Chimp Test
Memorize the numbers, click 1 — the rest hide. How far can you go?
🎮 Our new game is coming
Town Guardians launches in 2026 — drop your email and be first to play.
More from Pong Studio →About the Chimp Test
Tap the tile marked 1 and every other number on the board disappears. From there you have to place 2, 3, 4 and onward from memory, in order, on a grid you were only allowed to look at before you started. The name is not marketing — it comes from a real memory study, and the full story behind it is worth knowing before you play.
How it works
- Numbers are laid out across a 5×5 grid, 25 tiles in total. The first round shows 4 numbers; each round you clear adds one more, up to a maximum of 20.
- Look at the board for as long as you want — nothing is timed until you act. Tap the tile marked 1 to begin, and the rest of the numbers vanish immediately. You then have to tap 2, 3, 4 and so on, in order, purely from where you remember them being.
- You get 3 lives. Tap the wrong tile and the board reveals what it actually was, you lose a life, and you repeat the same round — same count of numbers, new positions — rather than being sent back to the start. Lose all 3 lives and the run ends.
Your score is the highest count of numbers you ever placed correctly in full, not the count you happened to be on when your lives ran out.
What the number actually tells you
Because a wrong tap costs a life and replays the round instead of ending it, this score reflects your best run across up to three tries per level — not your first attempt. That is a real design choice, not a flaw: it means one shaky tap early on doesn't erase a genuine result later. It also means two players who both reach 12 didn't necessarily play the same way. One might have cleared every round first try; the other might have needed all three lives on half of them. The final number can't tell you which.
It's also worth being exact about what this page doesn't reproduce. The study this game is named after flashed its numbers for a fixed 210 milliseconds before hiding them — a single glance, no choice about pacing. Our version lets you study the board for as long as you like before you tap 1. That makes for a fair game, but it is not the same task, and a score here is not a number you can hold up against the study's results.
The chimp that beat the humans, and the humans who beat it back
In 2007, Sana Inoue and Tetsuro Matsuzawa published a study in Current Biology testing chimpanzees and university students on a masked-number task much like this one. A young chimpanzee named Ayumu held around 80% accuracy even when the numbers were shown for only 210 milliseconds. Students given the same fraction of a second fell to about 40%. It ran around the world as proof a chimpanzee's short-term memory could beat a human's, and it is where this game gets both its name and its premise.
In 2010, Cook and Wilson published a rebuttal, in Science and separately in Psychonomic Bulletin & Review, that most retellings of the story leave out. They gave human participants the amount of practice on the task that Ayumu had already had before he was ever tested. The gap closed. Trained humans matched Ayumu's accuracy, and some beat it. The original result wasn't a species advantage — it was a practice advantage that the 2007 comparison hadn't accounted for.
Most pages that borrow this game's name repeat the 2007 headline and stop. We'd rather tell you both halves, because the second one is the more interesting finding: memory this good is achievable, and Ayumu's edge came from years of drilling this exact task, not from being a different species. Our own Elite and Pro labels here are tiers we made up for a browser game, not thresholds from either paper — the real spread of scores, and how many people they're drawn from, is on our stats page.
Why one bad round doesn't mean much
- This is a span test: your score is the point where memory finally failed, not an average of how you performed. A single lapse of attention on an otherwise good day can cost you several numbers, and that's a much noisier signal than a timing test, where one slow click barely moves an average.
- Grouping the numbers into a rough path or shape instead of memorizing each tile in isolation is the obvious strategy, and most players find it helps. It isn't cheating — it's exactly the kind of trick that makes this a memory-and-strategy game rather than a pure capacity test.
If your score jumps around between attempts, that's expected. Play it a few times before you treat any single number as your ceiling.
Frequently asked questions
Why is it called the Chimp Test?
It's named after a 2007 study by Inoue and Matsuzawa at Kyoto University, in which a chimpanzee named Ayumu recalled masked numbers more accurately than university students shown the same brief glance. The name refers to that experiment, not to any claim that this game measures the same thing the study did.
Did a chimpanzee really beat humans at memory?
In the original 2007 comparison, yes. But a 2010 follow-up by Cook and Wilson found the gap came from unequal practice — Ayumu had trained on the task far longer than the students had. When humans were given comparable practice, they matched or beat him. Both results are real; the second one is the reason the first shouldn't be repeated on its own.
What is a good chimp test score?
We don't have a research benchmark for this exact browser format, so we can't give you a scientifically defensible number. This page used to say most players reach 8–10 and that the chimps in the original study "handled 9 with ease" — neither figure had a source, and we've dropped them. The Elite, Pro and Sharp labels you see after a run are tiers we set ourselves for this game, not measurements from any study. The actual distribution of scores from real players, and how many players it's based on, is on our stats page.
Why does a wrong tap replay the round instead of ending the game?
You get 3 lives. A mistake costs one, reveals the board, and repeats the same count of numbers with new positions — it doesn't send you back to 4 numbers or end your run outright. Your recorded score is the highest count you ever completed in full, so a rough patch early doesn't erase a strong finish later.
Is this the same task as the original study?
Not exactly. Inoue and Matsuzawa's version flashed numbers for a fixed 210 milliseconds before masking them. This game lets you look at the board for as long as you want before you tap 1 to start. That's a real difference in how much control you have over the memorization phase, so treat this as a game inspired by the study, not a browser copy of it.
How can I remember more numbers?
Grouping the numbers into a rough shape or path, rather than memorizing 25 tiles independently, is the strategy most players land on. It isn't a trick against the game — it's a normal memory technique, and using it is part of what the game is testing.