Mahjong trains memory the way a gym trains muscle: through repetition under load. Every hand asks you to hold a shifting thirteen-tile shape in mind, keep a running tally of what has already been discarded, and rebuild your plan the moment someone claims a tile. That is a real cognitive workout rather than a cure-all, and the honest evidence picture is more interesting than the usual claims about ancient games and sharper brains.
What one hand demands of your brain
From outside the table, mahjong looks unhurried. Inside a player’s head, four separate memory jobs run in parallel from the first draw to the last.
Working memory: the shape never sits still
You are not memorising a picture. You draw a tile, compare two possible futures for your hand, and throw one of them away — then the tile you needed appears in somebody else’s discard and the calculation restarts.
Experienced players survive this by chunking. Instead of tracking every tile as a separate object, they hold a handful of groups (a set here, a part-run there, a pair, a couple of loose tiles) and that is a far lighter load to carry from turn to turn.
Recognition memory: reading the discard pond
The tiles lying face up in the middle are public information, but only for players who have been paying attention. Knowing that three of the four West winds are already gone changes whether you keep the fourth, and that knowledge is pure recall.
Prospective memory: remembering your own plan
The hardest discipline in mahjong is carrying out a decision you made three turns ago. You judged the hand unwinnable and chose to play defensively; holding that line while a tempting tile arrives is prospective memory plus impulse control.
Japanese riichi mahjong turns this into a rule of its own: furiten, which makes what you have already thrown away part of how the hand can be won. The exact conditions are set out in Wikipedia’s entry on Japanese mahjong and are worth reading before you play that variant. The point for memory is the same in any tradition — your own discard pile is information, and keeping it in mind is part of playing the hand rather than an afterthought.
| Moment in the hand | What you are doing | Memory system exercised |
|---|---|---|
| Sorting your opening tiles | Grouping into pairs, runs and partial sets | Chunking and working memory |
| Choosing a discard | Comparing two possible hand shapes | Working memory and mental simulation |
| Calling pung or chow | Recognising a match in under a second | Recognition and pattern matching |
| Scanning the discards | Counting what has gone and who threw it | Visual-spatial and source memory |
| Folding a losing hand | Holding one plan across many turns | Prospective memory and inhibition |
| Scoring at the end | Recalling fan values or named hands | Long-term retrieval |
What the research supports, and what it does not
Mahjong is unusual among leisure games in having been tested rather than merely praised. Sheung-Tak Cheng and colleagues in Hong Kong assigned 62 older people with mild-to-moderate dementia to play mahjong either twice or four times a week for 16 weeks, and reported gains across every cognitive measure they used, the largest of them on digit-span memory, in the International Journal of Geriatric Psychiatry in 2006. Twice a week performed about as well as four times, and the gains were still measurable a month after the sessions were withdrawn.
The same research group later ran a cluster-randomised comparison in Hong Kong nursing homes, published in the American Journal of Geriatric Psychiatry in 2014: 110 residents, in homes assigned to mahjong, tai chi or simple handicrafts as the control, three times a week for 12 weeks. Mahjong and tai chi both preserved functioning or delayed decline in some cognitive domains, including among residents who were already significantly impaired. Both papers are indexed in PubMed if you want to read the abstracts yourself.
Two caveats deserve equal billing with the findings. Both were modest studies in care-home populations, so the effect sizes do not transfer neatly to a healthy adult playing on a Friday night. And neither tested prevention: every participant already had a dementia diagnosis when the study began, so nothing here shows that mahjong keeps dementia away. Treat the game as worthwhile mental exercise, not as a treatment or an insurance policy.
The wider picture is consistent and worth stating plainly: mentally demanding, socially engaged activity is repeatedly associated with healthier cognitive ageing. Mahjong happens to bundle three ingredients that most solo puzzles lack (hidden information, live opponents, and a reason to turn up every week) which is the argument made in more detail in our comparison of mahjong versus brain training apps.
Five drills that turn a casual game into memory practice
- Name the shape before you discard. Say the grouping in your head — the sets you have, the runs you are building, the pair, the loose tiles. Naming forces chunking rather than staring.
- Track one tile family. Pick the honour tiles (winds and dragons) and count copies as they appear. Each exists four times, so the arithmetic stays manageable while the habit forms.
- Rebuild the pond from memory. When the hand ends, before anyone shuffles, try to name which player discarded each honour tile. Check yourself against the table.
- Declare your wait out loud. Once you reach tenpai (one tile away from a complete hand) state which tile you need before you look. It converts a vague feeling into a checkable claim.
- Play one hand unsorted. Leave the tiles in the order you drew them. Removing the visual crutch of a tidy rack is the single fastest way to notice how much of your play was sorting rather than thinking.
Why a physical table asks more of you than a screen
Digital mahjong is excellent for volume. Hands arrive back to back, with no shuffling, no wall to build and no scoring to slow the table down, so an app will deal you far more practice per hour than a club night can. What it removes is exactly the part that loads memory hardest: the app sorts your tiles, highlights legal calls, counts the score, and remembers the discards for you.
At a real table nothing is highlighted. You have to notice that the player opposite stopped discarding bamboos six turns ago, which is the same skill described in our guide to reading opponents in mahjong, and one of the habits that separates steady players from the errors covered in common beginner mistakes.
Playing across generations
The cognitive case and the social case are the same case. A weekly game that mixes grandparents, parents and teenagers gives the oldest player conversation and structure and the youngest player patience, which is why mahjong keeps turning up as a genuinely all-ages game.
Practical notes for family sessions: children from about five can take part with an adult supervising, a session can be cut to a single East round instead of a full game, and beginners do better learning the draw-and-discard rhythm for a few hands before scoring is introduced at all.
A physical set is what makes any of this practice rather than theory, and a family that plays three or four times before deciding it is not for them has bought an expensive box. We hire tiles, tables and stools across Greater Sydney for exactly that reason, with collection at Brodie Spark Drive, Wolli Creek, and every set cleaned before it goes out again. Rates by duration are on the hire page if you want to run a few sessions before committing to a set of your own.
What players ask about mahjong and the brain
Does playing mahjong improve memory in everyday life?
Transfer is narrower than most headlines suggest. The reliable gains sit close to the game itself (faster pattern recognition, better tracking of what has already happened) while general-purpose memory improvement is not something any single game can promise.
How often do you need to play to get anything out of it?
The Hong Kong trials ran sessions two to four times a week for three to four months, and in the 2006 study twice a week produced much the same result as four times. Consistency over months, rather than an intense fortnight, is the pattern worth copying.
Is mahjong better for the brain than sudoku or crosswords?
No study has run that head-to-head comparison properly. What mahjong adds is hidden information, live opponents and a social commitment, and those three factors are the ones most consistently linked to healthy cognitive ageing.
Can someone already living with dementia still play?
Often yes, with adaptations: simplified scoring, fewer rounds, a helper managing the wall. The Hong Kong studies by Cheng and colleagues ran exactly this kind of structured session in care homes, though decisions for an individual belong with their clinicians rather than a blog.
Which variant is the most mentally demanding?
Riichi is the usual answer, because defence, furiten and dora awareness add layers on top of hand building. Hong Kong style rewards speed and reading, while American mahjong loads pattern memory through its annual card of legal hands.
Do I need my own set to practise properly?
Not to start. Apps are fine for drilling tile efficiency, but discard tracking is a different skill on a physical table, so borrowing or hiring a set for a few sessions is the cheaper way to find out whether the game sticks.
Looking for a set of your own? See our mahjong merchandise, or rent a full setup to try one first.
More in Information, Lifestyle & Social