Visual Memory Test

Every other memory test here feeds you information in an order. This one shows you everything at once, and that changes what is being measured. There is no sequence to repeat and no rhythm to lean on. You get a pattern. Either it fits in your visual memory or it does not.

That makes it a cousin of spot-the-difference experiments rather than of span tests. It also explains why people fall apart so suddenly. Going from eight lit squares to nine is not ten per cent harder. It is one step past a limit, and that limit is sharper than most people expect.

You can click the lit squares in any order, and that is deliberate. Demanding a set order would sneak sequence memory into a test of pattern memory. The two really are separate skills. People who are excellent at sequence memory are often ordinary here, and the reverse happens just as often.

Each level is also all or nothing, again on purpose. Award part marks and clicking half the grid starts earning points. The question worth asking is whether the whole pattern survived, not how many lucky squares did.

Almost everybody who does well here is doing something other than storing single squares. They read the pattern as a shape. They spot a symmetry. They split the grid into quarters and count each quarter. Or they name the shape after something it looks like.

This is not cheating. It is what visual memory does. Cowan's limit of about four items applies to items with nothing in common. A pattern that makes a recognisable shape counts as one item, not seven.

So if your score is climbing fast between runs, you are almost certainly getting better at grouping rather than holding more. How working memory works covers why that difference matters.

Level ten is around the middle of the reference distribution. Level fourteen is strong. The grid grows as you climb, so a level is not a fixed step of difficulty. Level twelve is harder than level ten, but by no defined amount.

Three lives per run also makes this score jumpier than a span test. One distracted level can cost you two or three. Take the middle of three runs if you want a number to keep.

Compare it against your digit span and Corsi results from the same sitting rather than against a stranger's screenshot. What counts as a good digit span explains why a span measures your technique as much as your capacity.