The message arrives at 4:12pm. Slight change of plan, hope that’s okay. And it is okay, technically. Nothing has gone wrong. Nobody is upset with you. The change is small enough that saying anything about it would make you the problem. But the evening you had already built, the one you have been quietly running in the background since about ten this morning, has to come apart now and be built again from nothing. You say no worries! You mean it, mostly. You also know you will be tired in a way you cannot explain to anyone by seven.
Predictability is not a preference for you. It is infrastructure. When you know what is coming, your brain can stop running expensive background calculations and spend that capacity on the actual day: the conversation, the work, the person in front of you. Routine is what makes everything else affordable. What gets called rigidity from the outside is nearly always this from the inside, not a refusal to bend but a load-bearing structure doing its job. Remove it and nothing collapses immediately. You simply start paying for the same day twice, and you keep paying until you run out.
What the research shows
- Given a gambling task, 32 autistic adults preferred the less profitable decks with predictable outcomes over the more profitable ones with unpredictable outcomes, where 31 non-autistic participants did not. If you have ever taken the worse option because at least you knew what it was, that is the same choice. Macchia et al. (2024)1
- In 187 adults, insistence on sameness was significantly higher in the autistic group, and part of its link to depression and anxiety ran through rumination and obsessive thinking. The mediation was partial, and it held for autistic and non-autistic participants alike. Cooper & Russell (2024)2
- Across ten studies, intolerance of uncertainty and anxiety were associated at r = 0.62, roughly 38% of the variance in anxiety. Nearly all of that evidence comes from autistic children and teenagers, which is worth knowing: the adult research is still thin. Jenkinson et al. (2020)3
- In autistic children and adolescents, results were consistent with a model in which intolerance of uncertainty mediates the relationship between autism and anxiety, meaning the anxiety may run through the uncertainty rather than sit beside it. Boulter et al. (2014)4
Rigidity is a word from the outside
Notice who ever uses it. You have almost certainly never described yourself as rigid in your own head. The word arrives from somewhere else: a report, a performance review, a partner in the middle of an argument, a parent thirty years ago. It is a word for describing a person you are watching, not a word for being a person.
That matters more than it sounds. Rigid describes the outside of the behaviour and nothing else. It records that you did not move, and it stops there. It has no room in it for why, no room for what moving would have cost, no room for the fact that you had already moved four times that week in ways nobody noticed. It is a word that treats the not-moving as the whole event.
Predictability is the word from inside. It describes what you are actually doing, which is managing a load. And the two words can sit over the exact same moment, the same Tuesday, the same declined invitation, and produce completely opposite conclusions about who you are.
You have probably been living inside that gap for a long time. It is one of the specific injuries of being autistic and undiagnosed, or diagnosed late: your regulation strategies get filed as personality defects, and you file them that way too, because everyone around you agrees. You end up apologising for the one thing that was keeping you upright.
What predictability actually buys you
Think about what happens when you know exactly how a day is going to go. Not a good day, necessarily. Just a known one.
You are not modelling. You are not running the four possible versions of the meeting, the three ways the conversation could turn, the exits from a room you have not entered yet. You are not holding a contingency plan open in working memory in case the first one fails. All of that is switched off, and what it leaves behind is capacity. Attention that can actually go to the work. Words that arrive when you reach for them instead of ten minutes later. The bandwidth to notice that your friend seems a bit off today, which you genuinely cannot notice while you are calculating a train timetable.
That is the whole transaction. Predictability is not the goal. It is the thing you buy so you can afford the goal. When your executive function is already expensive, removing unnecessary decisions is not indulgence, it is the only way the necessary ones get made.
This is also why the same routine can look wildly different from two angles. Your morning routine looks, from outside, like an inflexible sequence of small unimportant actions. From inside it is the reason you can hold a full conversation at nine. The sequence is not the point. The absence of decisions is the point.
And when the predictable structure goes, the cost does not show up as distress in the moment. It shows up later, as overwhelm you cannot trace to anything, or a shutdown on a day that looked fine on paper, or the social hangover after an evening you actually enjoyed. Nothing bad happened. You just spent everything you had, and the receipt came late.
Why uncertainty costs you more
Here is the part that gets left out of most explanations, and it is the part that lets you stop reading yourself as difficult.
Brains work by prediction. Yours is constantly generating a model of what is about to happen and then checking reality against it, and the mismatch between the two, the prediction error, is what has to be processed. Ordinarily a lot of small mismatches get discounted as noise and never reach you at all.
One influential theoretical account of autism, published by Van de Cruys et al. (2014)5 under the title Precise minds in uncertain worlds, proposes that this discounting works differently for you. On their model, prediction errors get treated as consistently high-precision: more meaningful, more worth attending to, and not turned down when the situation calls for it. It is a model rather than a measurement, and it has not settled the field. But if it is right, the practical translation is worth sitting with: the world delivers more surprise to you than it delivers to the people telling you to relax about it.
So when someone says it’s only a small change, they are not lying. It is a small change, for them. Their brain sanded it down before it arrived. Yours may not have. On that account you are both standing in the same room processing different amounts of information, and only one of you is being asked to explain themselves.
This also explains why the load stacks. It is not that a changed plan is uniquely hard. It is that a changed plan lands on top of the sensory input you were already filtering, the conversation you were already decoding, and the day you were already tracking, and something has to give. The predictability you build into your life is you reducing the number of open channels, so that when one of them surprises you, it does not take everything down with it.
The research on autistic anxiety lands in the same place from a different direction, though it comes mostly from studies of autistic children and teenagers rather than adults. Uncertainty may not sit next to the anxiety so much as generate it, which makes a life with fewer unknowns in it something other than avoidance. It looks like a reasonable place to intervene.
Not all of it is coming from the routine
This one matters, because a great deal of advice aimed at you gets the target wrong.
The standard framing goes: your need for sameness is a symptom, symptoms are bad, therefore we should reduce your need for sameness, therefore let us practise tolerating disruption until you need it less. Underneath that sits an assumption nobody states out loud, which is that the preference itself is the whole problem.
Cooper and Russell’s study of 187 adults2 complicates that usefully. Insistence on sameness was higher in the autistic group, and it did track with depression and anxiety. But a substantial part of that route ran through rumination and obsessive thinking, and the same pattern showed up in the non-autistic participants too. Two honest caveats: the mediation was partial, so the preference did not come out of it entirely blameless, and the study cannot tell us which way the causation runs.
Still, it splits one thing into two, and that is worth having. Your need for a predictable Thursday is one thing. What happens in your head for the four hours after Thursday changes is another: the replaying, the rehearsing, the trying to work out whether you overreacted, the thirty-eighth pass through a two-minute conversation. Those are not the same problem, they do not respond to the same approach, and the second one is a great deal more workable than most advice assumes. None of that work requires you to stop wanting a predictable Thursday.
The wider mistake is a familiar one, the same one people make about repetitive behaviours generally. When ten young autistic people were asked directly about their own repetitive behaviour, rather than described by their parents, Joyce et al. (2017)6 found they gave a range of reasons for it. You have reasons too. You always did. Being asked is the rare part.
You have been flexible your whole life
Take the inventory honestly, because you probably never have.
You learned a second language of facial expressions that does not come naturally to you and you speak it daily. You changed how you sit, how long you hold eye contact, how much you say about the thing you actually care about. You have eaten in restaurants that hurt, attended events you had to recover from for two days, and said nothing about either. You have adapted to every workplace, every family gathering, every social norm designed by and for people who process the world differently than you do, and you did it continuously, unpaid, without being thanked, for decades.
That is not the behaviour of an inflexible person. That is masking, which is flexibility performed at a cost so high it eventually takes your health, and it is why so many of us arrive at autistic burnout genuinely believing we are lazy.
So when the flexibility finally runs out, and it does run out, what people see is the one place you stopped bending. They do not see the twenty years of bending that preceded it. They call the last five percent your rigidity and never once count the first ninety-five.
“I got called rigid for about fifteen years and not one person ever asked what it cost me to be the one who moved. Because I did move. Every single time. They just never had to watch me do it.”
— Autistic adult, HeyASD community
There is a specific grief in working this out at forty, or fifty, and realising how much of your self-image was built out of other people’s incomplete descriptions of you. It is not a dramatic grief. It arrives quietly, usually while you are doing something ordinary, and it tends to want somewhere to go.
Being read as rigid for decades, and only finding out much later that you were managing something invisible, is one of the quieter griefs of a late diagnosis. The Unmasking Years sits with that stretch of misread time, and with what it takes to stop apologising for the things that were holding you up.
When predictability narrows into a cage
This needs saying plainly, and it needs saying without anyone panicking.
Sometimes the structure does narrow. Not because you are doing it wrong, and not because the earlier sections were wrong, but because structure built during a hard period tends to stay in place after the hard period ends. A routine you built at your worst can quietly become the shape of your whole life. You end up with a life that is stable and also smaller than the one you wanted, and the smallness is hard to see from inside because everything in it is working.
The useful question is not am I being rigid. That question just imports the outside word and makes you the defendant. The useful question is: is this structure holding me up, or is it holding a fear in place?
Some honest ways to tell the difference:
- A supportive routine gives you back capacity. A protective one absorbs it. If keeping the routine takes as much effort as the thing it was meant to make possible, it has stopped paying for itself.
- A supportive routine survives being broken. It costs you a day, and then you rebuild it. A protective one produces dread out of proportion to the disruption, and the dread arrives well before the disruption does.
- A supportive routine has a reason you can still say out loud. A protective one is often held in place by something you have not looked at directly in years.
- A supportive routine leaves things you want inside your reach. A protective one has been quietly removing them, one declined invitation at a time.
If some of that lands, the move is not to blow the structure up. Forcing yourself through disruption to prove you can tolerate it is how people end up in burnout holding a completed exercise. The gentler and more effective route is to widen at the edges while the centre stays fixed: same time, different café. Same route, different day. Same meal, one new ingredient. You are not building a tolerance for chaos. You are collecting evidence that a small unknown can happen and the structure will still be there afterwards.
And if what you find underneath is the loop rather than the routine, the rumination that starts when things move, that is the thing worth taking to a therapist who understands autistic adults. Not the routine. The routine can stay.
Portable predictability
Most advice about routine assumes you control your environment. You often do not. Work changes, people cancel, the train is the train. So the version worth building is predictability you carry rather than predictability you depend on the world to supply.
Anchor points, not schedules. A schedule is a sequence and any break in it breaks the whole thing. Anchors are fixed points with slack between them: coffee in the same cup, the same twenty minutes before anyone speaks to you, the same walk at the end of the day. Three or four anchors will hold a day together even when everything between them collapses. This is the same instinct behind autistic nesting, making one part of the world reliably yours so the rest can be uncertain without taking you with it.
Information instead of control. You usually do not need the plan to stay the same. You need to know what the plan is far enough in advance to build it. Those are different requests, and the second one is much easier for other people to grant.
Name the unknown so it stops floating. An unspecified something might come up on Saturday occupies far more processing than a specific Saturday might involve two hours at my sister’s or nothing at all. Both are uncertain. Only one has edges. Written down, with its two or three possible shapes, an unknown stops running as a background process.
Build the transitions in. Most of what gets called difficulty with change is actually difficulty with the join between two things. A buffer either side of anything demanding is not padding. It is the part where the switch actually happens, and if you do not schedule it your body will take it later, usually at a worse time. This is also why travel is disproportionately expensive: it is transitions, all day, with no buffers.
Keep the known things known. On a week where a lot is uncertain, hold everything else absolutely still. Same food, same clothes, same route, no new anything. This is not regression. It is deliberately reducing the number of open channels so the one that matters has somewhere to land. Pair it with time in a special interest, which for most of us is the most reliably predictable environment that exists.
Decide once. Every recurring decision you convert into a default is one less thing running. Same breakfast, same order, same seat. People will find this funny. It is buying you an hour of function a day.
Asking for information instead of asking for flexibility
The reason accommodation conversations at work so often go badly is that you end up asking for the wrong thing. I need more notice of changes sounds, to a manager, like a request to be handled differently. It puts you in the category of people who need managing.
Reframe it as information flow and it stops being about you at all:
- Can we put the agenda in the invite so I come prepared?
- Could you flag schedule changes in the morning rather than as they happen? I’ll turn things around faster.
- What’s the actual deadline, not the comfortable one? I plan better against the real number.
- I’d rather know a plan might change than be told it definitely won’t.
None of these asks anyone to change what they do. They ask for earlier, clearer versions of what is already happening, which most workplaces will give you without a formal accommodation conversation at all.
“I stopped asking my manager to be flexible and started asking her to tell me things earlier. Same job, same chaos, nothing about the work changed. But I could see it coming, and it turned out that was the entire difference.”
— Autistic adult, HeyASD community
At home the same principle applies with softer edges. I’m not upset, I just need ten minutes to catch up with the change does an enormous amount of work, because it separates the disruption from the relationship. Most of the conflict around changed plans is not about the plan. It is about one person reading the other’s distress as disappointment in them.
And when someone calls you rigid to your face, you do not owe them a defence. But if you want one: I’m not refusing to change. I’m telling you what it costs and asking whether it’s worth spending today. Said calmly, that reframes the whole exchange, because it is true and because it is not an apology.
Key points
- Rigidity is an outsider’s word: it records that you did not move and has no room in it for what moving would have cost.
- Predictability is not a preference, it is infrastructure that frees up the capacity you spend on everything else in your day.
- Uncertainty may generate autistic anxiety rather than sit beside it, which makes a life with fewer unknowns a reasonable place to intervene rather than an avoidance.
- A substantial part of the distress attached to insistence on sameness runs through rumination, so the loop that starts when plans change is a separate and more workable target than the preference itself.
- You have been flexible your entire life at enormous cost; people only ever count the five percent where you stopped.
- Structure built during a hard period can outlast it, so it is worth asking whether a routine is holding you up or holding a fear in place.
- Portable predictability, anchors, advance information and built-in transitions, works even when you do not control your environment.
- Asking for earlier information is far more likely to succeed than asking other people to be more flexible.
Questions about autism and predictability
Why do autistic people need routine so much?
Because routine removes decisions, and decisions are expensive. When a sequence is known, your brain stops generating and checking predictions about it, which frees up capacity for everything that actually needs you: the work, the conversation, the sensory environment you are already filtering. It is not that the routine itself is satisfying. It is that the absence of unknowns inside it is what makes the rest of the day affordable. There is also a well-developed theoretical account, predictive coding, proposing that uncertainty carries more processing cost for autistic brains than for non-autistic ones. If that holds, the same unpredictable morning is not the same event for two people standing in it.
Is a need for predictability the same as rigidity?
No, and the difference is not cosmetic. Rigidity describes behaviour observed from outside: you did not move. It contains no information about why, or about what moving would have cost, or about how many times you already moved that week without anyone noticing. Predictability describes the same moment from inside, where you are managing a processing load. The two words fit over the identical Tuesday and produce opposite conclusions about your character. If you have spent years accepting the outside word as an accurate description of yourself, it is worth knowing it was never a complete one.
Why do small changes to plans upset me so much when nothing bad happened?
Because the size of a change and the cost of a change are unrelated. What you built when you made the original plan was a working model of the next several hours, and a small change does not edit that model, it invalidates it. You rebuild from scratch. Meanwhile you are usually also managing whatever else the day is throwing at you, so the rebuild lands on an already loaded system. The reason you cannot justify the reaction is that the reaction is not to the content of the change. It is to the rebuild.
Should I try to become more flexible?
Not as a goal in itself, and not by deliberately disrupting yourself to build tolerance. That approach treats your need for structure as the entire problem, and the research is more interesting than that: a substantial part of the distress tied to insistence on sameness appears to travel through rumination and repetitive thinking. Only part, and the study cannot tell us which way the causation runs. But it does mean the loop that starts after the ground moves is a separate target from the wanting of steady ground, and a far more workable one. Where you do want more room in your life, widening at the edges while the centre stays fixed works far better than forcing yourself through disruption to prove a point.
Why am I fine with change I chose but not change that happens to me?
Because you were the one holding the model. Change you initiate arrives with all its consequences already processed: you built the new version before you acted, so there is nothing left to rebuild. Change that arrives from outside skips that entire step and hands you the outcome without the working. This is why you can move house voluntarily and be fine, then be floored by a meeting moving forty minutes. It was never about the magnitude. It was about whether you got to do the processing in advance or were handed the bill afterwards.
How do I explain my need for predictability without sounding difficult?
Ask for information rather than flexibility. Can we get the agenda in the invite and could changes come through in the morning rather than as they happen are requests about how information flows, not requests to be treated differently, and most people grant them without noticing they granted anything. It also helps to attach a benefit rather than a need: you turn work around faster, you plan better, you come prepared. That is true, and it means the conversation never gets to the point where anyone has to decide whether you are difficult.
Can my need for predictability get worse?
Yes, and it usually means something else is going on rather than the need itself changing. Structure tightens when capacity drops, so a sharp increase in how much sameness you need is often the earliest visible sign of autistic burnout, a sustained stretch of masking, or a sensory environment that has quietly got worse. The instinct is to treat the tightening as the problem. It is more useful to read it as a gauge. Something is costing more than it used to, and your system is compensating by removing every variable it can still control.
What is intolerance of uncertainty and do I have it?
It is a measured tendency to find not-knowing distressing in itself, separate from whether the possible outcomes are bad. In autism research it turns up repeatedly and strongly: one meta-analysis found it associated with anxiety at r = 0.62, and other work suggests the relationship between autism and anxiety runs through it rather than alongside it. Both of those come mostly from studies of children and teenagers, so read them as a direction rather than a settled fact about adults. If waiting for an answer is worse for you than getting a bad answer, that is the pattern. It is worth knowing about mostly because it suggests where to aim: reducing unknowns may be a more useful target for autistic anxiety than trying to reason yourself out of the anxiety directly.
How do I keep any predictability when my life is genuinely unstable?
Stop trying to make the schedule predictable and make yourself the predictable part. Three or four anchors carried through the day, the same cup, the same twenty minutes alone, the same walk, will hold a week together when nothing in the calendar will. Keep every non-essential variable frozen while the essential ones are moving: same food, same clothes, same route, no new anything. It looks like a very small life for a while. It is not. It is you reducing the number of open channels so the one that actually needs your attention has somewhere to land.