The example computes gray/yellow/gray/green/green for EAGLE against APPLE. Its candidate check retains APPLE and AMPLE, so the row does not uniquely identify the teacher target. An independent PUPPY round credits two Ps and rejects the extra third copy. Text labels accompany every position; the activity does not depend on color alone or claim to reproduce an official daily-game engine.
Announce the two-pass rule
Use five uppercase ASCII letters and positions 1 through 5. First mark exact matches and remove those copies from the target's available counts. Then visit unmatched guess positions from left to right. A remaining copy earns Elsewhere and consumes one token; no remaining copy earns gray. Keep the wording No remaining copy rather than always saying absent. Gray may indicate a genuinely absent letter, or an extra copy beyond those already credited elsewhere in the same row.
For target APPLE and guess EAGLE, L4 and E5 are exact. After removing them, A1 and two Ps remain. E1 has no remaining E, so it is gray; A2 is elsewhere; G3 has no target copy. The result is gray,yellow,gray,green,green. Ask which copy was consumed before E1 was considered. Pointing to the exact E5 gives a traceable explanation and prevents the mistaken conclusion that no E can appear in an answer.
Recompute the complete row for candidates
Use exactly APPLE,AMPLE,ANGLE,EAGLE,MAPLE as the teacher-reviewed candidate pool. Treat each as a possible target and compute feedback for EAGLE with the same algorithm. APPLE and AMPLE reproduce all five labels. ANGLE would make G3 exact; EAGLE would make every position exact; MAPLE would make A2 exact. Matching only the visible green suffix is insufficient. The rejected candidate needs one concrete conflicting position, and a survivor remains possible rather than established as the actual hidden target.
The page shows the teacher's target openly for debrief; a live learner packet can withhold that separate key until after reasons are written. Do not confuse the open explanation with secure assessment storage. The small pool is a declared task boundary, not a dictionary. Blank, wrong-length, non-ASCII and unlisted guesses are held for teacher review rather than silently accepted as known words or transformed into another string through normalization that changes the intended task.
Count the Ps in a fresh round
For APPLE/PUPPY, first credit P3. One P remains, which P1 consumes as Elsewhere. U2 is gray, P4 is gray because the remaining budget is zero, and Y5 is gray. The row is yellow,gray,green,gray,gray and credits exactly two Ps. Write the remaining P count after the exact pass, after P1, and at P4: 1,0,0. Keep this round independent rather than carrying previous selections or feedback into a new guess.
End with a count or position reason for each decision. The page computes the row and survivor set, while a teacher reviews the learner's written explanation. It has no streak, date-based puzzle, leaderboard or measured vocabulary outcome. Reset restores the first round; editing a guess clears stale results. Preserve the native worksheet and teacher key as the activity record, and use the bounded computation to make repeated-letter reasoning inspectable rather than treating speed as evidence of understanding.
Make your own working example
Build an original classroom two-pass word-feedback explainer. Teacher target APPLE is openly labeled for debrief. Default guess EAGLE. Credit exact matches first, decrement target counts, then scan remaining guess positions left-to-right for Elsewhere copies. Label each tile Exact, Elsewhere, or No remaining copy with position and letter, not color alone.
EAGLE expected gray,yellow,gray,green,green. Compare candidate pool APPLE,AMPLE,ANGLE,EAGLE,MAPLE by recomputing the entire same row with each candidate as target. Retain APPLE and AMPLE only and identify the first differing position for rejected candidates; do not claim a unique deduction.
Add PUPPY demonstration yielding yellow,gray,green,gray,gray, with two credited Ps and remaining counts 1,0,0 after exact/P1/P4. Provide editable guess, Check, EAGLE round, PUPPY round and Reset. Display the five-word approved pool beside the guess input and identify PUPPY as an extra demonstration guess, not a candidate target. For each rejected candidate show its first differing position with candidate and observed text labels. Accept only the five pool words plus PUPPY, five uppercase ASCII letters. Blank, other length, non-ASCII and unknown words are held. Input changes clear results. Keep authored protocol separate from official game behavior, no daily puzzle, scoring, storage or learner-effectiveness claims.
Use one self-contained HTML file with no remote dependency. Keep the initial sample, output, and labeled controls near the top. Use at least 16px text, strong contrast, visible focus, and a stacked layout at 360px without horizontal overflow. Use explicit type=button click handlers and explicit DOM selectors; do not depend on native form submission. Clear stale feedback when task inputs change. Render entered text safely. Keep practice data local to the current page; do not imply saving, sending, or production integration. Return only the finished HTML.
Clarify the duplicate-letter debrief
Keep the learner worksheet and separate teacher key. In an HTML explanation, NOS lets you revise eligible static instructions visually without AI regeneration. Clarify “No remaining copy” so the next group does not interpret a gray extra letter as universal absence.