Narrative voice and pacing
The Tell-Tale Heart
Edgar Allan Poe · 1843
Follow how repetition and a restricted point of view build tension. A different voice can achieve equally strong control.
Explore the referenceHOW EVALUATION WORKS
The work comes first.
Clear criteria, evidence from the work, and room to question a result.
Our evaluation today.
CWA uses AI-assisted evaluation of submitted writing. For rolling entries, a validated initial AI score and feedback become privately available after processing. Eligible scores can join the next monthly leaderboard with your permission; there is no routine manual approval or formal award. If assessment is paused or the work cannot be read, we show its status rather than invent a score. The competition rules determine eligibility, entry and participation arrangements.
Using AI to evaluate a work is separate from using AI to create it. Read the rules for permitted assistance and disclosure before submitting.
Read the competition rulesFive dimensions. A score you can understand.
Each dimension is scored from 0 to 100. The total is the weighted sum, calculated by the platform. Your report records the exact rubric used; an existing competition may use an earlier version.
Original expression
Specific choices, a developed perspective and purposeful use of familiar material. Do not award points for novelty claims or resemblance to a famous author.
Craft and technique
Control of the chosen form: imagery and rhythm in poetry; scene, voice and pacing in narrative; reasoning and evidence in essays.
Emotional impact
How concrete details, stakes and restraint earn the intended response; serious subject matter alone is not evidence of quality.
Structure and coherence
Relationships between parts, development and an earned ending. Experimental structures can succeed without conventional plot.
Language and style
Precision, rhythm and consistency of voice. Evaluate intentional dialect and multilingual choices in context.
90–100 · ExceptionalSustained, precise control with compelling evidence across the work; only minor limitations.
75–89 · StrongClear purpose and effective execution, with a few specific areas to develop.
55–74 · DevelopingRecognisable strengths alongside uneven execution or development.
35–54 · EmergingSome purposeful choices, but substantial gaps limit the intended effect.
0–34 · Early explorationLimited demonstrated control in readable, eligible work. Missing or unreadable work receives no score.
LEARNING FROM HISTORICAL WORKS
Study the choices. Find your own voice.
These famous works illustrate techniques in our framework. They are not scored reference entrants or a validated percentile database. We do not assign them invented scores or reward imitation.
Character, dialogue and social observation
Pride and Prejudice
Jane Austen · 1813
Study how dialogue reveals competing intentions and changes relationships; ornate or historical language is not a requirement.
Explore the referenceUnderstand your comparison group.
For rolling entries, your private report compares your work with the latest approved work from each other participant in the same competition and category, using the same rubric and model version. It shows the real number of people and the observation date. It does not claim an age-matched or worldwide sample.
A position and percentile appear only with at least 10 other comparable participants. Ties share a position; percentiles use the midpoint of ties. Your score stays the same as new peers join, while the live comparison can change.
IN DEVELOPMENT
A clearer path from evaluation to growth.
Read the complete work.
Check that the submitted material can be evaluated properly. Missing or unreadable content should lead to a request for review, not an invented score.
Check differences.
Independent evaluations, calibrated criteria and human review of significant disagreements are part of our planned process.
Explain the next step.
Private reports will connect strengths and suggestions to specific passages or visual evidence, with a link to ask questions through our help center.
A reference point, not a label.
A score describes one work under particular criteria. It does not measure a student’s potential or establish a global ranking. We do not currently offer comparisons against a validated historical reference collection.
Our planned reference collection will use appropriately licensed examples and clearly defined groups. We will explain its scope before offering comparisons.