Do Brain Training Games Actually Work? What the Research Says

The question comes up every time: a friend shares a "brain training" ad, a parent asks about kids on tablets, or you catch yourself opening a game wondering — is this actually doing something, or am I just killing time?

Short answer: both. The research splits cleanly into near transfer (improving at the exact thing you practice) and far transfer (improving at life in general). One is real and reliable. The other, despite a decade of hype, still is not.

The big ACTIVE study and what it actually showed

The largest randomized trial in this area is ACTIVE — Advanced Cognitive Training for Independent and Vital Elderly. Its initial results appeared in JAMA in 2002. Researchers assigned 2,802 adults aged 65–94 across six US cities to one of three training groups (reasoning, speed of processing, or memory) or a no-contact control group.

The speed-of-processing group was the standout: at the five-year follow-up, also reported in JAMA in 2006, that group reported slightly less decline on everyday instrumental tasks — noticing things quickly, managing household paperwork, reacting in traffic. The effect size was small, and it is worth being precise about what it was: not "reversing dementia," closer to "staying slightly sharper at noticing things." Still meaningful when you are 78.

What a meta-analysis of 20+ studies confirmed

In a 2013 meta-analysis published in Developmental Psychology, Monica Melby-Lervåg and Charles Hulme reviewed 23 working-memory training studies involving roughly 600 participants — the review that famously poured cold water on claims of "raising IQ." The result: performance on the trained tasks improved reliably, and the gains lasted weeks or months. IQ did not move. Fluid intelligence did not move. Attention in unrelated tasks showed mixed or zero effects.

Important nuance: this is not the same as saying "brain games do nothing." They do exactly what consistent practice of anything does — they build skill at that thing. The question is whether that skill matters to you.

So what clearly works — and for whom

• Older adults worried about cognitive lapses: speed-of-processing training of the kind ACTIVE tested showed real near-term value on the trained tasks and small functional benefits.
• Students and knowledge workers: working memory drills (like our memory digit-span game) reliably improve your performance on the trained tasks themselves — a genuine near-transfer gain that helps when you read dense papers or juggle information in the moment. Whether it raises capacity for everyday life is less certain.
• Anyone whose job demands visual scanning: controlled visual search is exactly the skill Schulte grids and reaction drills train, so they are a sensible practice tool for dispatchers, radiologists, first responders and gamers. Formal evidence of on-the-job improvement in those specific professions is still thin.

What does not happen — and why you should ignore the ads

There is no convincing body of evidence that an off-the-shelf brain training program permanently raises IQ, prevents dementia, or cures ADHD. Individual studies have hinted at protective effects — for example, long-term follow-ups of speed-of-processing training reported lower dementia risk — but those findings remain debated and do not justify the advertising claims. What ads do is cherry-pick: they show the near-transfer gains (which exist) and imply far-transfer gains (which have not been reliably shown).

The transfer problem is real and stubborn. Your brain is spectacularly good at getting better at the specific patterns it encounters repeatedly. It is much less good at moving that improvement to new patterns. A person who can do 25 digits in a memory game is not automatically better at remembering names at a party — unless they practice that specifically too.

How to actually get transfer

• Train the component skill, then use it immediately: a Schulte round right before proofreading a spreadsheet means your trained visual search is fresh when you need it.
• Vary the formats deliberately. Our grid shapes — square, circle, triangle, hexagon — force flexible scanning instead of memorizing one layout. Flexibility is what makes transfer possible.
• Keep escalating difficulty. Improvement happens at the edge of ability. If you always play the same 5×5 grid you stopped learning months ago.
• Sleep more. Most consolidation of trained skills happens during deep sleep. Playing a round and then pulling an all-nighter is self-sabotage.

Our honest take

We do not sell the "20 minutes a day raises your IQ" story. Free, short, varied games used as a pre-work warm-up and a daily habit? That is legitimate cognitive maintenance. It is the equivalent of 10 push-ups before you sit down — not a gym membership, but not nothing either.

Just do not mistake a leaderboard ranking for a life outcome. The person who can do the circle grid in 12 seconds is the person who practiced the circle grid a lot. They are not automatically smarter, more attentive in meetings, or less likely to forget their keys — unless they also practiced those things.