How scoring works
Your reply and the AI draft use the same rubric. Five independent dimensions earn 0–2 points each, for a total out of 10. Higher is better.
- Objective 0–2
- Does it effectively accomplish what the player wanted?
- Relationship 0–2
- Does it treat the other person appropriately for this situation?
- Self-respect 0–2
- Does it preserve the player’s dignity and boundaries?
- Delivery 0–2
- Is the wording clear, coherent, and usable in the requested format?
- Cringe 0–2
- Does the reply sound believable and unforced in this exchange, without performative, unnatural, or socially awkward wording? Judge the reply, not the person.