Real Work Doesn’t Move in Straight Lines
3 August 2026
Every plan you have ever drawn is a line: a bar filling, a burndown descending, a percentage climbing to one hundred. And every real project you have ever finished got there some other way. That gap is not a discipline problem. It is one of the best-documented facts in the psychology of work.
Start with the oldest result. In 1994, students were asked to predict when they would finish their theses. Their average guess was thirty-four days; reality took fifty-five. Asked for a worst-case estimate, assuming everything went as badly as it possibly could, they still undershot the truth. Kahneman and Tversky named the mechanism the planning fallacy: we forecast from an imagined smooth version of the task, and the smooth version does not exist. Douglas Hofstadter compressed it into a law that eats its own correction: it always takes longer than you expect, even when you take Hofstadter’s Law into account.
The modern data is harsher. Bent Flyvbjerg’s database of sixteen thousand projects shows overruns are not polite noise around a plan; they are fat-tailed, which means the plan-shaped line was never a description of the work. It was a wish drawn with a ruler.
The middle is supposed to look like that
What does real progress look like instead? Designers draw it as a scribble that slowly resolves into a line. Founders call the long flat stretch after the exciting start the trough. Research on teams found something stranger and more precise: projects hold their first approach through a long stretch of apparent inertia, then reorganize in a burst, almost exactly at the midpoint, regardless of how long the project is. Learning works the same way when you measure one person instead of averaging a crowd: plateaus, then jumps. Even the dip is functional. Getting better at almost anything requires abandoning a method that works for one that will work better, and performance falls before it leaps.
None of this is dysfunction. The wandering middle, the plateau, the step backward before the step up: that is what progress is actually made of. The straight line was the fiction all along.
Tools that lie with geometry
Now look at what our tools draw. A progress bar asserts a known distance to a known end: software engineers have a joke about that. The first ninety percent of the work takes ninety percent of the time, and the last ten percent takes the other ninety. Project researchers found the same thing without the joke: projects report “ninety percent complete” for months because the metric cannot see rework coming. The bar is not measuring the work. It is measuring the plan’s opinion of itself.
And the lie has a cost, measured carefully. The largest study of motivation at work, twelve thousand diary entries, found the strongest driver of a good working life is visible progress on meaningful work, and its dark twin: setbacks hurt two to three times as much as progress helps. A bar that slides backward, a date that slips, a line that flattens: each one manufactures a setback out of geometry, taxing morale for information that was never accurate in the first place.
What honest progress looks like
A map makes a different set of promises. Structure accretes: branches grow, and nothing renders backward. Small completions stay visible as themselves, a ring closing on a node, rather than dissolving into a rounding error on a bar. A branch that stalls just sits there, holding its shape, looking exactly like what it is: a part of the territory that is waiting. A branch you deliberately park is a decision you can see, not a failure the tool nags you about. And the whole picture answers the only progress question that matters, which was never “what percent,” but where are we, and what’s next?
There is even a finding about which way to point the camera. People speed up when they can see how far they have come; loyalty-card experiments doubled completion just by showing customers their position differently. Perceived position drives effort. Which means the honest representation is not a smaller lie drawn more carefully. It is a different shape entirely: show the territory, mark what is done, and let the line go wherever the work actually goes.
Real work doesn’t move in straight lines. Your map shouldn’t either.