Video games can sharpen specific cognitive skills — attention, spatial reasoning, visual tracking — but the evidence that they raise general IQ is genuinely contested. Two major meta-analyses published in the same journal in 2018 reached nearly opposite conclusions. Bediou and colleagues found that habitual action gamers outperform non-gamers on attention and spatial tasks by about half a standard deviation (g = 0.55), a sizeable gap. Sala, Tatlidil, and Gobet, looking across 359 training studies, found no causal evidence that gaming lifts general cognitive ability at all. Both can be right at once, and understanding why is the whole story. Where games do build something measurable, the demands map onto dimensions the CMIAS framework (created by DesperateMinds founder Dr. Sarwar Naseer) treats as distinct — Novel Problem Solving, Speed of Updating, and the contested one this debate turns on, Cross-Domain Transfer.
Video Games and IQ — Key Statistics
To see your own profile across all seven dimensions instead of a single number, the DesperateMinds CMIAS assessment measures novel problem solving, updating speed, and cross-domain transfer in one ~90-minute session.
Do Video Games Make You Smarter?
The question is badly posed, and rephrasing it is half the work. "Smarter" hides two different claims: that games improve narrow, trainable skills, and that games raise the broad, stable capacity an IQ test estimates. The first is well supported. The second is where the field splits.
Picture a teenager who has logged hundreds of hours in a fast first-person shooter. Ask her to track several moving objects at once, or spot a target in visual clutter, and she will likely beat a non-gamer handily. Ask whether those hours raised her underlying reasoning ability — the thing that would show up on a vocabulary-and-matrices battery years later — and the evidence gets thin and quarrelsome. Conflating those two questions is why this topic generates such contradictory headlines, with one outlet announcing games make you a genius and the next warning they do nothing.
So hold the two apart from the start. Games train skills. Whether trained skills generalise into intelligence is a separate, harder question — and the most honest writers in this field say so plainly rather than picking the conclusion that flatters their priors.
What Action Games Actually Improve
The modern science starts with a single elegant study. Green and Bavelier (2003), publishing in Nature, showed that habitual action-game players had measurably better visual selective attention than non-players across four experiments — and then, in a fifth, trained non-players on an action game and watched their attention improve. That fifth experiment mattered because it pointed past mere correlation toward cause: the playing itself produced the change.
Fifteen years of work built on that foundation, and Bediou et al. (2018) gathered it into a meta-analysis covering studies from 2000 to 2015. Across cross-sectional comparisons, action gamers led non-gamers by g = 0.55 — roughly half a standard deviation — with the most robust gains in top-down attention and spatial cognition. Their separate analysis of training interventions, the more causally informative kind, found a smaller but real effect of about a third of a standard deviation. The pattern is consistent: action games reliably build the attentional and spatial machinery the genre demands. Those spatial gains overlap heavily with what the research on spatial intelligence describes, and a demanding game doubles as a sustained workout for the working memory that underpins IQ.
One honest caveat about the Bediou numbers, which their own team flagged. After publication, the authors identified clustering issues in cases where participant samples partly overlapped, and they published a correction with added sensitivity analyses. The headline conclusion held, but the episode is a useful reminder that even a careful meta-analysis is a living estimate, not a final verdict.
Discover Your Profile Across All Seven CMIAS Cognitive Dimensions in 90 Minutes
Instead of one score, the CMIAS assessment separates novel problem solving, updating speed, and cross-domain transfer — the abilities gaming research keeps circling.
Take the CMIAS Assessment →The Meta-Analysis That Says Otherwise
Now the counterweight — and it landed in the same journal, one issue later. Sala, Tatlidil, and Gobet (2018) ran three separate meta-analyses in Psychological Bulletin: one on the correlation between game skill and cognitive ability (310 effects), one comparing players with non-players (315 effects), and one on whether game training causally improves cognition (359 effects). All three returned small or null overall effects. Their conclusion was blunt: overall cognitive ability and video-game skill are only weakly related, and there is no evidence of a causal link.
| Bediou et al. (2018) | Sala et al. (2018) | |
|---|---|---|
| Focus | Action games specifically | Video games broadly + training |
| Cross-sectional effect | g = 0.55 (attention, spatial) | Weak correlation overall |
| Training (causal) effect | ~0.33 (near domains) | Near-null; no far transfer |
| Verdict | Games enhance attention & spatial cognition | No causal boost to general ability |
Sources: Bediou et al. (2018), Psychological Bulletin 144(1); Sala et al. (2018), Psychological Bulletin 144(2).
How can two rigorous teams disagree so sharply? Largely because they asked different questions. Bediou focused tightly on action games and on attention and spatial outcomes — exactly where the effects concentrate. Sala cast a wider net across game types and outcomes and tested the more demanding claim of transfer to general ability. Read together, they are less a contradiction than a boundary line: games build the skills they exercise, and stop there.
The Far-Transfer Problem
This is the crux, and it has a name. Far transfer is when a skill learned in one domain improves performance in an unrelated one — and it is one of the most stubbornly elusive effects in all of psychology. Working-memory training, music lessons, brain-training apps, even chess expertise: in case after case, practice makes you better at the trained task and its close cousins, and rarely at anything further afield. The same wall that defeats other interventions stands in the way of games, which is the heart of Sala and Gobet's argument and the reason their finding aligns with the broader literature on brain training and IQ.
Here the question maps cleanly onto the CMIAS framework, and not by coincidence. Its designer, founder Dr. Sarwar Naseer, treats Cross-Domain Transfer — recognising the same underlying structure across unrelated fields — as its own measurable dimension rather than something an IQ score assumes. The entire far-transfer debate is, in those terms, a debate about whether gaming raises Cross-Domain Transfer or merely the in-domain skills beneath it. The evidence so far says mostly the latter. A gamer's attentional edge is real; the leap from that edge to better reasoning about an unrelated problem is the leap the data does not reliably support.
I will register one qualification to the skeptical position, because Sala's framing can be read too harshly. "No far transfer" is not "no benefit." Sharper attention and faster visual processing are genuinely useful capacities, and the relationship between trained skills and the flexible, novel reasoning captured by fluid intelligence may yet prove more graded than a flat null implies. The strict version of the claim is well evidenced; treating it as proof that games are cognitively worthless overreaches.
"The cleanest reading of this whole literature is that games are excellent at building the skills they demand and unremarkable at building everything else. That is not a disappointment — it is just how human skill learning works."
— Adam Imran, MS Clinical Psychology · DesperateMinds
Are Skilled Gamers More Intelligent?
Yes, modestly — but the direction of that arrow is the catch. Kokkinakis, Cowling, Drachen, and Wade (2017), publishing in PLoS ONE, found that performance in two complex strategy games, League of Legends and DOTA 2, correlated with fluid intelligence: the better players tended to score higher on reasoning measures. They proposed, intriguingly, that commercial games might one day serve as informal "proxy" tests of cognitive performance.
There is a lovely historical footnote buried in that paper. Back in the 1980s, researchers correlated IQ-test scores with performance in a psychologist-built arcade game called Space Fortress and found that intelligence did not predict how players started — it predicted how fast they learned and where they eventually plateaued. That distinction has aged well. It suggests intelligence shapes the trajectory of skill acquisition more than the starting line, which is exactly the pattern you would expect if smart players climb faster rather than gaming making them smart. The correlation is real; the causation runs at least partly the other way.
Children, Gaming, and IQ Gains
The strongest evidence that gaming might actually move intelligence comes from children, and from a design built to defeat the usual confounds. Sauce, Liebherr, Judd, and Klingberg (2022) tracked 9,855 US children in the Adolescent Brain Cognitive Development cohort, measuring intelligence at ages 9–10 and again two years later, while controlling for both polygenic scores for cognition and socioeconomic status.
The result cut against the cultural script. Children who gamed above the average gained more in intelligence over those two years than their peers, with a standardized effect of β = +0.17 — which the Karolinska team translated to roughly 2.5 IQ points above the cohort's average change. Watching and socialising online showed no such gain. Is this proof that games build intelligence in the young? Not quite, because even a strong observational design cannot fully rule out a habit that travels with gaming. But it is the cleanest evidence available that, for children, time spent gaming is not the cognitive liability it is often assumed to be — and may be a modest asset.
Which Games, and How Much?
Genre matters more than hours. The cognitive effects in this literature come overwhelmingly from games that impose real-time demand: fast action titles for attention and spatial skill, complex strategy games for planning and updating. A game that asks you to track a shifting battlefield, hold several goals in mind, and revise your plan the instant the situation changes is exercising genuine cognitive control. An idle tapper or a slot-machine loop is not.
For anyone weighing this practically — a parent, or an adult curious whether their hobby is doing anything — the realistic takeaway is measured. Demanding games are a defensible, even beneficial use of some leisure time, and they are not displacing intelligence the way the old "games rot your brain" line claimed. They are also not a substitute for the slower, broader inputs that build a mind, and the honest guide to how to increase IQ is candid about how modest any single lever really is. Play the demanding ones, enjoy them, and don't expect a controller to do the work of an education.
The Bottom Line
The fairest verdict the evidence allows is narrower than either camp's slogan. Action and strategy games reliably sharpen attention, spatial skill, and the speed at which you update a plan — and for children, the best-controlled data even hints at a small lift in measured intelligence. What games do not reliably do is produce far transfer, the jump from a trained skill to general reasoning that has defeated nearly every cognitive intervention ever tried. Anyone promising that a game will raise your IQ is selling the one thing the research has most consistently failed to find.
Two of the most rigorous meta-analyses ever run on this question appeared in the same journal months apart and reached opposite headlines — yet neither is wrong. Games genuinely build attention and spatial skill, and those gains genuinely fail to spread to general intelligence. The contradiction dissolves the moment you stop treating "trains a skill" and "raises IQ" as the same claim.
Frequently Asked Questions
They sharpen specific skills more than general intelligence. Bediou et al. (2018) found action gamers outperform non-gamers on attention and spatial tasks by about half a standard deviation (g = 0.55). Whether that lifts overall IQ is disputed, since Sala et al. (2018) found no causal evidence of broad cognitive gains from gaming.
The strongest evidence is for visual attention and spatial cognition. Green and Bavelier (2003) showed action games enhance visual selective attention, and meta-analysis confirms robust gains in top-down attention and spatial skills. Effects on memory and general reasoning are smaller and far less consistent across studies.
They measure different things. Bediou et al. (2018) studied action games and found real gains in attention and spatial cognition. Sala et al. (2018) tested whether those gains transfer to general ability and found they do not. Both can be correct: games train near skills without producing far transfer.
There is a modest link. Kokkinakis et al. (2017) found performance in strategy games like League of Legends correlated with fluid intelligence. But correlation does not prove gaming builds intelligence; smarter players may simply climb faster, so the arrow of causation stays uncertain.
The best-controlled study suggests a small positive effect. Sauce et al. (2022), tracking 9,855 US children while controlling for genetics and family background, found above-average gaming predicted larger two-year IQ gains (β = +0.17), roughly 2.5 points above the cohort average.
Far transfer is when a skill learned in one domain improves performance in an unrelated one. The central question in gaming research is whether attention gains from games transfer to general intelligence. Sala et al. (2018) concluded they largely do not, consistent with the broader difficulty of achieving far transfer.
Map Your Novel Problem-Solving and Updating Speed in One Assessment
The CMIAS protocol scores seven dimensions of cognition separately, so you see where your real strengths and gaps sit — not just a single composite.
Start the CMIAS Assessment →References
- Bediou, B., Adams, D. M., Mayer, R. E., Tipton, E., Green, C. S., & Bavelier, D. (2018). Meta-analysis of action video game impact on perceptual, attentional, and cognitive skills. Psychological Bulletin, 144(1), 77–110.
- Green, C. S., & Bavelier, D. (2003). Action video game modifies visual selective attention. Nature, 423(6939), 534–537.
- Kokkinakis, A. V., Cowling, P. I., Drachen, A., & Wade, A. R. (2017). Exploring the relationship between video game expertise and fluid intelligence. PLoS ONE, 12(11), e0186621.
- Sala, G., Tatlidil, K. S., & Gobet, F. (2018). Video game training does not enhance cognitive ability: A comprehensive meta-analytic investigation. Psychological Bulletin, 144(2), 111–139.
- Sauce, B., Liebherr, M., Judd, N., & Klingberg, T. (2022). The impact of digital media on children's intelligence while controlling for genetic differences in cognition and socioeconomic background. Scientific Reports, 12(1), 7720.