Is your hot streak also your most disruptive streak?

There is a flattering idea floating around science-of-science, and it goes like this: your best work doesn’t dribble out evenly across a career. It arrives in a burst. A few golden years when almost everything you touch lands, and around them, before and after, the ordinary stuff.

That burst has a name. Liu et al. called it a “hot streak” in 2018: a bounded stretch in a creative career during which a person produces their highest-impact work. The unsettling part of their result was that the streak seems to strike at a random time. You cannot schedule it. You get one, if you get one at all. I find this idea seductive, and I distrust things I find seductive, so a new paper caught my attention.

What the study found

Hongkan Chen, Lutz Bornmann, and Yi Bu asked a narrow, testable question. Not “when does your best work happen,” but “when your best work happens, is it also your boldest?” (Humanities and Social Sciences Communications, 2025) Their answer, from the careers of 21,271 economists: yes. “Disruptiveness of scientific publications is higher during researchers’ hot streaks than in their other periods of careers.” During the streak you’re not only cited more. You are also more likely to publish work that shoves the field in a new direction instead of reinforcing the old one.

What “disruptive” means here

That word needs a definition, because it is doing a lot of work. Some papers consolidate what we already know; later work leans on them and on their predecessors together. Other papers disrupt: later work cites the new paper and quietly stops citing what came before, as if the new paper made the old scaffolding unnecessary. The disruption index puts a number on that difference. A high score means your paper eclipsed its own references. The claim here is that hot-streak papers score higher on it.

Two findings I did not expect

First, the boldness is random too. Just as hot streaks land at unpredictable times, so does disruptive work: “we also observe the randomness rule for the occurrence of disruptive work.” You do not graduate into disruptiveness at a fixed career age.

Second, and this is the one I keep chewing on: “researchers’ most disruptive years are oftentimes earlier than their most highly cited years.” Your boldest paper and your most-cited paper are usually not the same paper, and the bold one tends to come first. The recognition lags the risk, sometimes by years.

Where I would push back

I like this paper, and I don’t fully believe it yet.

It is economists. 21,271 of them is a large sample, but it is one field with its own citation culture. I wouldn’t bet the pattern transfers unchanged to molecular biology or to poetry.

The disruption index is contested. People argue about what it actually measures and how much it wobbles with database coverage and reference counts. Read “more disruptive” as “scored higher on one specific metric,” not as a verdict handed down by nature.

It is a correlation. Hot streaks and disruption travel together in this data. The paper doesn’t, and can’t, tell you that being disruptive triggers the streak, or that the streak makes you disruptive, or that some third thing drives both.

And one lever in the paper is not random at all: volume. “A higher publication volume of researchers increases the likelihood of entering hot streaks and producing disruptive publications.” Publish more, and you raise your odds of both. That is the least glamorous sentence in the paper and possibly the most useful.

Why care, if you are watching your own trajectory

If you’re staring at your own CV and wondering whether your good years are behind you or still ahead, this paper offers a small, cold comfort. The good years don’t seem to be something you earn on a schedule and then lose. They arrive when they arrive. What you can do is keep the tap open, because output feeds the odds. And if you’re sitting on a paper that feels riskier than your usual, the timing result hints that you might be earlier in a streak than the citation counts will admit for another few years.

The authors put the practical conclusion plainly, aimed at funders as much as scientists: “both scientists and funding agencies can assume the randomness of phases with important papers in scientific careers. But when these phases occur, impactful and disruptive papers can be expected.”

I’m glad they stopped there and didn’t try to sell me a recipe for manufacturing a hot streak. Because here’s what still nags at me. If the streak is random, and the disruption is random, and the two merely happen to move together, then the comforting story might collapse into something much plainer: sometimes a scientist has a good run, and during a good run good things cluster. Is that a law of creative careers, or is it just what any run of luck looks like from the inside? I’m not sure the data can tell those two apart. Are you?