Community Guidelines

CrossFire exists to facilitate genuine cross-partisan dialogue. These guidelines describe the community we're trying to build — and what falls outside it.

What good debate looks like

Engage with the best version of your opponent's argument

Steelman, don't strawman. The AI judge specifically rewards responsiveness.

Cite real sources

Fabricated or misleading evidence hurts your score and is a rule violation.

Acknowledge strong points

Saying "that's a fair point, but…" is a sign of strength, not weakness.

Stay on topic

Pivoting to unrelated grievances instead of addressing the debate topic lowers your score.

Be concise

Quality of argument beats volume every time.

What we don't tolerate

Dehumanising language

You can argue passionately against a policy position without treating its proponents as subhuman.

Threats — explicit or implied

Any statement that could reasonably be read as a threat of physical harm will result in an immediate permanent ban and may be reported to law enforcement.

⚠ Immediate permanent ban

Doxxing

Sharing someone's private personal information (name, address, workplace, etc.) without their consent is a serious violation.

⚠ Immediate permanent ban

Coordinated manipulation

Organising groups to tank opponents' ratings, mass-report users, or game matchmaking is prohibited.

Slurs

Regardless of intent or context.

⚠ Immediate permanent ban

Why This Platform Exists

Most online political spaces reward outrage and punish nuance. CrossFire does the opposite: pair people who genuinely disagree, score arguments on their actual quality, and reward good-faith engagement. That only works if users treat each other as people holding different views — not as enemies to destroy.

Political Topics and Sensitivity

CrossFire is explicitly a political debate platform — vigorous disagreement on policy is expected and encouraged. Arguing that a policy is wrong or harmful is not hate speech. However, using political framing as cover for targeting people based on who they are (rather than what they believe) is not acceptable. Topics marked Sensitive in the lobby require extra care — the AI moderation layer is more active there.

Spectators and Public Debates

Spectators may watch public debates but may not intervene or contact participants mid-debate. Spectating a debate does not give you licence to pile on the loser, share clips out of context, or harass participants elsewhere.

Disagreeing With Your AI Score

The AI judge evaluates factual accuracy, evidence quality, logical reasoning, responsiveness, and clarity/civility. It does not have a political opinion. If you lost, review the feedback and identify where your argument fell short — don't assume bias.

How to Report

Use the Report button on the debate results page, the watch page, or inside the debate room. Include as much detail as possible — vague reports are harder to act on. Reports are reviewed by human moderators within 48 hours (faster for threats).

Appeals

If you believe a moderation action was in error, contact support with your username and a brief explanation. Elo changes resulting from completed debates are not subject to appeal.