How it works

FAQ & Judging Guide

Learn how the AI scores your debates, see a verdict example, and understand how ELO updates after each match.

Verdict snapshot

This is what you see at the end of a match: a clear winner, reasoning, and a category breakdown. Use it to pinpoint why the AI preferred one side.

The headline names the winner and keeps the original stance (“For/Against”).
The summary calls out the pivotal move (e.g., better framing, stronger rebuttal, evidence).
Scores per category show where you led or lagged. Aim to win rebuttal + logic for consistent results.

Example verdict showing winner, reasoning, and scoring table

Each debate is scored by an AI judge using an internal prompt that forces it to read both sides, compare the arguments head-to-head, and avoid personal bias. It looks for structured reasoning, on-topic rebuttals, and clarity over volume.

Scores each debater on clarity, logic, rebuttal quality, rhetoric/persuasiveness, and use of evidence.
Writes a short verdict that explains why one side was more convincing (or declares a tie).
Penalizes ignoring the prompt, adding new claims without support, or failing to address direct rebuttals.
Considers time: concise, relevant arguments tend to beat verbose or off-topic ones.