Z Zeloku
Play today

Evidence-based · No hype

Sudoku as brain training — does it work?

A straight answer, with the studies. Then a five-minute daily routine that maximises the actual benefit.

Play today's Medium → Sudoku & memory loss

The short answer

Yes — but with a caveat you never see in brain-training app marketing. Sudoku genuinely improves specific cognitive skills: attention, working memory, and pattern recognition. Those improvements are real, measurable, and durable. What Sudoku doesn't do is raise your general intelligence, prevent dementia, or transfer meaningfully to unrelated tasks like language learning or driving skill. If a brain-training product claims otherwise, it's marketing, not evidence.

The studies

PROTECT (University of Exeter, 2019). 19,000 adults aged 50+. Regular number-puzzle players scored higher on tests of attention, reasoning, and working memory — with the effect equivalent to their cognitive age being roughly eight years lower. Correlational only, but a very large sample.

ACTIVE trial (Rebok et al., JAGS 2014). Randomised, controlled. Reasoning training in older adults produced benefits still detectable 10 years later. Sudoku wasn't the specific training used, but it exercises the same underlying skill.

What the studies don't show: A causal effect on general intelligence, prevention of Alzheimer's, or transfer to unrelated cognitive tasks. Anyone claiming those things is selling something.

A five-minute daily routine that works

  1. One puzzle a day, at the same time. Morning coffee, evening tea — pick a slot. Consistency beats duration in every cognitive-training study.
  2. Match the difficulty to your effort tolerance. If you're on autopilot, go up a level. If you're frustrated, go down. The right level makes you think — not sweat.
  3. Solve without pencil-marks once a week. Forces working memory to do the tracking. This is the closest Sudoku gets to actual brain-training research protocols.
  4. Learn one new technique a month. A new mental pattern is the highest-value cognitive investment. See the techniques library.
  5. Stop when it stops feeling good. Grinding is not training. If a puzzle stops being fun, save it, come back tomorrow. The pattern-recognition benefit accrues during rest, not during frustration.

Also see