Skip to content
Lizely

Pattern Memory Game

Memorize five original 3×3 patterns, choose each exact hidden record from four options, and finish with exactly 1,000 points.

Privacy: your files never leave your device. All processing happens locally in your browser.

How to use

  1. 1.Memorize the filled and empty cells in the 3×3 target during its 1.5-second Reveal phase, or press Hide now when ready.
  2. 2.After the target is hidden, choose the identical grid from four options with pointer controls or keys 1–4.
  3. 3.Match all five original records. Each exact choice awards 200 points, producing a strict final score of 1,000.

About Pattern Memory Game

Pattern Memory Game is a five-record visual recall challenge built from original three-by-three binary grids. A filled cell and an empty cell are both part of the answer. Each round first displays one target pattern by itself. After 1.5 seconds, or when you press Hide now, the target disappears and four candidate records become available. Your task is to choose the exact arrangement you just saw.

The five targets and every distractor are fixed literal fixtures. They are encoded as nine disclosed filled-or-empty cells rather than images or generated art. The correct answer position changes across the route: option 3 in round one, option 1 in round two, option 4 in round three, option 2 in round four, and option 3 in round five when choices are numbered one through four.

The option tray is not revealed while the target is visible. During the Reveal phase it uses uniform placeholders, so the target cannot be compared side by side with the correct candidate. Once the record is hidden, all four real grids appear in a stable layout. This preserves the intended memory step while avoiding layout shift between phases.

The reveal transition uses a single cancellable timeout and a pure elapsed-time reducer. It does not read time during React rendering, create an animation loop, or accumulate intervals in the background. Hide now provides a deterministic alternative for players who do not want to wait and for exact browser verification. A new 1.5-second reveal begins after every correct choice.

Pointer and touch players use the four large option buttons. Keyboard players press H or Space to hide the current target, then keys 1 through 4 to choose a record. R restarts when focus is outside a button. The shared game shell adds a visible Restart control, best-score storage, and the double-Escape Boss Key. Focus treatments and minimum 44-pixel targets support keyboard and touch navigation.

Each exact match awards 200 points and opens the next reveal. Five correct choices therefore produce exactly 1,000. One distinct mismatch is recoverable. Choosing the same wrong option again on the same round is deduplicated and leaves the error count unchanged. A second different wrong option, whether in the same or a later round, closes the memory file until Restart.

Choice actions during Reveal are atomic no-ops. Negative, fractional, out-of-range, or unknown option indexes are rejected. Nonfinite, zero, or negative timer advances are also no-ops. Completion and deadlock freeze timer, hide, selection, and scoring actions. Restart returns to round one, full 1.5-second reveal, zero score, no errors, and no previous result.

Independent tests own all five target strings, all twenty candidate strings, and all five correct indexes as literal fixtures. They confirm the four deterministic answer positions, exact timer boundaries, Hide now, the complete five-step 1,000 route, repeated-error deduplication, second-error deadlock, illegal inputs, phase restrictions, Restart, and terminal freezing. Expected patterns are not generated from the production answer function.

All work stays in the browser. There is no account, upload, eye tracking, camera, API, model, random service, or new dependency. The score records completion of these five disclosed patterns only. This is entertainment, not a medical or psychological screen, standardized memory assessment, hiring test, educational placement decision, or statement about a player's cognitive ability.

Methodology & sources

Encode five original 3×3 targets and four literal candidate strings per round, with correct positions varying across all four indexes. Separate Reveal and Choose phases; expose a pure timer reducer and Hide now action, and display real candidates only after the target is hidden. Award 200 for each exact index and exactly 1,000 after five, deduplicate a round-and-option error signature, deadlock after the second distinct mismatch, reject illegal timer and option input atomically, and freeze terminal states. Independently restate all twenty-five patterns, answer indexes, timer boundaries, and full win/loss routes.

Frequently asked questions

Can I compare the target with the answer choices?
No. The real choices appear only after the target is hidden; the Reveal phase shows uniform placeholders in their place.
How long is each pattern visible?
Each target is shown for 1.5 seconds, and Hide now lets you move to the choices sooner.
Are the patterns generated randomly?
No. All five targets, twenty candidates, and answer indexes are fixed original fixtures verified by independent literal tests.

Related tools

Mini Games guides

View all