The Week I Could Not Stop Puzzling

Last week I wrote about the psychology of not wanting to give up on a puzzle. Then I gave up on it. I started another one and could not put it down, to the point where the people in my house noticed before I did. This is the uncomfortable half of the same story, and I think it is worth telling even though I sell puzzles for a living.

Two weeks, one mechanism

Last week's piece was about whether it is OK to give up on a jigsaw puzzle. I was stuck in a large field of grass and flowers where every piece was green and most of them were lying to me. I wrote the article, and then I took my own advice and boxed it. I was not making enough progress to enjoy it. The small hit of placing a piece was arriving too rarely to outweigh the low-grade stress of not finding one.

So I opened the next one. It was the puzzle I had bought for our thirtieth wedding anniversary, a gift for my wife which, as she pointed out, was largely a gift for me.

I was gone. Not pleasantly absorbed, gone. Sitting at that table when I should have been somewhere else, in the room but not really in it.

Here is the part I did not expect to find when I went looking. Both weeks are the same variable. Mihaly Csikszentmihalyi's work on flow describes the state as a balance point between the difficulty of a task and your skill at it. Too hard and you get anxiety, which is exactly the grass and the flowers. Too easy and you get boredom. Land in the middle and you get flow, which he characterised by a loss of self-consciousness and a distorted sense of time.

Flow is not a side effect of a good puzzle. Flow is the product. It is what you are actually buying. Which is an awkward thing for me to notice, because the state I sell is the state that has no natural end.

Why a puzzle has no natural stopping point

Almost every other leisure activity has a seam built into it. A book has chapters. A match has a final whistle. An episode has credits, and the streaming services had to invent an autoplay countdown specifically to paper over the gap they knew you would use to stop.

A jigsaw has nothing. It is a continuous field of small wins with no structural pause anywhere in it. You never arrive at a moment that announces itself as a sensible place to leave off.

The chemistry does not help. The reward signal people loosely call the dopamine hit does not fire hardest when you place the piece. Wolfram Schultz's work on reward prediction shows the dopamine response tracking anticipation and uncertainty rather than the satisfaction itself. The searching is the engine. The placing is only the confirmation, and the instant it lands you are already scanning for the next one.

Add the pull of the unfinished task, which is well documented. A 2025 meta-analysis by Ghibellini and Meier found the effects behind that nagging sense of incompletion to be reliable, if modest in size. So you have a continuous activity, an uncertain reward on a fast cycle, and a standing urge to return to anything left undone.

The line that actually matters, and it is not hours

The most useful thing I read was not about puzzles at all.

In 2003 Robert Vallerand and colleagues published a model that splits passion into two kinds. Harmonious passion is something you have taken up freely and which sits in balance with the rest of your life. You control it. Obsessive passion is something you feel internally compelled to do, which takes up disproportionate space and starts conflicting with other parts of life. It controls you. Across a long run of studies since, harmonious passion tends to track with better outcomes and obsessive passion with worse ones, within the very same activity.

Read that last part again, because it is what surprised me. Same activity. The difference is not what you are doing, and it is not how long you do it for. It is whether you chose it in the moment, and what it is pushing out of the way.

By that measure the grass-and-flowers puzzle was fine and the anniversary puzzle was not, and the anniversary puzzle is the one I was enjoying. That is an unsettling result and I think it is the honest one.

It was not the puzzle, and I have the evidence

This matters commercially and I would rather say it out loud than have you wonder. If puzzling were the cause, the responsible thing for me to do would be to sell you fewer puzzles. It is not the cause. It is a well-designed absorbing activity, and it will find whatever tendency you already have and give it somewhere to go.

The part I could not see on my own

I did not notice. That is the whole problem.

I spent twenty five years managing money, and in my experience nothing in that job is harder than seeing your own bias from the inside. Nobody has ever felt themselves becoming overconfident. You find out from the record, or from somebody willing to tell you.

The same applies at a puzzle table. The information that I had gone missing for a week did not come from me. It came from the people closest to me, and my first instinct was to be a little defensive about it. That instinct is worth overriding. Criticism from the person who lives with you is the cheapest and most accurate data you will ever get about your own behaviour. It is free, it is timely, and it comes from someone with no incentive to be wrong.

There are two roads from there. One is to go looking for where it comes from, which is slow, occasionally useful, and in my case still unfinished. The other is to accept you may never fully find out and to build the guard rails anyway. They are not alternatives. The second one works while you are doing the first.

Rules beat willpower

Angela Duckworth and colleagues make a point in their work on self-control that lands exactly here: strategies that change the situation tend to beat strategies that rely on resisting temptation in the moment. Ariely and Wertenbroch found something similar with self-imposed deadlines. People who set their own binding constraints in advance do better than people who intend to behave well later.

Applied to a puzzle table, the decision has to be made while you are standing up, not while you are sitting down with eleven pieces of sky in your hand.

1. Set the end before the start. A time, a timer or a piece count. Anything external. Deciding that you will stop at a natural break is deciding nothing, because there is no natural break.
2. Stop mid-flow, not at a gap. This feels wrong, and it is the one that works. Leave it in the middle of a section you are enjoying, and you will come back to it happily tomorrow.
3. Make ending the session one physical action. This is why a board or a mat matters more than it looks. If closing the puzzle means closing a lid, you will do it. If it means abandoning the dining table mid-build, you will keep going.
4. Puzzle with someone, or near someone. A shared table is self-limiting in a way a solo table is not. It is also more fun, which is not a small point.
5. Let the household tell you. Ask, and then believe the answer. You are not the reliable narrator of your own attention.

I am doing all five. I will report back honestly on whether they hold, including if they do not.

What I am not saying

I am not telling you to puzzle less. I do not think most people have this problem, and the evidence on the activity itself is good. A 2018 study led by Patrick Fissler found that jigsaw puzzling draws on several distinct cognitive abilities at once, including visual perception, mental rotation and working memory, and the authors put it forward as a potential protective factor in cognitive ageing. It is absorbing, it is screen-free, and it is one of the few things left that a household will do together in the same room.

What I am saying is that absorption is powerful, and powerful things are worth putting a little structure around. Noticing it is not a reason to stop. It is a reason to run it on your own terms.

There is a coda that makes all this more interesting rather than less. Part of what hooked me on that anniversary puzzle was the cut. The pieces were genuinely unusual shapes rather than the standard grid, and it changed the activity rather than only the picture. I am writing about piece shapes properly next, because I did not expect the cut to matter as much as it did.

Common questions