasions are what an assessment works from.
Go sooner if low mood, anxiety or difficulty concentrating is sitting underneath it, since treating those often does more for the behaviour than working on the behaviour does. Go sooner too if the self-criticism afterwards is severe, or if you have started avoiding opportunities in advance to spare yourself the whole cycle.
Contact your local emergency services or a crisis helpline if you feel unsafe or have thoughts of harming yourself.
How MyFreud can help
MyFreud is a mobile app that helps you find solutions to problems that have affected your mind and productivity. Live coaching sessions give you somewhere to work out what a particular behaviour is protecting, which is the question the rest of it turns on, and each session ends with an actionable plan rather than encouragement to believe in yourself. Daily tracking shows the pattern you cannot see from inside it, including the timing that gives the relationship version away. The notepad is where the fear gets written as a specific sentence, which is harder to do in your head than it sounds.
Download MyFreud and start today: App Store or Google Play.