Does brainmaxing work? The evidence, graded

“Brainmaxing” is a label the internet put on a pile of different practices: sleeping properly, exercising, studying differently, playing timed tasks, and taking things. Asking whether it works is like asking whether a supermarket works. The parts have different answers, and mixing them is how a strong claim about the cheap part ends up selling the expensive one.

So this page grades them separately. A grade here belongs to one claim, in one population, for one preparation — never to a practice as a whole and never to this article. The table at the foot is the part designed to be quoted, because it carries the population and the weakest link that a quotation normally strips.

Sleep first, and it is not close. Pooling seventy articles and 147 cognitive tests, Lim and Dinges (2010) found that short-term total sleep deprivation, under 48 hours, degraded simple attention, working memory and processing speed, with the largest effect on lapses in simple attention [3]. That is a literature about the cost of losing sleep rather than a promise about gaining any, which is what the grade below says — but it means one night without sleep is visible on tasks as plain as pressing a key when a light changes.

Movement next. Ludyga and colleagues (2020) pooled 80 randomised trials of long-term exercise in healthy individuals and reported a small summary effect on cognition, moderated by who was exercising, by the type of exercise and by the length of the programme and its sessions [4]. Small, real, and free.

Then how you study, which is the part you can change today. Dunlosky and colleagues (2013) rated ten study techniques and gave practice testing and distributed practice their highest utility rating, with summarising, highlighting and re-reading among the five rated lowest [5]. Cepeda and colleagues (2006) pooled 839 assessments of distributed practice from 317 experiments and found that the gap between study sessions and the time until the test work together: the longer you need to remember something, the longer the gap that gives the best retention [6]. These are claims about retaining what you studied, which is narrower than “better thinking” and more useful than it.

Brain games are where the evidence is weakest. Practice on a task improves that task; training on one task moves performance on tasks that look very like it, modestly and by an amount that depends on who is training [2]. Reaching further — to untrained abilities, to school results, to work — is where it stops: once placebo effects and publication bias are allowed for, the estimated effect is close to zero [1][2].

That is not a fringe reading. It is where a comprehensive review of consumer brain training landed: extensive evidence that training improves the trained tasks, less for closely related tasks, and little for distantly related tasks or everyday cognitive performance [1]. It is the reason this site reports what you did on a task rather than what a task is supposed to do for you.

Supplements are the part this page does not grade, and that is on purpose. A verdict on supplements as a category would be an article-level grade in another form: it would let the weakest compound borrow the strongest one's evidence, or the other way round. So each compound this site covers has its own note in the Fuel section, with a grade per outcome, per population and per preparation, and its sources on the same card. If you want to know about one, read its card. This page makes no claim about any of them.

And two things that get sold as brainmaxing have no honest place here at all: anything measured as an intelligence score, because there is no norming sample behind a browser task and nobody should be sold a number about themselves that cannot be sourced; and anything promising an outcome in school, work or health, which is a claim this product does not make and would not know how to support.

A grade is the weakest of four links: how good the studies are, whether the preparation studied is the one you would use, whether the population studied is one you belong to, and whether the outcome measured is the thing you actually care about. A row graded moderate with “population match” as its weakest link is telling you that the studies are fine and the people in them may not be like you.

That is why you will not find a single letter at the top of this page. An article-level grade would let its weakest claim borrow the strength of its strongest, which is the exact move this whole page exists to refuse.

Protect your sleep first: losing it carries the best-graded evidence on this page, and protecting it is free. Move most days at a dose you will still be doing in a month. If you are learning something, test yourself instead of re-reading it, and space the attempts. Those three are the whole of the well-graded part, and none of them requires a purchase.

Play the timed tasks because a repeatable number about a narrow thing is useful — set against your own earlier runs, it can show you how that one task moves after a night without sleep — and not because the number is going to become something else. That is the honest version of the answer, and it is also the least sellable one, which is roughly how you can tell.

A grade belongs to one claim, in one population, for one preparation — never to a compound, a technique or an article. The weakest link is the part of the chain between the evidence and the thing you would actually do that is doing the least work.