The uncomfortable truth about brain training
An enormous industry rests on a simple promise: practise these games and you will become generally smarter. The evidence does not support that promise, and it is better to say so on the first line than to bury it. What repeated, well-designed studies find is that people get markedly better at the trained task, somewhat better at tasks that look almost identical, and no better at anything else. The term for this is near transfer without far transfer, and it has survived a great deal of hopeful re-analysis. A 2016 review of the field put it bluntly: the improvements are real, and they stop at the edge of the exercise.
So what is training actually good for?
Three things, all of them worth having. First, measurement: a weekly benchmark tells you how your sleep, stress and workload are affecting you, which is genuinely useful information about your life rather than your brain. Second, skill: mental arithmetic, typing speed and reading speed are trained skills, and training them works exactly as advertised because there is no transfer being claimed. Third, the habit itself: ten focused minutes is a small, pleasant, measurable commitment, and the people who keep it up tend to be doing other things right too.
A ten-minute daily routine
| Day | Exercise | Why |
|---|---|---|
| Every day | Reaction time, 5 rounds | A one-minute baseline that reveals fatigue |
| Mon / Thu | N-back, then sequence memory | Working memory under load, then span |
| Tue / Fri | Symbol search, then math test | Processing speed and arithmetic fluency |
| Wed / Sat | Visual memory, then chimp test | The spatial side, which most people neglect |
| Sunday | One full reasoning test | A longer sitting to check the whole picture |
What actually improves cognition
This is the part the industry says less about, because none of it can be sold as a subscription. Aerobic exercise has the best evidence of anything studied, with effects on attention and processing speed that survive replication. Sleep is not a lifestyle detail but the single largest short-term variable in every measurement on this site. Treating hearing loss, managing blood pressure and staying socially engaged all have better evidence behind them than any puzzle game. Learning something genuinely new — an instrument, a language, a craft — produces broader gains than any amount of practice on a task you have already mastered.
How to train without fooling yourself
- Fix the conditions. Same device, same time of day, same quiet room, or you are measuring your environment.
- Track a range, not a record. Your best-ever score is noise; your typical score across a week is signal.
- Stop after three attempts. Fatigue makes attempt four worse and tells you nothing.
- Expect a plateau. Most of the improvement on any task arrives in the first week and then stops.
- Do not read a rising score as rising intelligence. You are getting better at the game, which is a smaller and more honest claim.
Where to start today
Take reaction time and n-back now and write the two numbers down. Repeat them at the same time tomorrow. Within a week you will have something more useful than any training claim: a personal baseline and a clear picture of how much your own performance swings from day to day.
Frequently asked questions
Does brain training actually work?
It reliably makes you better at the trained task and at very similar tasks. Study after study finds no meaningful transfer to unrelated abilities, so treat improvement as skill in that exercise rather than a general gain.
How long should I train each day?
Ten focused minutes is plenty. Longer sessions add fatigue, and fatigue lowers scores, which makes the measurement less useful rather than more.
Which exercise is the most demanding?
N-back, by a distance. It loads working memory continuously rather than in bursts, which is why it feels harder than a memory span task even at low levels.
What improves cognition better than brain games?
Aerobic exercise and sleep have the strongest evidence, followed by treating hearing loss, managing blood pressure and learning something genuinely new such as an instrument or a language.
Why has my score stopped improving?
Most of the gain on any task arrives in the first week as you learn the strategy, then it plateaus. That plateau is the normal shape of the curve rather than a sign you are doing it wrong.
Is this brain training free?
Yes, every exercise is free with no account, no email and no subscription.