A word list recall test built around Word Chain Memory runs across five rounds, each one adding exactly one new word to a growing ordered chain that the player must then type back from memory. The fixed sequence, beginning with the single-token file and ending at the five-token file, is ember; ember harbor; ember harbor velvet; ember harbor velvet cactus; and ember harbor velvet cactus orbit, and finishing all five rounds yields exactly 1,000 points at 200 per round. The test is deterministic: the chain is fixed, the score formula is fixed, the keyboard flow is fixed, and the only variable is how accurately the player recalls the ordered chain. Two distinct wrong recalls trigger a visible deadlock, one recoverable mistake is allowed, and case plus repeated whitespace are normalized so EMBER HARBOR matches ember harbor without changing where one word ends and the next begins. Each round is short, the entire game fits in a single browser break, and the answer route is identical on every load so the score is auditable rather than lucky.
The fixture is original product content rather than a vocabulary standard, educational assessment, or imported word list. Those ordinary tokens are used only as game pieces, and they make no factual category, spelling-level, diagnostic, or language-proficiency claim. Players come to the test for the ordered-recall mechanic itself, not for graded difficulty or external meaning.
What a Word List Recall Test Looks Like in Word Chain Memory
The structure of the test is unusually simple to describe because every variable except recall accuracy is held constant. Five immutable cumulative word arrays contain one through five original tokens, and each round reveals a chain that contains every earlier word plus one new word. Round one shows ember. Round two shows ember harbor. Round three adds velvet for ember harbor velvet. Round four adds cactus for ember harbor velvet cactus. Round five adds orbit and closes the file at ember harbor velvet cactus orbit.
Because the chain only grows and never rearranges, the difficulty curve is dictated entirely by the length of the chain rather than by content variation, time pressure, or surprise words. There is no timer bonus, no random multiplier, no speed penalty, and no hidden difficulty adjustment. The same five rounds appear in the same order on every load, which keeps the score auditable and lets the player focus on ordered memory rather than luck.
How the Reveal and Hide Phases Work
Reveal and recall are separate interface states, and the difference between them is what makes the test a real recall test rather than a copying exercise. During Reveal, the current ordered list is visible as individual word blocks, and the player reads the chain from left to right. Choosing Hide words and recall, or pressing the H key, unmounts that list from the document before the answer input appears.
Recall does not merely blur, cover, recolor, or visually move the answer off screen. Inspecting the active recall region does not expose the sequence. The visible chain is removed from the page, so the only way to pass the round is to reconstruct the chain from memory and type it back in. This separation is what gives the word list recall test its measurable memory load instead of being a transcription task.
The keyboard flow is complete. Press H to remove the revealed chain, type the answer into the normal text field, and press Enter to submit. Pointer and touch users get the same behavior through labeled Hide and Check controls, each sized with a practical minimum touch target so the test can be played on a phone.
How to Run a Word List Recall Test in Five Rounds
- Open Word Chain Memory and wait for the first Reveal state, which shows the single word ember as an individual word block.
- Read the current chain from left to right, then press H or choose Hide words and recall to remove the visible list and reveal the answer input.
- Type the complete ordered chain for the current round, separated by single spaces; for round one that is ember, and the validator accepts any case and any amount of repeated whitespace between tokens.
- Press Enter or choose Check to submit the recall. If the chain matches exactly, the score advances by 200, the next Reveal state appears with one more word, and the round counter increments.
- Repeat the read, hide, and type cycle for the two-word chain ember harbor, then for ember harbor velvet, then for ember harbor velvet cactus, and finally for ember harbor velvet cactus orbit.
- Finish all five clean recalls to land on a final score of exactly 1,000 points, or use Restart at any point to return to round one with zero score and zero errors.
Each correct chain awards exactly 200 points and advances to a new Reveal state, so five clean recalls produce exactly 1,000 points. If you make a recoverable mistake, finish the round, fix your recall on the next attempt, and continue. If a second different wrong sequence appears, the run closes in a visible deadlock and the only recovery path is Restart.
Scoring, Errors, and the Two-Mistake Deadlock
The scoring formula is the simplest part of the contract and worth internalizing before you start a run. Each round is worth 200 points, there are five rounds, and there are no modifiers. 200 plus 200 plus 200 plus 200 plus 200 equals 1,000, which matches the verified five-round route through the 200, 400, 600, 800, and 1,000 score milestones. A perfect run therefore ends on a round number that is easy to remember and easy to compare against the saved best score in the same browser.
| Round | Revealed chain | Length | Cumulative score on correct recall |
|---|---|---|---|
| 1 | ember | 1 word | 200 |
| 2 | ember harbor | 2 words | 400 |
| 3 | ember harbor velvet | 3 words | 600 |
| 4 | ember harbor velvet cactus | 4 words | 800 |
| 5 | ember harbor velvet cactus orbit | 5 words | 1,000 |
Error handling is also explicit. A first distinct wrong recall is recoverable, so the round does not end on the first mistake. The same normalized wrong sequence submitted again is recognized as a repeat and does not consume another error, which prevents accidental double-counting from a single typo. A second different wrong sequence closes the run in a visible deadlock, after which answer and phase actions freeze, and Restart is the only way back to round one.
The independent logic test confirms all of this behavior. It restates all five chains as literal fixtures rather than computing expected answers from the game table, executes the complete 200 to 1,000 score route, checks case and whitespace normalization, proves glued words stay glued, verifies blank and invalid input atomicity, repeats the same wrong answer, triggers a second-error deadlock, and confirms both the terminal freeze and the deterministic restart.
How the Validator Handles Capital Letters, Spaces, and Glued Words
The validator trims leading and trailing whitespace, accepts repeated spaces between tokens, and compares ASCII letters without case sensitivity. EMBER HARBOR therefore matches ember harbor, and ember harbor with two spaces matches ember harbor with one space. Case and whitespace are the two properties the validator normalizes, and they are the only ones, so any other input difference becomes a real difference.
Word boundaries are never invented. emberharbor stays a single token and cannot stand in for ember harbor. The validator splits only on whitespace, so two words glued together count as one long word, and any chain whose split-on-whitespace length does not match the expected chain length cannot be correct, regardless of the underlying letters. Punctuation, digits, and blank submissions are rejected as harmless no-ops, which means typing nothing, typing a number, or typing punctuation alone does not consume an error and does not advance the score.
This combination of normalization and strict boundaries is what turns the test into a genuine word list recall test rather than a fuzzy matching exercise. The player must remember the order and the boundaries of every word, and small allowances on case and spacing keep the focus on memory rather than on typing precision.
Other Browser Memory Tests Worth Trying
Word Chain Memory is one of several short memory challenges in the same browser collection, and the differences between them are useful for picking the next thing to play. Number Memory Test uses fixed digit strings instead of word chains and follows the same reveal, hide, and keyboard-entry flow without a timer. Backward Digit Span asks for the digits in reverse order, which adds a transformation layer on top of the same ordered recall mechanic. Visual Memory Test keeps the recall load on a grid pattern instead of a typed answer, while Pattern Memory Game tests pattern recognition from four multiple-choice options.
None of these alternatives use a timer bonus or a random multiplier, which keeps the comparison straightforward. Pick the one whose recall material matches what you want to practice, and the same audit-style scoring will apply.
Everything in the test runs locally in the browser. There is no account, no upload, no microphone, no camera, no remote word service, no network request, and no downloaded asset, and the game stores only the shared best completed score in the same browser. Current input and round state disappear when the page is closed or restarted, so a word list recall test in Word Chain Memory is a clean, repeatable short break rather than a persistent profile.
For a deeper look, see Word Families vs Morphological Families: 6 Differences.