Do brain games actually work?
Short answer: yes for some things, no for others, and the industry has spent years blurring the line. Here's where we stand — with the receipts.
Why we don't make big promises
For years, brain-training was sold on sweeping claims: play a few minutes a day and you'd delay memory loss, guard against dementia and Alzheimer's, and sharpen your performance at school and work. The trouble is that the science never supported promises that broad — and a lot of people felt misled by them.
That gap between hype and evidence is the reason CortexCrunch exists in the shape it does. We treat it as a design brief: build something genuinely fun and measurable, and never, ever sell it as medicine.
What the evidence actually supports
The research community draws a critical distinction between two kinds of improvement:
- Near transfer — getting better at the trained task and very similar ones. This is real and reliable. If you practise a working-memory task, your working-memory-task scores go up. If you drill reaction time, you get faster at reaction time. Recent meta-analyses continue to find measurable gains in processing speed, attention and short-term memory from cognitive training.
- Far transfer — those gains spilling over into unrelated abilities like general intelligence, reading or everyday problem-solving. Here the evidence is weak and hotly contested. Reviews repeatedly find no convincing proof that brain training makes you broadly smarter.
In plain terms: brain games make you better at brain games. That improvement is genuine and worth something — it's just much narrower than the marketing of the 2010s implied.
So why play at all?
Because "narrow" isn't the same as "worthless." There are honest, defensible reasons to spend five minutes a day here:
- Measurement. A consistent reaction-time or memory-span number is a useful personal baseline. You can see the effect of sleep, caffeine, stress or a long day — real, immediate feedback about your own state.
- Engagement and enjoyment. These are well-designed little puzzles. Enjoying a mental challenge is a perfectly good reason to do it, the same way you might do a crossword or Wordle.
- Focus practice. Tasks like Stroop and N-back are demanding. Practising sustained, effortful attention is a reasonable thing to do — we just won't pretend it cures anything.
What we will never claim
To be completely clear, CortexCrunch does not:
- treat, prevent, diagnose or delay any medical or cognitive condition;
- protect against dementia, Alzheimer's, ADHD, or age-related decline;
- raise your IQ or make you "smarter" in any general sense;
- replace medical advice — if you're worried about your memory or focus, talk to a doctor.
About our percentiles
The percentile bands we show after each game are estimates, built from published distributions for tasks of this type — not clinical norms calibrated to your age, device or population. They're there to give your score a friendly sense of scale and something to beat, nothing more. Your raw numbers (milliseconds, tiles, points) are the honest measurement; the percentile is just context.
The sources
If you'd like to read the evidence yourself, good starting points include the 2014 consensus statements from the cognitive-science community (where one group of scientists endorsed brain training and a larger group of neuroscientists publicly pushed back), and recent meta-analyses on cognitive training in healthy adults and older adults. We'll keep this page updated as the science evolves — honestly, in both directions.
Now go have some fun
No promises, no pressure. Just six good games and your own honest numbers.
Play the daily set →