Do Brain Training Games Work

Answer to the question

Do brain training games make you smarter?

Not in the way the advertisements said. Brain training makes people better at the games they practise, and in one large trial some of that lasted ten years. A trial that enrolled 52,617 people and a pooled analysis found no reliable carry-over to untrained abilities once expectation and publication bias were controlled.

Key takeaways

• In the ACTIVE trial of 2,832 older adults, 73.6 per cent of the reasoning group still performed above their starting level on the trained ability ten years later, against 61.7 per cent of controls, and 70.7 against 48.8 per cent for speed training. Memory training showed no difference.
• Owen's 2010 trial enrolled 52,617 people and analysed 11,430. Training improved the trained tasks and not the untrained tests, and the control group improved by similar amounts.
• In a 2016 George Mason experiment, 50 people did the same one-hour session. Only the 25 recruited by a flyer promising brain training improved on two intelligence tests.
• Sala and Gobet's 2019 second-order meta-analysis found the far-transfer effect and its true variance both came out at zero once placebo effects and publication bias were controlled.
• On 5 January 2016 the US Federal Trade Commission announced that Lumos Labs would pay 2 million dollars, with a 50 million dollar judgment suspended, saying Lumosity "simply did not have the science to back up its ads."

Two months after a consensus statement criticised commercial brain games, a group of scientists published an open letter arguing that some commercially available exercises have the potential to bolster cognition. That is correct, and it is the strongest objection to the answer given here.

Brain training works in the plain sense that people who practise a task get better at it, and in one large trial a benefit was still measurable ten years on. What it has not been shown to do is make people smarter in the way the advertisements said.

Every claim about brain training sits on one side of a single distinction. Near transfer means improving at the trained task and at tasks that closely resemble it. Far transfer means improving at something else, such as following a conversation, managing a workload or holding a plan together. The trials measure the first. The marketing promised the second.

The Case for Training

The best evidence in favour is ACTIVE, a randomised trial of older adults. Each trained group attended ten sessions of an hour or so, spread over about five weeks, in memory, reasoning or speed of processing, while a control group received nothing. Some participants were randomly chosen for later booster sessions.

The trial began with 2,832 volunteers whose average age was 74. Ten years later, 73.6 per cent of the reasoning group still performed above their starting level on the trained ability, against 61.7 per cent of controls. For speed training the figures were 70.7 and 48.8 per cent. Memory training showed no difference from controls. The follow-up reached 44 per cent of the original sample.

All three trained groups also reported less difficulty with daily tasks than the controls did. The standard performance tests of everyday function showed no differences between the groups.

Daniel Simons and six co-authors reviewed, in a 2016 paper in Psychological Science in the Public Interest, the studies that leading brain training companies cite on their own websites, presumably the best evidence those companies had. They found extensive evidence of improvement on the trained tasks, less on closely related tasks, and little on distantly related tasks or everyday cognition. Many of the studies had design or analysis flaws serious enough to block firm conclusions, and none met all the practices the authors judged essential for showing a benefit in daily life.

The Largest Trial Yet

Adrian Owen's team ran the largest test of far transfer, working with a BBC science programme, Bang Goes the Theory, and reported it in Nature in 2010. A total of 52,617 people aged 18 to 60 joined. The 11,430 who completed at least two training sessions and both test batteries formed the analysis. One group trained reasoning, planning and problem-solving. A second trained a broad mix that included short-term memory, attention, visuospatial processing and mathematics. A control group spent the time searching the web for answers to obscure questions. Training ran for six weeks, a minimum of ten minutes a day, three times a week, and the average participant completed about 24 sessions.

Everyone on a training programme improved at the tasks it contained. The gains did not carry over to untrained tests. Improvements there were small, with a largest effect size of 0.35, and the control group improved by similar amounts. The authors wrote that the results gave "no evidence for any generalised improvements in cognitive function."

Everyone on a training programme improved at the tasks it contained. The gains did not carry over to untrained tests.

The control group's improvement is the detail to hold on to. People score higher on a cognitive test the second time they take it, so any trial of brain training needs a comparison group, and what that group does shapes what the result can mean.

When Expectation Raises Scores

Walter Boot, Daniel Simons and two colleagues argued in 2013 that an active control group, one that does some other activity, rules out a placebo effect only if its members expect to improve as much as the trained group. Almost no psychological intervention checks that, they wrote, which is a design flaw that "potentially undermines any causal inference." They illustrated the point with video-game training research. The wider logic is the same one behind [how expecting an effect can produce it](https://thewhyweproject.com/articles/when-expecting-side-effects-causes-them).

A 2016 experiment at George Mason University made the argument concrete. Fifty people were recruited with one of two flyers. One advertised brain training and stated that numerous studies had shown working memory training can increase fluid intelligence. The other looked similar and said nothing about either. All fifty then did the same one-hour training session, 25 in each group. Only the group drawn by the first flyer improved on two tests of fluid intelligence, with effect sizes of 0.50 and 0.39. The authors judged that one hour of practice had almost certainly not changed anyone's intelligence.

Giovanni Sala and Fernand Gobet pooled the pooled evidence in 2019. Their second-order meta-analysis covered existing meta-analyses of working memory training, video games, music, chess and exergames. Working memory training produced near-transfer effects whose size varied with age group. Far-transfer effects were small or null, and once placebo effects and [publication bias and failed replications](https://thewhyweproject.com/articles/the-psychology-you-learned-didnt-replicate) were controlled, the overall far-transfer effect and its true variance both came out at zero. They called the failure of skills to generalise "an invariant of human cognition."

Letters and a Settlement

The first statement came from the Stanford Center on Longevity on 20 October 2014. The statement granted that cognitive training produces statistically significant improvement in practised skills. It said compelling evidence of general and enduring positive effects had "remained elusive", that there was "little evidence that playing brain games improves underlying broad cognitive abilities", and that promotion of brain games was "frequently exaggerated and at times misleading." The December letter agreed that claims for brain games are often exaggerated. Its worry was different: that the first statement would discourage academic research and leave a gap for unvetted commercial products.

On 5 January 2016 the US Federal Trade Commission announced that Lumos Labs, the company behind Lumosity, would pay 2 million dollars to settle charges of deceptive advertising. A 50 million dollar judgment was entered and suspended because of the company's financial condition, on condition that the payment was made. The Commission said the company had claimed its games would improve performance at school, at work and in athletics, delay age-related cognitive decline, and reduce impairment linked to a range of medical conditions. In the Commission's words, Lumosity "simply did not have the science to back up its ads."

The order required competent and reliable scientific evidence before similar claims were made again. It also required the company to notify subscribers on auto-renewal plans and give them a way to cancel. The settlement concerned advertising rather than the research literature, and it did not rule on whether any exercise works. It put a price on the gap between near and far transfer.

The settlement concerned advertising rather than the research literature. It put a price on the gap between near and far transfer.

An Advocate Citing Sceptics

The Distracted Mind, published by MIT Press in 2016, makes its case for cognitive training and video-game training in chapter 10. At the point where that case is made, it prints Owen's null result and Boot's placebo paper as consecutive endnotes. The chapter also grades training as a signal rather than a prescriptive intervention on an evidence ladder of the authors' own devising, and gives the top rung to physical exercise. By our count about 88 per cent of its 93 endnotes are peer-reviewed journal articles, the most rigorous sourcing in the book. Few advocacy chapters hand the sceptic both of the sceptic's best citations.

The chapter also recommends a game built in one author's own laboratory, and the journal paper behind that game states the author's commercial interest in it. The same book's account of [self-interruption](https://thewhyweproject.com/articles/self-interruption-is-half-the-story) covers the other side of attention, what people do to it themselves.

The advocates keep what is theirs. Practice on a task improves that task, and in ACTIVE part of the improvement outlasted a decade. Someone who enjoys a training game, or who wants to get quicker at one particular exercise, gets exactly that, a game that makes them better at that game, and nothing in the evidence disputes it.

Common questions

Do brain training games make you smarter?

Not in the way the advertisements said. Brain training reliably improves the trained task, and in the ACTIVE trial some gains lasted ten years. A trial that enrolled 52,617 people found no generalised improvement, and Sala and Gobet's 2019 meta-analysis put far transfer at zero once placebo effects and publication bias were controlled.

Did the Lumosity settlement prove brain training does not work?

No. The Federal Trade Commission's case in January 2016 was about advertising claims, and the company paid 2 million dollars of a suspended 50 million dollar judgment. The Commission said the firm lacked the science to support its claims about everyday performance and age-related decline. It did not test any exercise itself.

Does the ACTIVE trial show brain training helps older adults?

It shows a narrow benefit. Ten years on, the reasoning and speed groups were more likely than controls to perform above baseline on the ability they trained, and the memory group was not. All trained groups reported less daily difficulty, but performance tests of everyday function showed no group differences.

What is the difference between near and far transfer?

Near transfer is improvement on the trained task and on tasks that closely resemble it. Far transfer is improvement on something different, such as everyday memory or reasoning at work. Brain training trials reliably show the first. The evidence for the second is where the dispute lies, and the large trials and pooled analyses have not found it.

Sources

Gazzaley, A., & Rosen, L. D. (2016). The Distracted Mind: Ancient Brains in a High-Tech World. Cambridge, MA: MIT Press, chapter 10, pp. 190-199

Owen, A. M., Hampshire, A., Grahn, J. A., Stenton, R., Dajani, S., Burns, A. S., Howard, R. J., & Ballard, C. G. (2010). 'Putting brain training to the test.' Nature, 465(7299), 775-778. doi:10.1038/nature09042

Rebok, G. W., et al. (2014). 'Ten-Year Effects of the Advanced Cognitive Training for Independent and Vital Elderly Cognitive Training Trial on Cognition and Everyday Functioning in Older Adults.' Journal of the American Geriatrics Society, 62(1)

Simons, D. J., Boot, W. R., Charness, N., Gathercole, S. E., Chabris, C. F., Hambrick, D. Z., & Stine-Morrow, E. A. L. (2016). 'Do brain-training programs work?' Psychological Science in the Public Interest, 17(3), 103-186. doi:10.1177/1529100616661983

Boot, W. R., Simons, D. J., Stothart, C., & Stutts, C. (2013). 'The pervasive problem with placebos in psychology: Why active control groups are not sufficient to rule out placebo effects.' Perspectives on Psychological Science, 8(4), 445-454. doi:10.1177/1745691613491271

Foroughi, C. K., et al. (2016). 'Placebo effects in cognitive training.' Proceedings of the National Academy of Sciences, 113(27), 7470-7474

Sala, G., & Gobet, F. (2019). 'Near and far transfer in cognitive training: A second-order meta-analysis.' Collabra: Psychology, 5(1), 18

Stanford Center on Longevity & Max Planck Institute for Human Development (2014). 'A Consensus on the Brain Training Industry from the Scientific Community.' 20 October 2014

Federal Trade Commission (2016). 'Lumosity to Pay $2 Million to Settle FTC Deceptive Advertising Charges for Its 'Brain Training' Program.' Press release, 5 January 2016

Related concepts

• Placebo Effect
• Cognitive Training
• Publication Bias
• Transfer
• The Active Control
• Working Memory

People

• Adrian Owen