Advanced · Advanced · 6 rounds · ~8 min · free with an account
Prompt Evaluation Rubrics
Build your own grader. The Chip-Huyen-style eval rubric that catches drift before it ships.
Prompt Arcade · 6 rounds · free with an account
Prompt Evaluation Rubrics
Decide what 'good' means before you judge outputs: a short rubric of checkable criteria with clear levels. Then you can compare prompts with evidence, not vibes.
- Build it
- Spot it
- Order it
- Match it
- Spot it
- Build it
Five games are open to everyone — start with Prompting Mindset 101. The other 31 need a free account (no card) so your scores and progress are saved.
What you'll learn
- Write the rubric: Few criteria, observable, with clear levels.
- Bad criteria: A criterion is good if two people would score it the same.
- Compare two prompts: Rubric first, same inputs, blind scoring, check the worst cases.
- Read the results: Look at averages and at the worst outputs.
- Judge bias: Blind, varied, and rubric-first.
- AI as judge: Anchor the AI judge with examples, then check it against your own scores.
More advanced games
Self-Consistency VotingRun the same prompt 3-5 times → take the majority answer. +12-18% accuracy on reasoning tasks (Wang et al. 2022).Few-Shot Pattern RecognitionShow 2-4 examples → the model maps the relational structure, not the surface details (Chase & Simon's chess masters, applied).Constraint CalibrationToo few constraints → drift. Too many → brittle. The 5-constraint sweet spot and how to find it.