# Moderation, Trust, Safety, and Fair Play Policy

Fan Passport should feel competitive, social, and celebratory without becoming hostile, spammy, or unfair. This policy governs user-generated content, leaderboards, rewards, social loops, and incident response.

## Surfaces covered

This policy applies to:

- Display names.
- Usernames and handles.
- Avatars.
- Squad names and descriptions.
- Match memory captions.
- Prediction comments if enabled.
- Sticker trade messages if enabled.
- Share card captions generated or edited by users.
- Leaderboard names and ranks.
- Referral links and invite messages.
- Support messages.
- Uploaded media if later supported.
- Public or semi-public recap pages.

## Policy principles

1. Celebrate fandom without attacking people.
2. Protect minors and vulnerable users.
3. Keep football debate from turning into hate, harassment, or threats.
4. Prevent spam, scams, and reward abuse.
5. Keep leaderboards fair and auditable.
6. Give users clear reasons for enforcement and a route to appeal.
7. Do not over-collect personal data for safety when less invasive controls work.

## Allowed content

Users may:

- Support teams, players, countries, clubs, and fan communities.
- Make respectful predictions and football arguments.
- Celebrate wins and mourn losses.
- Use light rivalry banter that does not target protected characteristics or encourage harm.
- Share their own passport progress, badges, stickers, and memories.
- Create squads around families, friends, workplaces, schools, pubs, or fan groups when names are appropriate.

## Restricted content

Restricted content may be limited, hidden, age-gated, or reviewed depending on context:

- Profanity that is not targeted harassment.
- Political references that are not hateful or threatening.
- Sensitive match incidents or injuries.
- User claims about officiating corruption without credible sourcing.
- Sponsor, betting, alcohol, crypto, or financial product discussion.
- External links in user captions or squad descriptions.
- Repeated promotional content.
- Claims of official affiliation by users, squads, or partners.

## Prohibited content

The following content should be removed and may lead to account enforcement:

- Hate speech or dehumanizing language based on protected characteristics, nationality, ethnicity, race, religion, gender, sexual orientation, disability, caste, or immigration status.
- Threats, incitement, praise of violence, or instructions for harm.
- Harassment, bullying, doxxing, stalking, or targeted abuse.
- Sexual content involving minors or any child safety violation.
- Non-consensual intimate content.
- Graphic violence or gore.
- Terrorist or extremist praise, recruitment, or propaganda.
- Scams, phishing, malware, credential theft, or impersonation.
- Illegal betting solicitation or attempts to evade age and jurisdiction controls.
- Reward fraud, botting, multi-account abuse, leaderboard manipulation, or sale of accounts and rewards.
- Unauthorized use of official marks in a way that implies partnership, verification, or endorsement.
- Private personal information such as addresses, phone numbers, emails, government IDs, precise location, or payment details.
- Attempts to bypass moderation filters or enforcement.

## Enforcement ladder

Enforcement should be proportionate to severity, history, and risk.

| Level | Action | Typical use |
| --- | --- | --- |
| Notice | User receives explanation and guidance without penalty. | Confusing rule, accidental external link, mild off-topic content. |
| Content removal | Specific content is hidden or deleted. | Inappropriate caption, squad name, or share text. |
| Warning | Account receives formal warning. | Repeated low-level violations or targeted profanity. |
| Feature restriction | User loses access to captions, squad creation, trading, or sharing for a period. | Spam, repeated abusive content, suspicious social activity. |
| Leaderboard hold | Points or rank are temporarily excluded pending review. | Suspicious scoring, referral spikes, prediction anomalies. |
| Reward hold | Reward is withheld pending eligibility review. | High-value prize, fraud signal, duplicate account pattern. |
| Temporary suspension | Account cannot participate for a defined period. | Harassment, evasion, repeated abuse, moderate fraud. |
| Permanent ban | Account is removed from participation. | Severe hate, threats, child safety, scams, confirmed major fraud. |
| Law enforcement escalation | Evidence preserved and escalated according to legal process. | Credible threats, child safety, serious illegal activity. |

## Moderation workflow

1. Content enters moderation through user reports, automated detection, staff review, sponsor escalation, support ticket, or anomaly detection.
2. System attaches context: user ID, content ID, surface, timestamp, locale, report reason, prior enforcement, and relevant event history.
3. Low-risk obvious violations can be actioned by trained moderators.
4. Ambiguous, severe, regulated, or high-profile cases escalate to Community Lead.
5. Legal/Compliance reviews cases involving threats, child safety, regulated categories, official rights, or law enforcement.
6. User receives an enforcement notice unless doing so would create safety, fraud, or legal risk.
7. Enforcement is logged for audit.
8. User may appeal eligible enforcement.
9. Appeals are reviewed by someone who did not make the original decision when staffing allows.
10. Confirmed errors are reversed and user trust is restored through clear messaging.

## User reports

Report reasons should be simple and consistent:

- Hate or harassment.
- Threat or violence.
- Spam or scam.
- Private information.
- Inappropriate squad or profile.
- Cheating or leaderboard abuse.
- Reward fraud.
- Impersonation.
- Other safety issue.

Report acknowledgement should be immediate. Outcome details can be limited to protect privacy, but users should know the report was reviewed.

## Service-level targets

| Case type | Initial review target | Resolution target |
| --- | ---: | ---: |
| Child safety or credible threat | 15 minutes | Immediate escalation and ongoing handling. |
| Hate, harassment, doxxing, scam | 2 hours during live coverage | 24 hours. |
| Leaderboard top-rank anomaly | 2 hours during tournament | Before reward payout or public winner announcement. |
| High-value reward fraud signal | 4 hours | Before fulfilment. |
| General profile or squad name report | 24 hours | 72 hours. |
| Appeal | 72 hours | 7 days. |

## Leaderboard integrity

Competitive leaderboards are only valuable if users trust them.

### Scoring integrity rules

- Score-changing events must be recorded server-side.
- Prediction submissions require immutable timestamps and lock validation.
- Challenge completion should be idempotent so repeated requests cannot duplicate points.
- Daily challenge XP caps should be enforced per account and per campaign.
- Referral points should require referred users to complete a meaningful passport action.
- Suspicious users can remain active while their leaderboard points are held.
- Public winner announcements should occur only after fraud review.

### Fraud signals

- Impossible completion velocity.
- Multiple accounts on same device claiming the same reward.
- Repeated referrals among a closed account cluster.
- Abnormally high share or referral conversion from low-quality traffic.
- Prediction submissions after lock time.
- Check-ins outside configured match windows.
- Repeated quiz submissions faster than human reading time.
- Many accounts sharing the same reward redemption pattern.
- Sudden leaderboard jumps inconsistent with event history.
- Use of automation, scripts, emulators, or API abuse.

### Leaderboard response actions

- Soft hide suspicious entries while review is pending.
- Freeze leaderboard payout for the affected segment.
- Recalculate points after invalid events are removed.
- Publish a general integrity note if public ranks changed materially.
- Preserve audit logs for all score-changing events.

## Reward safety

High-value rewards require stronger controls:

- Age and territory gating where required.
- Terms accepted before entry or redemption.
- Manual review before fulfilment.
- Clear expiration and inventory limits.
- No transfer or resale unless explicitly allowed.
- Sponsor fulfilment contacts available during campaign.
- Support escalation path for missing, invalid, or expired codes.

If a reward campaign experiences fraud, pause new redemptions before revoking issued rewards. Determine whether honest users were affected and remedy them first.

## Match and national rivalry risk

International football can create intense national, ethnic, religious, and political tension. During high-risk fixtures:

- Increase moderation staffing.
- Pre-review squad names and public captions appearing on leaderboards.
- Monitor reports by fixture and country segment.
- Avoid inflammatory push copy.
- Use neutral editorial tone for controversial incidents.
- Disable user-edited public captions temporarily if abuse spikes.
- Coordinate with support for confused or upset fans after eliminations.

## Minors and youth-safe defaults

- Avoid paid random reward mechanics that could resemble gambling.
- Age-gate regulated sponsor categories.
- Do not allow public display of age.
- Avoid precise location sharing.
- Use strict filters for sexual content and adult solicitation.
- Provide parent or guardian support path where appropriate.
- Keep reward terms and odds clear.

## Appeals

Eligible enforcement notices should include:

- What rule was violated.
- What action was taken.
- How long the action lasts.
- Whether content can be edited and resubmitted.
- How to appeal.

Appeals should consider context, translation issues, local football slang, user history, and whether the user intended harm. Severe safety and confirmed fraud cases may have limited appeal detail to protect investigations.

## Incident response for safety events

A safety incident exists when abuse, fraud, or harmful content affects many users, high-profile surfaces, public trust, or legal risk.

1. Community Lead declares incident severity.
2. Live Ops Lead coordinates cross-functional response.
3. Engineering disables affected surface if needed.
4. Moderation clears highest-risk content first.
5. Analytics identifies affected users, surfaces, and spread.
6. Support prepares user-facing response.
7. Legal reviews external communication if safety, regulated content, or rights are involved.
8. Incident closeout documents cause, impact, response, and prevention.

## Public trust language

Use plain language when communicating integrity actions:

- "We removed accounts that broke the challenge rules."
- "We are reviewing unusual leaderboard activity before awarding prizes."
- "A fixture data issue affected prediction scoring for a small group of users. We corrected the score and restored eligible rewards."
- "We removed content that violated our community policy. Football rivalry is welcome; hate and harassment are not."

Avoid naming accused users or revealing detection methods.