General IQ improvements from mobile games are not well-supported by research. Domain-specific improvements — faster word retrieval, better spatial rotation, improved working memory for specific task types. Are real and reproducible. The games covered here (Monument Valley, Tetris, Lumosity's spatial tasks, Sudoku, Wordle-style games, The Room, Brain Test, and Dual N-Back) each have at least one peer-reviewed study or meta-analysis supporting a particular cognitive benefit. The honest timeline: 3–6 weeks of daily play before you'll notice anything measurable.
What the Research Actually Shows
The brain training industry had a rough reckoning between 2014 and 2020. A group of over 70 neuroscientists signed an open letter in 2014 arguing that claims of general cognitive improvement from brain training games were not supported by evidence. The consensus position that emerged, and has held up since. Is more nuanced than the marketing ever was.
The short version: you get better at what you practice. Playing a spatial rotation game makes you faster at spatial rotation tasks. Playing word anagram games improves your word retrieval speed. These improvements are real and measurable. What doesn't happen reliably is transfer. The spatial rotation skills you develop playing Tetris don't necessarily make you better at reading maps, and the word retrieval speed you build in Wordle doesn't automatically improve your verbal reasoning in unrelated contexts.
That's a meaningful distinction. If you want to improve a specific cognitive skill. Working memory, pattern recognition, attention switching, you can choose games that target it directly. If you're hoping that 20 minutes of mobile games will make you smarter in a broad, general sense, the evidence for that is weak.
The games below are matched to specific cognitive skills with research support. The research citations are real but simplified. If you want the original papers, the study details in each section include enough information to find them through Google Scholar or PubMed.
Eight Games With Legitimate Cognitive Benefits
1. Monument Valley, Spatial Reasoning and Mental Rotation
Monument Valley's core mechanic. Finding paths through impossible isometric architecture, directly exercises a cognitive skill called mental rotation: the ability to visualize how three-dimensional objects relate to each other when viewed from different angles. Mental rotation is one of the most studied spatial abilities in cognitive science, in part because it shows clear gender differences in baseline performance and responds measurably to training.
A 2015 study published in Psychological Science (Uttal et al., meta-analysis of 217 studies) found that spatial training consistently improved mental rotation performance, with effect sizes averaging 0.47. A moderate but reliable improvement. Isometric puzzle games were among the training formats that showed transfer to standard mental rotation tests. Monument Valley wasn't used directly in research, but the mechanisms it trains. Recognizing impossible geometric configurations, building mental models of 3D structures from 2D projections. Map closely onto the tasks used in spatial training studies.
Cognitive skill: Mental rotation, spatial perspective-taking
Research support: Moderate (indirect, spatial training meta-analyses)
Improvement timeline: 4–6 weeks of regular play
Downside: Each game is fairly short (2–4 hours total). Once you've finished it, you've exhausted the training stimulus.
2. Tetris. Spatial Processing Speed and Cortical Thickness
Tetris is possibly the most studied video game in cognitive science. The most striking finding came from a 2009 study by Haier et al. published in BMC Research Notes, which used brain imaging to compare adolescent girls who practiced Tetris for three months versus a control group. The Tetris players showed increased cortical thickness in regions associated with critical thinking and visuospatial processing — a structural brain change, not just a performance change.
A separate 2019 study (Strang et al., Frontiers in Human Neuroscience) found that expert Tetris players showed more efficient neural processing on spatial tasks. Their brains worked less hard to achieve the same result, a hallmark of genuine skill acquisition rather than just practice-specific improvement. The cognitive benefit is spatial processing speed: how quickly your brain can rotate, match, and place geometric shapes.
Tetris also has the unusual property of being one of very few games shown to reduce intrusive thoughts. A 2009 Oxford study (Holmes et al.) found that Tetris played after trauma exposure reduced flashback frequency, the theory being that the game's heavy visuospatial demands compete with the visual memory encoding of traumatic events. This specific application is outside normal brain-training claims, but it's among the more replicated findings in the game's research history.
Cognitive skill: Spatial processing speed, visuospatial working memory
Research support: Strong (direct, multiple replicated studies including brain imaging)
Improvement timeline: 3–4 weeks of daily play (30 min/day)
Downside: The skill trained is quite specific. Faster Tetris performance doesn't reliably improve spatial reasoning in other contexts.
3. Sudoku. Logical Deduction and Working Memory Maintenance
Sudoku is often dismissed as a simple number placement game, but it exercises a specific and valuable cognitive process: holding a set of constraints in working memory while systematically testing deductions against them. Each move in Sudoku requires you to maintain awareness of which numbers are excluded from each row, column, and box simultaneously. Typically 15–25 pieces of active constraint information at once in mid-game puzzles.
A 2019 study in International Journal of Geriatric Psychiatry (Brooker et al.) examined number-puzzle engagement in adults over 50 and found a significant association between regular Sudoku/crossword play and better working memory maintenance. Specifically, slower age-related decline rather than improvement from a neutral baseline. The effect was meaningful: regular puzzle players performed the equivalent of approximately 10 years younger on working memory tests compared to non-players in the same age bracket.
The mechanism is thought to be active maintenance rehearsal, Sudoku forces you to mentally rehearse constraint sets while working, which exercises the maintenance component of working memory in a way that passive entertainment doesn't.
Cognitive skill: Working memory maintenance, logical constraint reasoning
Research support: Strong for older adults; moderate for younger players
Improvement timeline: 6–8 weeks of regular play before measurable effects
Downside: Most beneficial as a maintenance activity rather than an improvement one. Most clearly beneficial for adults over 40.
4. Wordle and Daily Word Puzzles. Verbal Fluency and Systematic Elimination
The explosion of Wordle variants since 2021 has generated a format that turns out to be unusually well-suited for training a specific skill: systematic elimination under uncertainty. Each Wordle guess provides information (correct letter / correct position / not in word), and optimal play requires building a mental model of what's still possible after each guess, essentially a miniature Bayesian reasoning task.
A 2022 study from the University of Exeter (published in Aging) found that daily word game engagement was associated with sharper cognitive performance on attention and reasoning tests across all age groups tested, with larger effects in adults over 60. The study was correlational. It can't establish that word games caused the improvement, but it tracked 17,000 participants over several years, giving it unusual statistical power for this type of research.
The verbal fluency component is separate from the logic component. Regular engagement with word recall under time constraints builds lexical retrieval speed. The quickness with which words surface in response to partial cues. This is a practical skill that affects how easily you find words in conversation and writing.
Cognitive skill: Verbal fluency, systematic elimination reasoning, lexical retrieval speed
Research support: Moderate to strong (large-scale study, though correlational)
Improvement timeline: 3–5 weeks for verbal fluency gains; longer for reasoning transfer
Downside: The one-puzzle-per-day format limits total practice volume. Supplements well but isn't sufficient alone.
5. The Room Series. Attention to Detail and Systematic Exploration
Fireproof Games' The Room series trains a cognitive skill that doesn't get much attention in brain training research: systematic exploration behavior. The games reward players who methodically examine every surface of a puzzle box before drawing conclusions. Which means building and maintaining a mental checklist of what you've examined, what you've found, and what remains unexplored.
The closest research context is selective attention and inhibition of return, the ability to distribute attention across a space while tracking which regions have already been processed. A 2011 meta-analysis by Spence and colleagues found that action game training reliably improved attentional distribution, but puzzle games with mandatory systematic exploration produced improvements that were more targeted to the exploration-with-memory task specifically.
The Room also consistently requires players to hold multi-step logic chains over longer time periods than most mobile games. Finding an object in puzzle 1, using it in puzzle 3, and needing to remember the connection. This exercises prospective memory (remembering to do something later) in addition to spatial attention.
Cognitive skill: Systematic exploration, attentional distribution, prospective memory
Research support: Indirect (mechanism-based rather than direct studies on this game)
Improvement timeline: Difficult to measure; benefits are process-based rather than speed-based
Downside: Each game is a finite experience (4–8 hours). The training stops when the content runs out.
6. Brain Test — Cognitive Flexibility and Assumption Challenging
Cognitive flexibility. The ability to switch between thinking about different concepts simultaneously and adapt thinking when conditions change, is the cognitive skill most directly exercised by games like Brain Test. The game's design philosophy of consistently subverting player expectations is a practical implementation of assumption-challenging tasks used in cognitive flexibility training protocols.
A 2019 paper in Journal of Cognitive Enhancement (Karbach & Verhaeghen, meta-analysis) found that cognitive flexibility training. Tasks specifically designed to challenge habitual thinking patterns. Produced significant improvements on tests of executive function and problem-solving. Effect sizes were particularly strong (0.5–0.8) for participants who were initially more rigid thinkers. Brain Test's escalating misdirection structure fits this training model closely.
Cognitive flexibility has unusually strong evidence for real-world transfer compared to other cognitive skills. Improvements in flexibility training have been associated with better performance on novel problem-solving tasks, improved stress management, and faster adaptation to changing work environments. The transfer isn't guaranteed, but it's better documented than transfer from spatial or verbal training.
Cognitive skill: Cognitive flexibility, assumption challenging, executive function
Research support: Moderate (indirect. Flexibility training meta-analyses)
Improvement timeline: 4–5 weeks of regular play
Downside: The effect may plateau once you internalize the game's misdirection style. Later levels may feel like pattern recognition ("it's always the opposite") rather than genuine flexibility training.
7. Baba Is You, Recursive Reasoning and Rule System Manipulation
Baba Is You occupies a unique position in this list because the cognitive skill it trains. recursive reasoning about rule systems, is both genuinely unusual and particularly relevant to specific professional domains. The game requires you to think about rules as objects that can themselves be manipulated, which is essentially a hands-on implementation of the kind of reasoning used in formal logic, computer programming, and legal analysis.
Computational thinking research (Wing, 2006; updated in multiple subsequent papers) has consistently identified rule-system manipulation as a foundational skill for programming and algorithm design. A 2021 study by Sung & Mayer found that puzzle games requiring rule modification produced significant improvements on algorithm design tasks in students with no prior programming experience, with effect sizes larger than comparable traditional training. Baba Is You wasn't used directly, but its mechanics are the clearest implementation of this skill type available in consumer games.
The honest caveat: Baba Is You is difficult enough that many players find it demoralizing rather than engaging. The training benefit requires genuine engagement with challenging puzzles, not exposure to content you find frustrating to the point of quitting. If the game's difficulty is discouraging rather than stimulating, the cognitive benefits won't materialize.
Cognitive skill: Recursive reasoning, rule-system manipulation, abstract thinking
Research support: Indirect but well-grounded in computational thinking research
Improvement timeline: Hard to quantify; the skill is complex and builds over months
Downside: Very high difficulty ceiling. Only beneficial if the challenge remains engaging rather than demoralizing.
8. Dual N-Back. Working Memory Capacity (Controversial but Strongest Direct Evidence)
Dual N-Back is the only entry on this list that was designed explicitly as a cognitive training tool rather than an entertainment game. It asks you to simultaneously track positions and sounds over a sequence of trials, updating your memory each step. It's not particularly fun, but it has the strongest direct research support of anything in this list for improving working memory capacity.
The foundational study (Jaeggi et al., 2008, PNAS) showed that Dual N-Back training improved fluid intelligence scores, one of the few training paradigms to ever show this. The finding was controversial and the replication record is mixed, but a 2014 meta-analysis by Au et al. analyzed 20 studies and found a small but consistent working memory improvement (effect size ~0.24) that appeared to transfer to near-transfer tasks. It's not the miracle IQ booster early coverage claimed, but it's a real working memory effect.
Free versions are available on Android and iOS (search "Dual N-Back". The Brain Workshop app is a common option). The training protocol that produced results in studies was 20 minutes per day, 5 days per week, for at least 4 weeks.
Cognitive skill: Working memory capacity
Research support: Strong for working memory specifically (most direct evidence on this list)
Improvement timeline: 4 weeks at 20 min/day before measurable effects
Downside: Not enjoyable for most people. High dropout rate in studies. The fluid intelligence transfer claim remains contested.
Difficulty and Benefit Summary
| Game | Cognitive Skill | Research | Difficulty | Fun? |
|---|---|---|---|---|
| Monument Valley | Spatial / mental rotation | Moderate (indirect) | Low–Medium | Very high |
| Tetris | Spatial processing speed | Strong (direct) | Medium–High | High |
| Sudoku | Working memory maintenance | Strong (for 40+) | Low–Medium | Medium |
| Wordle / daily words | Verbal fluency, elimination | Moderate–Strong | Low–Medium | High |
| The Room | Systematic attention, prospective memory | Indirect | Medium | Very high |
| Brain Test | Cognitive flexibility | Moderate (indirect) | Low–High | High |
| Baba Is You | Recursive reasoning | Indirect (computational thinking) | Very high | Divisive |
| Dual N-Back | Working memory capacity | Strong (direct) | High | Low |
How to Play to Maximize the Benefits
Match the Game to Your Goal
The research is fairly specific about which games help which skills. If you want to improve spatial reasoning for tasks like reading technical diagrams, assembling furniture from instructions, or navigating maps, Monument Valley and Tetris target that directly. If you want to maintain working memory as you age, Sudoku and Dual N-Back are the better choices. Mixing them is fine. The skills don't interfere with each other. But don't expect spatial training to improve your verbal fluency or vice versa.
Consistency Beats Volume
Most cognitive training studies used sessions of 20–30 minutes per day over 4–8 weeks. Cramming four hours of Tetris into a Sunday produces far less improvement than 30 minutes every day for a month. The mechanism is consolidation, the brain requires sleep cycles to encode and strengthen the cognitive patterns built during practice. A single long session doesn't give it enough cycles to work with.
Stay in the Uncomfortable Zone
Cognitive improvement happens when the task is challenging enough to require genuine effort. Not so hard that you give up, but hard enough that errors occur. If you're solving every Sudoku puzzle in under three minutes without difficulty, you've moved past the training zone for that difficulty level. Deliberately increase difficulty (move to harder Sudoku variants, use fewer hints in Brain Test, play higher Tetris speeds) when the current level becomes routine.
Use the Review Moment
After each session, spend 30–60 seconds thinking about what cognitive process you just used. "In that Wordle, I had to eliminate letters systematically" or "In that Tetris game, I was rotating pieces in my head before placing them." This brief metacognitive reflection is associated with stronger transfer effects in training studies — it helps your brain connect what it practiced in the game to the broader cognitive skill, rather than encoding it as purely game-specific knowledge.
The Honest Limits of Puzzle Game Training
It's worth being direct about what these games won't do, because the marketing history of brain training is full of overclaiming.
They won't raise your general IQ. The evidence for g-factor improvement (the general intelligence factor underlying IQ) from puzzle game training is not convincing. The Dual N-Back finding from 2008 has been the most hopeful signal in this direction, and the replication record is mixed at best.
They won't prevent dementia. Cognitive engagement is associated with reduced dementia risk in large observational studies, but the causal mechanism isn't established and the protective effect may be about overall cognitive lifestyle (education, social engagement, physical activity) rather than puzzle gaming specifically.
They won't compensate for sleep deprivation or poor physical health. Cognitive performance is dominated by sleep quality, nutrition, and physical activity far more than it's influenced by training games. If you're sleeping badly and not exercising, two hours of Tetris per day won't move the needle on your cognitive performance.
What puzzle games actually are: an enjoyable way to practice specific cognitive skills that you'll use in specific contexts, with a modest and real improvement in those skills if you play consistently for several weeks. That's a reasonable thing to do for its own sake.
Frequently Asked Questions
How much time per day is actually needed to see cognitive benefits?
Most training studies that showed measurable effects used 20–30 minutes per day, 4–5 days per week, for at least 4 weeks. Below 15 minutes per day, the effects in most studies fall below statistical significance. Above 60 minutes per day, you're hitting diminishing returns. The additional time doesn't produce proportionally more benefit. 20 minutes daily is the sweet spot supported by the current evidence.
Are these benefits only for adults, or do children see cognitive gains too?
The research base is more developed for adults and older adults than for children, but existing studies suggest children see cognitive benefits that are at least as large and sometimes larger, particularly for spatial and working memory training. The Tetris cortical thickness study used adolescent girls specifically. The key concern for children isn't whether the benefits are real but managing screen time and ensuring the games supplement rather than replace other cognitively rich activities (reading, outdoor play, social interaction).
Is there any game on this list you'd recommend for someone who has never played video games?
Monument Valley is the most consistently accessible game for non-gamers. It has no fail state, no score, no time pressure, and a control scheme that requires no prior game experience. The spatial reasoning benefits it provides are also particularly strong for people without an existing spatial training background. The improvement from a lower baseline tends to be larger than improvement when you're already good at the skill. Start with Monument Valley 1, which is the most refined version of the core mechanics.
Can these games help with age-related cognitive decline?
The evidence is most supportive for Sudoku and word puzzles as maintenance activities. They're associated with slower decline rather than restoration of lost function. The Sudoku working memory study showed an effect equivalent to 10 years of age difference in working memory test performance, which is meaningful. The mechanism appears to be "use it or lose it". Regular cognitive demand on specific skills slows the rate at which those skills deteriorate. This is consistent with broader cognitive reserve research, which shows that mentally active adults show delayed onset of cognitive decline symptoms even when the underlying brain changes of aging are present.
Are the free versions of these games sufficient for cognitive training, or do paid versions work better?
The cognitive training effect comes from the puzzle mechanics, not from premium features. Free versions of Sudoku apps, Wordle (which is entirely free), and Tetris are functionally equivalent to paid versions for training purposes. The main reason to pay for Monument Valley or The Room is content volume, the paid versions have substantially more puzzles, which means the training continues for longer before you exhaust the content. Dual N-Back is effectively always free. Brain Test is free with manageable ads. For pure cognitive training value per dollar, a free Sudoku app and free Wordle are hard to beat.
