BrainFinch
Articles

Brain training games: what the research does and does not show

Research shows one thing clearly: people get better at the brain games they practice. It has not established that those gains carry over to general intelligence or to everyday life. This article covers the main studies, a regulator’s case against one brain training company, and what both mean for how to treat any brain game, including the ones on this site.

Do people get better at brain training games?

Yes. Practice makes people better at the task they practice, and this is the most consistent finding in the field. Play a grid memory game every day for a month and your scores in that game will very likely rise.

Nobody in the research argues about this part. Part of the gain is learning the game itself: where to look, how the rounds are paced and which strategies pay. The open question is whether anything changes beyond the game.

What is the difference between near transfer and far transfer?

Transfer means getting better at something you did not practice. Researchers split it by how far the new task is from the trained one.

Type What improves State of the evidence
Practice effect The game you trained on Reliable
Near transfer Tasks that closely resemble the trained one Sometimes found
Far transfer General intelligence, school or work results, everyday functioning Not established

Near transfer would be training on one grid memory game and then scoring higher on a different grid memory test. Far transfer would be training on that game and then doing better at tasks that look nothing like it. Brain training advertising has mostly been about far transfer, and that is the part with the weakest support.

What did the large studies and reviews find?

Three sources are cited most often, and they point the same way.

Owen and colleagues published a study in Nature in 2010, run with the BBC, in which about 11,000 people trained online for six weeks. Participants improved on the tasks they trained on. The authors found no evidence that the improvement transferred to untrained tasks, including tasks that were closely related to the training.

Melby-Lervåg and Hulme published a meta-analysis of working memory training in 2013. Pooling the studies available at the time, they found that training produced short-term gains on working memory tasks like the trained ones. They found no convincing evidence that the gains generalized to other skills, such as verbal and nonverbal ability or arithmetic.

Simons and colleagues published a long review in 2016 called “Do brain-training programs work?” in Psychological Science in the Public Interest. They examined the studies that brain training companies cited in support of their products. Their conclusion was that the evidence is extensive for improvement on the trained tasks, weaker for improvement on closely related tasks, and slight for improvement on distantly related tasks or everyday cognitive performance. They also reported that many of the studies had design problems, such as small samples and weak control groups.

“Not established” is a precise phrase. The studies do not prove that far transfer can never happen. They show that, when it has been tested carefully, convincing evidence for it has not turned up.

What was the Lumosity FTC settlement?

In January 2016 Lumos Labs, the company behind Lumosity, agreed to pay $2 million to settle charges brought by the US Federal Trade Commission. The FTC said the company had advertised that its games could help users do better at work and at school and could hold off cognitive decline, and that it did not have the evidence to back those claims.

The case was about advertising. The FTC did not rule on whether the games were enjoyable or well made. It objected to promises of real-world benefit made without adequate support, which is the same gap the research reviews describe.

How should you treat a brain game, then?

Treat it as a game, and treat the score as a measurement of how you played. Both of those are true of every brain game, and neither depends on transfer.

A score in a well specified game is a real number. If Flash Grid says you tapped 14 correct squares in 30 seconds, that is what happened. You can compare it with your score yesterday, or with a friend’s score on the same game. What the number cannot do is stand in for a wider claim about your memory or your mind.

A few practical rules follow from the evidence.

  1. Play a game because you like playing it. That reason holds up whatever the next study finds.
  2. Read a rising score as “I got better at this game.” The evidence supports that reading and no broader one.
  3. Be wary of any product that promises results at school, at work or in daily life from playing games. That is a far transfer claim.
  4. Do not read a score as a test result about yourself. A game round is short, the conditions are uncontrolled, and the score reflects the device and the day as well as the player.

What does BrainFinch promise, and what does it not?

BrainFinch promises games that take about a minute or less, clear rules, and scores you can compare. It does not promise that playing will improve your memory, attention, intelligence, or performance at school or work, and it does not claim to protect against anything.

Flash Grid, the game at the top of this page, shows how that works in practice. Some squares on a grid light up for a second or two, then the grid goes blank and you tap the squares that were lit, in any order. Each correct square is one point and a round lasts 30 seconds. Clear a pattern and the next one is harder: more squares light up, the grid grows from 3 by 3 to 4 by 4 and then 5 by 5, and you get less time to look.

With practice you will probably score higher at Flash Grid. You may learn to see the lit squares as one outline, such as a line or an L, which the game’s own tips suggest. That is a practice effect, and it is the kind of improvement the research supports. Whether it would show up on a different memory task is a near transfer question, and the honest answer is “sometimes, in some studies.” Whether it would show up anywhere else in your life is a far transfer question, and the evidence so far says not to count on it.

Some games on the site are modeled on tasks that researchers use. N-Back is a version of the n-back task, and Sequence Tap follows the Corsi block task. Sharing a format with a research task does not turn a one-minute round into a research measurement, and the site does not present it as one.

What you get from a round is a score, a personal best to beat, and on the daily challenge at /daily a leaderboard where everyone plays the same game that day.

Common questions

Do brain training games work?

They work in a narrow sense: people reliably get better at the games they practice. Reviews such as Simons et al. (2016) found little evidence that this practice improves general thinking skills or everyday performance.

What is the difference between near transfer and far transfer?

Near transfer is improvement on a task that closely resembles the one you practiced. Far transfer is improvement on something quite different, such as general intelligence or how you do at work or school. Near transfer is sometimes found, and far transfer is not established.

What happened in the Lumosity FTC case?

In January 2016 Lumos Labs, the company behind Lumosity, agreed to pay $2 million to settle US Federal Trade Commission charges that its advertising claims were not backed by evidence.

Does BrainFinch claim to make you smarter?

No. BrainFinch offers one-minute games and scores you can compare. It does not promise any improvement in memory, attention, intelligence or daily life.

Games in this article

Take all 62 games with you

The BrainFinch app plays offline, keeps your streak and reminds you when the daily challenge is ready.

Get it on Google Play