A racing mind at bedtime often has a storyline. You replay a conversation. Plan tomorrow. Solve a problem. Rehearse something that might happen. Return to the same worry for the tenth time.
The Cognitive Shuffle takes almost the opposite approach. Rather than trying to force your mind to become blank, it gives it something deliberately disconnected to think about.
The technique was developed by cognitive scientist Luc Beaudoin at Simon Fraser University and is more formally known as Serial Diverse Imagining. The idea grew from Beaudoin's theory of sleep onset: certain forms of mental activity - planning, rumination, problem-solving - are poorly suited to falling asleep, while the transition toward sleep is often accompanied by increasingly fragmented and loosely connected mental imagery. Cognitive shuffling attempts to deliberately create something closer to that second state, rather than a single connected story.
Beaudoin's do-it-yourself version uses a neutral seed word such as BEDTIME. Start with B and think of different things beginning with B, briefly picturing each one. When you become bored with that letter or can't easily think of another image, move to E, then D, and so on. The images are deliberately unrelated - you're not supposed to build a story connecting them.
There is some experimental evidence, though considerably less than the technique's online popularity might suggest, and it's worth being precise about exactly what it shows. In 2016, Beaudoin and colleagues studied 154 university students who reported excessive cognitive arousal before sleep - 137 of them women - randomly assigned to Serial Diverse Imagining, Structured Problem-Solving, or both together. Importantly, the version tested wasn't quite the DIY word-spelling method described above: participants used an app that played concrete words one at a time, about eight seconds apart, and were asked to form and hold a mental image of each word before the next arrived.
After one week and/or one month, cognitive and physical pre-sleep arousal had decreased, sleep effort had decreased, and self-reported sleep quality had improved compared with baseline, performing about as well as Structured Problem-Solving on those measures.
That's encouraging, but there are real reasons not to overstate it. The study involved university students, not a broad population with clinically diagnosed insomnia. It relied on subjective measures. There was no untreated control group. And although sleep diaries were collected, the published results didn't report that data - so the study doesn't establish that participants actually fell asleep faster by any specific number of minutes, only that they felt less aroused and rated their sleep as better. The findings were published as a conference abstract rather than a full peer-reviewed clinical trial, meaning the methods and results haven't yet been scrutinized to the same depth most other citations in this library have been.
So the evidence here is genuinely promising, not established. There's a plausible theory behind the technique, and real preliminary human evidence supporting it - but not yet the kind of large, independently replicated evidence base CBT-I has built up.
For someone whose main obstacle to falling asleep is a mind that keeps planning, analyzing, or rehearsing, that may still make it a sensible, low-effort technique worth trying. You're not trying to win an argument with your thoughts - you're giving the mind something harmless and fragmented to do until coherent thinking loses its grip.
