Do Brain-Training Games Actually Work?

The honest, evidence-based verdict on whether brain games make you smarter — or just better at brain games.

Updated July 16, 2026

The Billion-Dollar Industry Built on a Half-Truth

Somewhere between 2005 and 2015, "brain training" turned into a real industry. Apps promised to make you sharper, quicker, and more resistant to age-related decline, all through games you'd play in fifteen minutes over coffee. By the early 2010s, industry analysts were pricing the market at well over a billion dollars a year, with venture money pouring into the category and headlines promising that a few minutes of daily play could rewire your brain.

The half-truth in that pitch is the word "rewire." Playing games does change your brain — any repeated skill does that, from chess to crossword puzzles to the piano. The dishonest part was what companies implied came next: that getting better at their specific game would make you generally smarter, sharper at your job, or protected from dementia. That leap, from "you got better at the game" to "you got better at life," is exactly what a decade of research has spent picking apart.

The argument got loud enough, and personal enough, that it produced two competing open letters from scientists in the same year, a multimillion-dollar FTC settlement, and a body of research that's more nuanced — and more useful — than either side's press release ever was. Here's what actually holds up.

2014: Two Open Letters, One Public Fight

On October 20, 2014, the Stanford Center on Longevity and the Max Planck Institute for Human Development published "A Consensus on the Brain Training Industry from the Scientific Community," signed by roughly 70 cognitive psychologists and neuroscientists. Its position was blunt: the industry, it said, was exploiting consumer anxiety about aging and cognitive decline with claims that outran the evidence. The signatories argued there was little evidence that brain games improve everyday cognitive performance beyond the training tasks themselves, and no credible evidence they slow or reverse age-related decline.

The rebuttal came fast. A group calling itself Cognitive Training Data — researchers with ties to the training industry alongside independent academics — gathered roughly 130 signatures on a counter-statement arguing the Stanford/Max Planck letter had overstated its case. Their position: the field had a real, if uneven, body of supportive evidence, and dismissing an entire category of research because some of it was weak, or because some of it was marketed badly, was itself unscientific.

Neither letter was wrong about everything. The skeptics were right that marketing had sprinted ahead of the data. The defenders were right that "brain training doesn't work" was too broad a claim for a category covering wildly different games, doses, and populations. What the fight actually clarified, once the dust settled, was that the real question was never a simple yes or no. It was which games, trained how, produce benefits that generalize to what.

Near Transfer vs. Far Transfer, in Plain Terms

Near transfer means improvement on tasks that closely resemble what you trained on. Practice a visual memory grid and you'll likely get better at other visual-span tasks built on a similar grid-and-recall structure. Far transfer means improvement on tasks that don't resemble the training at all — your working-memory game making you better at following a complicated recipe, remembering names at a party, or scoring higher on an IQ test.

Near transfer is well established. Play a game daily for weeks and you get measurably better at that game and at close cousins of it — that's not controversial, it's just practice. Far transfer is where the field split, and it's also where the marketing lived, because "you'll get better at this specific memory game" doesn't sell subscriptions the way "you'll get smarter" does.

The clearest case study is dual n-back, the working-memory task where you track two independent streams at once — say, a spoken letter and a spatial position — and flag whenever either one repeats from n steps earlier. A widely cited 2008 study reported that dual n-back training raised scores on fluid-intelligence tests, a striking far-transfer claim that made headlines. Later, better-controlled replications and systematic reviews found the near-transfer gains held up — people got reliably better at n-back-style tasks — but the far-transfer bump to general intelligence mostly evaporated once studies used active control groups and accounted for expectancy effects. For a deeper dive on what working-memory training can and can't do, we cover it in its own guide; Scotix's own Dual N-Back Blitz runs on the same core mechanic.

The FTC vs. Lumosity: What Overclaiming Actually Costs

If the 2014 letters were an academic argument, the Lumosity case was the regulatory version of the same fight, with a dollar figure attached. In January 2016, the Federal Trade Commission announced a settlement with Lumos Labs, maker of the Lumosity brain-training app, requiring a $2 million payment and a court order to stop specific advertising claims.

The FTC's complaint wasn't that Lumosity's games were useless. It was that the company's ads told customers the games would help them perform better at work and school, delay age-related cognitive decline, and protect against memory loss associated with conditions like Alzheimer's disease, dementia, PTSD, ADHD, and chemotherapy-related cognitive impairment — claims the FTC said weren't backed by the science Lumos Labs actually had. The agency called out the company for playing on consumers' fear of cognitive decline in older age without the evidence to justify the promise.

The settlement is a useful marker precisely because it wasn't really about whether brain games "work" in the narrow sense. It was about the gap between what the research supported — near-transfer gains on trained tasks — and what the advertising promised: protection from dementia. That gap is the entire reason this topic needs a skeptical eye, not because the games are worthless, but because the claims built on top of them frequently aren't honest.

What the Careful Reviews Actually Found

Away from press releases and lawsuits, a 2016 review published in Psychological Science in the Public Interest did what the dueling 2014 letters couldn't: it systematically graded the existing brain-training literature study by study. The conclusion tracked what's described above — solid evidence for near transfer, and evidence for far transfer that was inconsistent at best, often undermined by small samples, missing active control groups, and outcome measures that shifted after the fact.

The one place broader effects show up with any consistency is a specific, unusually large trial: ACTIVE, a National Institute on Aging–funded study that trained thousands of older adults in memory, reasoning, or speed-of-processing tasks. A follow-up analysis years later associated the speed-of-processing arm specifically with a lower rate of dementia diagnosis. It's a genuinely interesting result, but it's one training protocol, in one trial, examined with one type of after-the-fact analysis — not a blanket endorsement of "brain games" as a category. Treat it as a promising research lead, not a guarantee that applies to whatever app happens to be on your phone.

What Games Can Reliably Do For You

Strip away the overclaiming and there's a genuinely useful core left standing:

• You get better at the specific skill you're training. Practice mental arithmetic and you get faster at mental arithmetic. Practice a visual-span memory grid and your visual-span recall improves. That's real, replicated, and doesn't need a marketing department to be true.
• Near-transfer spillover to closely related tasks. Skills that share structure with what you trained — other short-term recall formats, other rapid-arithmetic layouts — tend to improve alongside it, even when the game itself never taught them directly.
• A built habit of focused practice. Ten focused minutes a day is a discipline most people don't otherwise build. The game is the excuse; the habit of sitting down and concentrating on one thing without a phone buzzing is the actual value, and it's one you keep whether or not it "transfers" anywhere.
• Feedback you can actually see. Scores, streaks, and rating systems turn an abstract goal like "be sharper" into a measurable one like "beat last week's time." That measurability is what keeps people coming back long enough to get any of the above benefits at all — motivation isn't a side effect here, it's load-bearing.

What They Can't Honestly Promise You

None of that adds up to the claims that got Lumosity fined or that the Stanford/Max Planck letter was written to push back on:

• Raising your general IQ. The far-transfer evidence for fluid-intelligence gains from consumer brain games is thin, and the strongest early claims — dual n-back and fluid intelligence, most famously — didn't hold up well under replication.
• Preventing or reversing dementia. No consumer brain-training app has evidence anywhere near strong enough to promise protection against Alzheimer's disease or other dementias. The one large trial with a suggestive dementia-related finding used a specific clinical protocol, not an off-the-shelf game, and even that finding is a single result, not a proven cause and effect.
• Treating a diagnosed condition. ADHD, traumatic brain injury, chemo-related cognitive fog, PTSD — these are exactly the claims the FTC found unsupported in the Lumosity case. If you're managing one of these, a game is not a substitute for clinical care.
• Making you "smarter" in some general sense. Getting good at a game measures getting good at that game. It doesn't measure a change in your underlying cognitive capacity, however much the interface makes it feel that way.

So Should You Play? A Grounded Verdict

Yes — for the same reason people do crosswords, sudoku, or chess puzzles: they're a fun, low-stakes way to exercise a specific mental skill, and measurable feedback makes practice satisfying in a way that open-ended "try to be sharper" advice never is. Play the games because ten minutes of focused mental arithmetic or memory recall is a good use of a coffee break, not because it's a shortcut to genius or a shield against getting older.

That framing isn't a downgrade — it's the honest version of what was always true. A specific skill, practiced deliberately, with feedback you can track over time, genuinely improves. That's worth doing on its own terms. The moment any brain-training product, including this one, starts implying it'll raise your IQ or keep dementia away, that's your cue to be skeptical, because the actual science never signed up for that promise.

Frequently Asked Questions

Play These on Scotix

More Guides