The pitch is familiar. Play a few short games each morning, and your memory, attention and processing speed improve. Some apps even give you a brain age.
Twenty years of research has now tested that claim thoroughly, and the answer is unusually clear for a psychology question: you get better at the games. Almost everything else is marketing.
That does not mean puzzle games are worthless. It means the reason to play them is different from the reason you were sold.
The Distinction Everything Turns On
Cognitive training research separates two kinds of improvement.
Near transfer is improvement on the trained task and tasks that closely resemble it. Practice a digit span task, get better at digit span tasks. This is well documented, uncontroversial and happens quickly.
Far transfer is improvement on genuinely different abilities. Practice a memory game, and your work performance, reasoning or everyday forgetfulness improves. This is what the apps sell, and it is what the evidence does not support.
| Claim | Evidence |
|---|---|
| You improve at the specific game | Strong. Happens within days. |
| You improve at very similar tasks | Moderate. Diminishes as the task differs. |
| Working memory capacity increases | Weak. Meta-analyses find no durable transfer. |
| Fluid intelligence increases | Not supported. |
| Everyday memory and focus improve | Not supported. |
| Dementia risk falls | Not established. |
What the Big Studies Found
The FTC action. In January 2016 the US Federal Trade Commission settled with Lumos Labs over Lumosity's advertising, which claimed the games improved everyday performance and delayed cognitive decline. The regulator's position was that the company preyed on consumer fears without the science to back the claims. The settlement carried a 2 million dollar judgment.
The scientists' letter. In 2014 a group of around seventy cognitive psychologists and neuroscientists, organized through the Stanford Center on Longevity and the Max Planck Institute for Human Development, published a consensus statement saying the brain game industry's claims outran the evidence. A rival letter from other researchers argued the opposite, which tells you the field was genuinely contested at the time. The meta-analyses that followed mostly favored the skeptics.
Working memory training meta-analyses. Reviews by Melby-Lervag and Hulme and later replications found the same pattern repeatedly: reliable short term gains on tasks resembling the training, and no convincing evidence of lasting transfer to reading, arithmetic or fluid intelligence.
The ACTIVE trial. This is the strongest evidence in the other direction and deserves an honest reading. It trained 2,832 older adults in memory, reasoning or processing speed. Ten years later, participants still performed better in the domain they trained. Effects on real world functioning were far smaller and inconsistent across the three arms. So: training endures, transfer stays narrow.
Why Far Transfer Fails
Practice makes you better at a task in two ways. You can improve the underlying capacity, or you can learn a strategy specific to that task.
Almost all measurable improvement in brain games appears to be the second one. You learn the pattern, the chunking trick, the specific layout. That knowledge is genuinely useful for that game and useless everywhere else, which is exactly what a narrow transfer profile looks like.
Add two artifacts that inflate results in weaker studies:
- Test familiarity. Many studies measure improvement with a test that resembles the training, so near transfer gets reported as if it were general improvement.
- Expectation effects. Participants who paid attention to being enrolled in brain training expect to improve, and passive control groups do not correct for that. Studies with active control groups, where the comparison group also plays something, consistently show smaller effects.
What Does Improve Cognition
If the goal is a mind that works well over decades, the evidence points somewhere less convenient than a subscription app.
- Aerobic exercise. The most consistently supported intervention for cognitive health in adults.
- Sleep. Memory consolidation happens during it, and short sleep degrades attention measurably the next day. Our guides on how much sleep you need by age and falling asleep faster are the more useful reading here than any training app.
- Learning something genuinely hard. A language, an instrument, a craft. Complex, effortful, sustained over months. This is the closest thing to real cognitive training that exists, and it is nothing like a five minute game.
- Treating hearing loss and managing blood pressure. Both appear repeatedly in the modifiable risk factor literature for cognitive decline.
- Social engagement. Consistently associated with better cognitive aging.
None of these are what people want to hear. All of them beat brain games in the evidence.
The One Honest Benefit of Puzzle Games
There is a real finding that is often confused with brain training, and it is worth separating out.
Visually demanding puzzle games appear to compete for the same mental resources used to hold intrusive visual memories. A line of research starting with Tetris studies, later tested with trauma patients in an emergency department setting, found reduced intrusive memories when the game was played shortly after a distressing event. It is a specific, narrow effect about visual interference, not a general brain benefit, and it should not be stretched into one.
Beyond that, the real benefits are the ordinary ones nobody bothers to dispute: a puzzle game is absorbing, it fills the awkward ten minutes in a queue, and it delivers a small clean sense of progress. That is a legitimate reason to play something. It just is not a medical one.
If you are trying to spend less of your day on the phone rather than more, the honest counterweight is our reduce screen time guide and the 7 day dopamine detox plan.
How to Pick a Puzzle Game Worth Playing
Since the cognitive claim is off the table, judge on the things that actually determine whether you enjoy it.
1. **Session length that respects you.** A good puzzle game is fine to leave after two minutes. Anything built around energy timers is designing around your schedule, not for it.
2. **Works offline.** The queue, the flight and the underground are when you actually want it.
3. **No forced video ad between every round.** One optional ad for a reward is a fair trade. A mandatory 30 second ad after every attempt is a slot machine with extra steps.
4. **Real depth rather than difficulty inflation.** Good puzzles get more interesting. Bad ones get slower until you pay.
5. **No brain age score.** If a game claims to measure your cognitive age, you now know exactly what that number is worth.
Chain is built on the merge and evolve loop, where combining matching pieces creates the next tier and the board fills while you look for the chain that clears it. The appeal is the planning problem, not a promise about your neurons. It plays offline, sessions end when you want them to, and the difficulty comes from the puzzle rather than from waiting.
For more in the category, see the best free offline iPhone games.
The Bottom Line
Brain training games make you better at brain training games. Near transfer is real, far transfer is not, and the industry was fined for blurring the difference.
If you want cognitive health, the boring interventions win: move, sleep, learn something genuinely difficult, stay connected, look after your hearing and blood pressure. If you want a good puzzle game, pick one that is fun and fairly built, and play it for the reason people have always played games. Neither goal is served by pretending the second one accomplishes the first.