Language: English, EN
Download
Chess and maths: what the studies say

Learn chess

Chess and maths: what the studies really say

A parent's guide to chess and maths: the Sala and Gobet meta-analysis, the English trial with 4,009 pupils, and how to read a study without being misled.

"Chess improves maths." You have probably read it in an after-school club leaflet, a news story or a parents' group chat. Sometimes it comes with a number ("20% better") and almost always with the words "it's proven". This article is here so you can read those claims calmly and know what lies behind them. Honest spoiler: the story is more interesting than the slogan, and the practical conclusion is good news for chess, just not for the reasons you were told.

The essentials in one minute

  • Older studies, taken together, show a moderate effect of chess on maths.
  • Almost none compared chess with another new, motivating activity. When that is done, the difference disappears.
  • The largest randomised trial, in England, with 4,009 pupils in 100 schools, found an effect of 0.01: none.
  • Strong players are, on average, a bit better with numbers, but that is correlation, not cause.
  • The European Parliament's 2012 declaration is a political request, not scientific proof.
  • Conclusion: chess is well worth it as a game. As a shortcut to better maths grades, there is no evidence that it works.

The table of studies

StudyDesignSampleResultCaveat
Sala and Gobet, 2016Meta-analysis24 studies; 2,788 children with chess and 2,433 withoutModerate effect on maths (g = 0.38), smaller on reading (g = 0.25)Almost no study had an active control
Sala and Gobet, 2017Two experiments with active controls (checkers and go)233 and 52 childrenNo significant difference in maths problem-solvingMedium samples, 25 hours of chess
Jerrim and others (EEF), 2016Cluster randomised trial4,009 pupils, 100 schools0.01 on maths (95% CI -0.15 to 0.16); nothing on reading or scienceMeasured a year later, with national tests
Rosholm and others, 2017Quasi-experimental with administrative dataPupils in grades 1 to 3 in Aarhus (Denmark)Tends towards positive effects, larger in bored childrenNot randomised; replaced a maths lesson
Aciego and others, 2012Quasi-experimental170 with chess, 60 with football or basketballCognitive and social-emotional gainsSelf-selection, no randomisation, teacher not blind
Burgoyne and others, 2016Meta-analysis of correlationsPlayers of different agesChess skill and cognitive ability correlate (r = 0.24)Correlation, not cause
Sala and Gobet, 2017Review of three meta-analysesChess, music and working memory trainingThe better the design, the smaller the effectReview with a strong thesis

Let's look at them one by one.

The meta-analysis: a moderate effect with a catch

A meta-analysis pools many studies and calculates an average effect. Giovanni Sala and Fernand Gobet gathered 24 studies, with 40 effect sizes, and found a moderate overall effect of chess on academic and cognitive skills, slightly larger on maths (g = 0.38) than on reading (g = 0.25). They also found that more hours of instruction went with larger effects (Sala and Gobet, 2016).

So far, good news. But the authors flag something crucial: no study used the ideal design, and almost none had an active control group.

What is that? Picture two classes. One starts chess with a new, enthusiastic instructor. The other carries on as usual. If the first improves, is it the chess, or the novelty, the motivation and the instructor? You cannot tell. An active control group is a third class doing another new activity that is just as appealing (checkers, go, a workshop). If chess beats that too, then it points to chess.

With an active control, the effect goes away

The same authors ran exactly that experiment. In the first, 233 children in third and fourth grade: some had 25 hours of chess, some played checkers, and some did nothing special. In the second, 52 children: chess against go. Neither showed significant differences in mathematical problem-solving (Sala and Gobet, 2017).

Their later review, which pulls together this kind of result for chess, music and working memory training, sums up the pattern: effects are small to moderate, and the better the design, the smaller they get (Sala and Gobet, 2017). Their conclusion is that far transfer (from a game to a quite different school subject) rarely happens.

A fair caveat: these are medium-sized samples and 25 hours is not a lot. These studies do not rule out a small effect; what they show is that there is no large effect that is easy to see.

The English trial: the most serious test, with a null result

The Education Endowment Foundation (EEF), a British charity that evaluates education programmes, funded the largest trial so far (Jerrim and others, 2016). The exact figures:

  • Design: cluster randomised controlled trial. Schools were assigned to chess or no chess by lottery.
  • Sample: 100 schools in 11 local authorities in England, 4,009 Year 5 pupils (aged 9 to 10).
  • Intervention: a year of chess lessons during the 2013/2014 school year.
  • Measure: the national Key Stage 2 tests in June 2015, a year after the programme ended.
  • Maths result: effect size 0.01, with a 95% confidence interval from -0.15 to 0.16. On the EEF scale, "0 months" of progress.
  • Reading and science: no evidence of an effect either.
  • Pupils eligible for free school meals (a marker of lower-income families): 0.01, with an interval from -0.18 to 0.19.

An effect of 0.01 is, for practical purposes, zero. And the interval tells us that, with reasonable confidence, any real effect would be small. The EEF rates the security of these findings as high. The journal version of the same trial reaches the same conclusion (Jerrim and others, 2017).

Does it have limits? Yes, and the authors point them out: it was measured a year after the programme ended, using general exams rather than tests tied to chess. Some supporters of school chess also argue there may have been a ceiling effect (Sala, Foley and Gobet, 2017), although that is an opinion piece and one of its authors belongs to the organisation that delivered the programme. Those criticisms explain why the result might differ under other conditions; they do not turn a zero into a positive effect.

A nice detail from the same report: half the pupils said they liked the chess lessons a lot, and only 8% said they did not. Teachers perceived improvements in concentration. Grades did not go up, but the children enjoyed it.

Denmark: the positive study, and why it is not enough

In Aarhus, one of the weekly maths lessons in grades 1 to 3 was replaced by a chess-based lesson. Using administrative data and two statistical methods, the authors "tend" to find positive effects on maths, larger for children who felt unhappy or bored at school (Rosholm and others, 2017).

It is an interesting result, but it is not a classic randomised trial. And notice a detail: it did not add chess hours on top of maths; it changed how one maths lesson was taught. What worked may have been making maths more playful, not chess itself.

The Tenerife study: useful, with limits

This study is quoted a lot in Spain. It compared 170 schoolchildren aged 6 to 16 doing after-school chess with 60 doing football or basketball, measuring at the start and end of the school year. It found gains in cognitive skills, coping, problem-solving and social-emotional development (Aciego and others, 2012).

Its limits: it was not randomised (children chose the activity, and those who choose chess may differ from the start), the comparison group did sport rather than another mental activity, and the teacher's report was not blind. The authors acknowledge that results depended on the pupil's profile. Nor is it a study of maths grades.

Correlation is not cause: strong players and numbers

It is true that strong chess players tend to have somewhat higher cognitive ability. A meta-analysis found an average correlation of 0.24 between chess skill and various cognitive abilities, stronger with numerical ability than with verbal ability (Burgoyne and others, 2016). Another found that players outperform non-players on skills related to intelligence (Sala and others, 2017).

But a correlation runs both ways. Chess may train those abilities, or children who find numbers easier may enjoy chess more and go further. Correlational studies cannot tell the difference. To know whether something causes an improvement you need an experiment, and the best experiment we have (the English one) did not find it.

The European Parliament declaration: what it is and what it isn't

Many leaflets say "the European Parliament recommends chess in schools". That is half true. It refers to Written Declaration 50/2011, backed in 2012, which asks the Commission and the member states to encourage the "Chess in School" programme and fund it (official text; transcription consulted). It says chess "can improve children's concentration, patience and persistence".

What it is not: a law, a scientific evaluation or a research finding. A written declaration is a non-binding political request, in this case promoted by chess organisations. In fact the text itself calls on the Commission to take into consideration "the results of any studies on the effects of this programme on children's development". Quoting it as proof that chess improves maths is like quoting a manifesto as if it were an experiment.

Five questions for reading any study

Next time you read "it's proven", ask:

  1. Was there a control group? Without one, you cannot know whether the children would have improved anyway over time.
  2. Was it active? If the other group did nothing new, novelty could explain the gain.
  3. Was entry decided by lottery? If children chose, they may already have been different.
  4. How many children, and who measured? Twenty children and an instructor who knows who is who are not worth the same as thousands of pupils and an external exam.
  5. Who is telling you? A press release, a leaflet or a political declaration is not a study.

The practical conclusion

With what we know today, chess is not a shortcut to better maths grades. If your child needs help with maths, what works is practising maths.

And still, it is well worth it. It is a deep game that a 6-year-old can start and an adult never finishes. It teaches you to look before you move, to lose and try again, to enjoy thinking. Many children love it (half the pupils in the English trial said so). And it is a great thing to share as a family. Those reasons are enough.

If you want to start, Chess Academy Kids has a lesson-based course with a knight professor who explains everything with a voice, and puzzles with hints to practise. We do not promise better grades; we promise chess that is well explained. And if your child finds it hard to stay focused, see what to do if your child can't focus.

To see which other benefits have more or less support, read benefits of chess for kids, and to get off to a good start, how to teach chess at home.

Is it proven that chess improves maths?

No. There are older studies with moderate effects, but their designs were weak. When chess is compared with other games the effect disappears, and the largest randomised trial, with 4,009 pupils, found an effect of 0.01.

So is chess useless?

Far from it, as a game: it is fun, demanding and teaches you to think before acting on the board. What has not been shown is that this carries over to grades in other subjects.

What is an active control group?

It is a group of children doing another new and appealing activity, such as checkers or go, while another group learns chess. It tells you whether the gain comes from chess or simply from doing something new and motivating.

Sources

All accessed on 2026-10-26.

  • Sala, G., and Gobet, F. (2016). Do the benefits of chess instruction transfer to academic and cognitive skills? A meta-analysis. Educational Research Review, 18, 46-57. https://doi.org/10.1016/j.edurev.2016.02.002
  • Sala, G., and Gobet, F. (2017). Does chess instruction improve mathematical problem-solving ability? Two experimental studies with an active control group. Learning & Behavior, 45(4), 414-421. https://doi.org/10.3758/s13420-017-0280-3
  • Sala, G., and Gobet, F. (2017). Does far transfer exist? Negative evidence from chess, music, and working memory training. Current Directions in Psychological Science, 26(6), 515-520. https://doi.org/10.1177/0963721417712760
  • Jerrim, J., Macmillan, L., Micklewright, J., Sawtell, M., and Wiggins, M. (2016). Chess in Schools: Evaluation report and executive summary. Education Endowment Foundation. https://files.eric.ed.gov/fulltext/ED581100.pdf
  • Jerrim, J., Macmillan, L., Micklewright, J., Sawtell, M., and Wiggins, M. (2017). Does teaching children how to play cognitively demanding games improve their educational attainment? Journal of Human Resources, 53(4), 993-1021. https://doi.org/10.3368/jhr.53.4.0516.7952r
  • Sala, G., Foley, J. P., and Gobet, F. (2017). The effects of chess instruction on pupils' cognitive and academic skills: State of the art and theoretical challenges. Frontiers in Psychology, 8, 238. https://doi.org/10.3389/fpsyg.2017.00238
  • Rosholm, M., Mikkelsen, M. B., and Gumede, K. (2017). Your move: The effect of chess on mathematics test scores. PLoS ONE, 12(5), e0177257. https://doi.org/10.1371/journal.pone.0177257
  • Aciego, R., García, L., and Betancort, M. (2012). The benefits of chess for the intellectual and social-emotional enrichment in schoolchildren. The Spanish Journal of Psychology, 15(2), 551-559. https://doi.org/10.5209/rev_sjop.2012.v15.n2.38866
  • Burgoyne, A. P., Sala, G., Gobet, F., Macnamara, B. N., Campitelli, G., and Hambrick, D. Z. (2016). The relationship between cognitive ability and chess skill: A comprehensive meta-analysis. Intelligence, 59, 72-83. https://doi.org/10.1016/j.intell.2016.08.002
  • Sala, G., Burgoyne, A. P., Macnamara, B. N., Hambrick, D. Z., Campitelli, G., and Gobet, F. (2017). Checking the "academic selection" argument. Intelligence, 61, 130-139. https://doi.org/10.1016/j.intell.2017.01.013
  • European Parliament. Written Declaration 50/2011 on the introduction of the programme "Chess in School" in the educational systems of the European Union. Official text: https://www.europarl.europa.eu/doceo/document/DCL-7-2011-0050_EN.html. Transcription consulted: https://en.odfoundation.eu/a/2203,written-declaration-on-the-introduction-of-the-programme-chess-in-school-in-the-educational-systems-of-the-european-union/

What do you think?

Tell us what you like, what's missing or what you'd improve. We read everything.

Your email app will open with the message ready to send.