Do Puzzles Make You Smarter? An Honest Answer
We run a puzzle site, so we would love to say yes. The evidence says something more modest, and in some ways more useful.
It would be very convenient for a puzzle site to tell you that ten minutes a day will sharpen your mind, protect your memory and add points to your IQ. Plenty of people have made that pitch. We are not going to, because the evidence does not support it, and you deserve a straight answer.
What the research says
The question researchers care about is called transfer. Nobody doubts that practising a task makes you better at that task. Do a Schulte Grid every day for a fortnight and your times will fall. The question is whether that improvement carries over to anything else: remembering names, concentrating at work, staying sharp in old age.
On the whole, the answer has been disappointing. In 2014 a large group of cognitive scientists and neuroscientists signed a public statement, organised through the Stanford Center on Longevity, objecting to claims that brain games offer a scientifically grounded way to reduce or reverse cognitive decline. In 2016 the company behind one of the best-known brain training programmes agreed to pay two million dollars to settle charges from the US Federal Trade Commission that its advertising had gone beyond what its evidence could support.
That does not mean puzzles do nothing. It means the strong claims have not been proved. You get better at what you practise, and the benefit mostly stays close to home.
So why bother?
Because "it will not raise your IQ" is a strange reason to give up something enjoyable. Nobody asks whether a walk in the park transfers to other skills. Here is what we think puzzles honestly offer.
A clean break. A puzzle occupies your attention completely for a few minutes. You cannot solve a Binary Grid and worry about your inbox at the same time. For a lot of people, that is the whole appeal: a short spell of thinking about one thing only.
A finish line. Much of ordinary work has no clear end. A puzzle does. The last square goes in, the tick appears, and something is actually done. That small feeling is worth having in a day.
Practice at being stuck. This one is underrated. Every decent puzzle puts you in a spot where nothing seems possible, and then shows you that something was. Getting used to that feeling, and learning not to panic in it, is a habit rather than a measurable skill. We still think it is a good habit.
Particular methods. Some puzzles teach a way of thinking that you can name. Two Jugs is about measuring differences. Tower of Hanoi is about breaking a problem into smaller copies of itself, which is an idea programmers use every day. You may never need them. They are still nice things to own.
A fair way to use the memory games
The two memory puzzles in the notebook report a number: a time for the Schulte Grid, a level for Pattern Recall. It is tempting to read those as a score for your brain. Please do not.
They are better used as a weather report. Your result on a given day says more about how you slept and how distracted you are than about your abilities. Played that way, they are quite useful: a thirty-second check on whether now is a good time to start something difficult.
What we promise instead
GridPencil will not tell you that its HTML5 games make you cleverer. What we can say is narrower and true:
- Every puzzle has a real solution that you can reach by reasoning.
- You will get better at each one the more you play it.
- It costs nothing and takes about five minutes.
If you finish a puzzle and feel a little pleased with yourself, that is the product working as intended. Anything beyond that is a bonus we cannot guarantee, and we would be wary of anyone who says they can.
If your interest is in keeping your mind healthy over the long term, the dull advice has far better evidence behind it than any game: sleep, exercise, company and learning things that are properly new to you. For anything that worries you about your memory, talk to a doctor, not a website. Then, by all means, do a puzzle.