A verbal memory test shows you words one at a time and asks a single question about each: have you seen this one before? The list never stops growing, so the load rises continuously until it beats you. Everyone loses eventually — the score is where, not whether. You can take it on [Verbal Memory](/games/verbal-memory), and it takes about two minutes.
Most memory tests give you a fixed amount of material: seven digits, a grid of cells, a list of ten words. You either hold it or you do not, and the test ends.
A verbal memory test has no fixed list. It keeps adding words while you are still responding, so what it measures is not capacity but how gracefully your memory degrades as the load grows past capacity. That makes it one of the most demanding formats there is, and it is why the scores people reach surprise them in both directions.
Recognition, not recall
You are never asked to produce a word — only to say whether you have met it. Recognition is a much lower bar than recall, which is why the test stays playable at loads where a recall test would collapse. Your recognition vocabulary is several times larger than your production vocabulary, and the same gap applies to memory for a list.
Stop trying to hold a list
This is the single most important thing. People who try to maintain an explicit list of everything they have seen fail at around forty words, without exception, because maintaining a list of forty items is not something anyone can do.
People who reach high scores rely on a feeling of familiarity instead. Recognition memory can operate without any recollection of the event — you simply know you have met the word, without being able to say when. That feeling is far more capacious than deliberate rehearsal, and acting on it beats deliberating.
Answer fast
Deliberating about familiarity makes it less reliable, not more. The first impression is the familiarity signal; anything after that is reasoning, and reasoning about whether you saw a word is not something you have evidence for.
Note the unusual ones deliberately
Unusual words are both the ones you can actually hold and the ones most likely to come back. Spending a fraction of a second on a distinctive word is worth more than spreading attention evenly.
Expect the middle to be where you fail
The serial position effect applies here as everywhere: the first words got rehearsal and the most recent are still active, leaving the middle of your history least protected. Most errors are words from the middle of a long run.
Verbal memory in the clinical sense — list learning, delayed recall — is a standard part of neuropsychological assessment, and difficulty with it is one of the changes people notice first in themselves. This game is not that assessment and no score here means anything clinically, but the underlying ability is the same one.
If you want the closer analogue of the clinical task, Word Recall presents a fixed list and asks you to produce as many words as you can, and Audio Word Memory reads the list aloud so you cannot re-read it. Both are recall rather than recognition, and both are considerably harder.
Switch to familiarity. If you are currently trying to hold a list, this one change will move your score more than anything else.
Speed up your responses. Slow answers are reasoned answers, and reasoning is worse than the signal here.
Build a story for the distinctive words. Narrative encoding is the strongest mnemonic technique there is, and it works even applied selectively.
Play regularly rather than long. Five minutes daily beats thirty minutes weekly for every task on this site.
The other word games worth pairing it with are Word Recall, Audio Word Memory and Missing Object, which tests the same familiarity signal on images rather than words. They all live on the memory games hub.
Words appear one at a time and you say whether you have seen each one before in the run. The list keeps growing, so the load rises continuously until it beats you.
Compare against your own history. Scores depend on presentation speed and on how many lives the version allows, so cross-site numbers are not comparable.
By relying on familiarity rather than trying to maintain a list. An explicit list fails at around forty words for everyone; familiarity scales much further.
On a genuinely uncertain word, yes — but your sense of familiarity is a better guide than any reasoning you can do about it.
No. Verbal list learning is used in neuropsychological assessment, but a browser game is not that assessment and no score here means anything diagnostically.
The trick to a verbal memory test is counterintuitive: try less. Stop maintaining a list, answer on the first flicker of familiarity, and let a system that is much larger than your deliberate memory do the work. [Take the test](/games/verbal-memory) and see how far it carries you.