What the number means
The most repeated result from Sauce and colleagues is that children who spent more time gaming than average showed about 2.55 additional IQ points of change over two years.
That is the modelled association in the follow-up sample. It is not a promise that a child will gain 2.55 points by gaming, and it does not mean a League profile can estimate an individual’s IQ.
Study design
The researchers used data from the US Adolescent Brain Cognitive Development study:
- 9,855 children aged 9 or 10 were included at baseline;
- 5,169 of those children had the two-year follow-up data used for the longitudinal analysis;
- children reported time spent watching digital video, socialising online, and gaming;
- intelligence was represented as a latent score built from several cognitive tasks;
- models included socioeconomic status, a polygenic score related to cognitive performance, age, collection site, and genetic population structure.
The genetically informed model is a notable strength. It addresses part of the selection problem that affects simpler screen-time studies. It does not measure or eliminate every relevant family, developmental, educational, health, or motivational factor.
Results in context
At baseline, gaming time was not correlated with intelligence in the reported model. Over two years, more gaming was associated with greater intelligence change, with a standardized coefficient of 0.17. The authors translated that effect to about 2.55 IQ points.
Watching digital videos was also associated with greater change in the main model, although that result was sensitive to how socioeconomic background was represented. Socialising online was not associated with the change.
This nuance matters. A simple story in which active media helps and passive media does not is not what the full result shows.
Read the Scientific Reports article for the model, measures, and sensitivity analyses.
Why longitudinal is not the same as causal
Following the same participants is stronger than comparing two groups once. It establishes temporal order for measured change. Several issues remain:
- screen time was self-reported;
- the study did not randomize children to gaming exposure;
- specific games and genres were not separated;
- almost half of the baseline participants were not in the follow-up sample;
- controls reduce known confounding but cannot remove unmeasured confounding;
- a population association does not predict the result for one child.
The authors used causal language in parts of the paper, but the observational design still requires care when translating the finding into advice.
What this study does not say about League
The exposure variable was broad gaming time, not League rank, match performance, champion choice, or any Saiki metric. The participants were children, not a representative sample of adult competitive players.
The result therefore does not support:
- estimating IQ from a Riot ID;
- calling a rank or economy statistic an intelligence score;
- recommending a gaming dose for cognitive development;
- claiming that playing League will increase intelligence;
- treating personality and intelligence as interchangeable.
Saiki does not currently administer an intelligence test. Its Strategic Play panel reports game economy and tempo, and the app explicitly states that it is not an IQ score.
A defensible takeaway
The study makes broad claims about screen time less tidy. It suggests that gaming exposure and cognitive development may be related even after several important controls, and that the relationship deserves better causal research.
For players, parents, and product teams, the responsible conclusion is not “games increase IQ.” It is “one large longitudinal model found a modest association that future work should test with better exposure measures, clearer game categories, and stronger causal designs.”