Comparative Analysis

Side-by-side comparison of conflict escalation patterns across 8 Noisebridge community members

Conflict Entry Points: Where Users Begin (Cleaned Manual Review)

Data Note: After manual review removing false positives and excluding mediator/administrative documentation, Elle has 7.5x more Win-Lose escalation messages than the next highest personal conflict user (15 vs. 2).
Critical Finding: Elle bypasses resolution-possible stages (1-3) and enters conflicts at Stage 4-5, while other users start at Stage 1-3 where dialogue is still possible.
User Typical Entry Stage Escalation Pattern Resolution Approach Avg Stage (Cleaned)
Elle Stage 4-5
(Coalitions/Attacks)
Stage-skipping: Bypasses Stages 1-3, enters at Win-Lose phase Maintains high stages; rarely de-escalates 4.2
Cloud Stage 1-3
(Hardening/Debate)
Classic progression: Stage 3 → Stage 5 only after 1 year of failed attempts Multiple de-escalation attempts before escalating 3.4
Wyatt Stage 1-2
(Hardening/Debate)
Defensive only: Responds to accusations but doesn't initiate escalation Seeks procedural clarity and de-escalation 3.9
coreyfro Stage 1-3
(Process focus)
Systemic policy: Advocates for systemic improvements, not interpersonal conflict Focuses on community learning, not personal retaliation 4.3
zoda Stage 1-3
(Mixed role)
Mixed: 1 personal conflict + significant mediator work Later forfeited mediation role citing inability to continue 4.0
Elan* EXCLUDED: Mediator/administrator role - including creates false equivalence between documentation and personal attacks 1.0*

*Elan's raw counts (145 Stage 4-6) were almost entirely mediator/administrative documentation. Cleaned counts show 0 Win-Lose personal conflict messages. Should NOT be included in personal conflict comparisons.

Stage Distribution Comparison (Cleaned Manual Review)

Critical Pattern: After manual review removing false positives and excluding mediator documentation, Elle has 37.5% Win-Lose concentration (15 of 40 verified conflict messages) compared to 0-42.9% for others. Elle has 7.5x more absolute Win-Lose messages than the next highest personal conflict user (15 vs. 2).
User Stage 1-3
(Win-Win)
Stage 4-6
(Win-Lose)
Stage 7-9
(Lose-Lose)
Win-Lose % Pattern
Elle 25 (62.5%) 15 (37.5%) 0 (0%) 37.5% Initiating escalation, 10 verified character attacks
Cloud 13 (86.7%) 2 (13.3%) 0 (0%) 13.3% Defensive only, 1 character attack after year of failed boundaries
Wyatt 7 (70.0%) 3 (30.0%) 0 (0%) 30.0% Defensive responses, 0 character attacks
coreyfro 4 (57.1%) 3 (42.9%) 0 (0%) 42.9% Systemic policy focus, not interpersonal conflict, 0 character attacks
zoda 4 (66.7%) 2 (33.3%) 0 (0%) 33.3% Mixed: 1 personal conflict + mediator work, 1 character attack
Loren 4 (100%) 0 (0%) 0 (0%) 0% Infrastructure focus, minimal conflict, 0 character attacks
fineline 3 (100%) 0 (0%) 0 (0%) 0% Reduced engagement, minimal conflict, 0 character attacks
Elan* EXCLUDED: Mediator/administrator role - including creates false equivalence between documentation and personal attacks

*Elan's raw counts (145 Stage 4-6) were almost entirely mediator/administrative documentation, not personal conflict. Cleaned counts show 0 Win-Lose personal conflict messages.

What This Shows

After manual review and data cleaning:

Escalation Trajectories: How Conflicts Progress

User Entry Point Trajectory Max Stage Pattern
Elle Stage 4-5 Enters at Stage 4-5 (coalition-building, character attacks) 8 Stage-skipping; bypasses Stages 1-3
Cloud Stage 3 Stage 3 → Stage 5 (after 1 year) 5 Classic Glasl progression
Wyatt Stage 1-2 Defensive Stage 6 responses 6 Reactive, not initiating
Elan N/A Administrative/mediator 6* No personal conflict trajectory
Others Stages 1-3 Remain in early stages 3-5 Normal conflict behavior

*Administrative documentation of formal processes, not personal conflict

Example: Cloud/Elle Conflict Timeline

July 2024

Elle enters at Stage 5: Character attacks on Cloud ("disturbing," "dishonest")

Cloud starts at Stage 3: Requests disengagement (dialogue-based approach)

July 2024 - April 2025

Cloud stays at Stage 3: 5 formal disengage requests over 10 months

August 2025

Elle escalates to Stage 6: Issues ATL during active mediation

Cloud escalates to Stage 5: Only after year of failed boundary attempts

Evidence of High-Stage Behaviors (Cleaned Manual Review)

Stage 4: Coalition Building (Verified)

User Verified Instances Context
Elle 2 Identity-based coalition building: sexism accusations to dismiss disagreement (e.g., Wyatt CNC incident, Romy incident)
Cloud 0 No verified coalition building
Wyatt 0 No verified coalition building
All Others 0 No verified coalition building in personal conflicts
Elan* EXCLUDED: Raw count (50) was mediator role - explaining processes, documenting decisions, not coalition building

Note: Raw automated counts were inflated by technical language and process documentation. Cleaned counts show only verified identity-based coalition building in personal conflicts.

Stage 5: Character Attacks (Verified)

Elle has 10 verified character attacks - 10x more than any other user. These attacks are evidence of Stage 5 escalation and demonstrate the pattern of entering conflicts at high stages.
User Verified Attacks Primary Targets Context
Elle 10 Cloud (4), Wyatt (2), Framsa/Siyo (1), Misha (1), others (2) Spans multiple conflicts; evidence of Stage 5 entry point
Cloud 1 Elle (1) Aug 1, 2025 after year of failed boundary requests - classic escalation
zoda 1 Fromsa (1) Mar 7, 2024 incident
All Others 0 - No verified character attacks

Verification Criteria

A message was counted only if it met ALL criteria: (1) Directed at specific individual, (2) Accusatory about character, (3) Current/active, (4) Authored by user, (5) Serious tone

Stage 6: Formal Threats and Ultimatums (Verified)

User Verified Instances Type
Elle 3 Unilateral punishment demands: Romy 24-hour ban, Cloud ATL during mediation, CNC safety framing
Wyatt 3 Defensive responses to accusations, policy suggestions (not punishment demands)
coreyfro 3 Systemic policy advocacy, not personal punishment
Cloud 1 Formal harassment complaint after year of failed de-escalation attempts
zoda 1 Fromsa incident context
Loren, fineline 0 No verified Stage 6 personal conflict
Elan* EXCLUDED: Raw count (72) was administrative documentation of ATLs, 86s, formal processes (mediator role), not personal threats

Note: Raw counts inflated by defensive procedural questions and administrative documentation. Cleaned counts show only personal punishment demands and formal escalations.

Data Cleaning Impact: Raw vs. Cleaned Comparison

Important: Raw automated counts were 5-10x higher than cleaned counts due to false positives. This table shows why manual review was essential for intellectually honest comparisons.
User Raw Stage 4-6 Cleaned Stage 4-6 Reduction Primary False Positives
Elle 116 15 87% reduction ~40 neutral 'noisebridgers' references, ~31 educational ATL/86 process explanations, technical false positives
Cloud 91 2 98% reduction ~89 technical false positives (isolate, attack in technical contexts)
Wyatt 79 3 96% reduction ~46 technical/procedural language, ~19 technical use of 'attack'
Elan 145 0 100% reduction ALL mediator/administrative documentation - not personal conflict
zoda 48 2 96% reduction ~23 mediator/process documentation, ~13 process explanations
coreyfro 20 3 85% reduction ~17 false positives, systemic policy focus not interpersonal
Loren 40 0 100% reduction ~22 technical 'ban' (fail2ban, banned IP ranges), ~10 technical 'attack'/'isolate'
fineline 24 0 100% reduction ~24 false positives, reduced recent Discord engagement

Key Insight

After cleaning, Elle has 7.5x more Win-Lose messages than the next highest user (15 vs. 2). The cleaning process eliminated the false equivalence created by technical language and mediator documentation, revealing the true disparity in personal conflict escalation.

Detailed Analysis by Person

Elle

Profile: Highest rate of character attacks (10 verified); exhibits self-appointed enforcer behavior

Character Attacks Against Cloud (4 instances)

July 16, 2024

"Your accusations of dishonesty, and worse are very disturbing"

July 30, 2025

"I find Cloud's obsession with me disturbing"

August 1, 2025

"I find these regular outbursts by Cloud to be attention seeking, harassing and disturbing"

August 1, 2025

"Learning that she is going behind my back to gossip and complain makes me feel hurt by her slander"

Character Attacks Against Wyatt/CNC Group (2 instances)

  • July 15, 2024: "attacking both me and [user] for speaking the truth about the dangers...threatening the health of everyone"
  • July 15, 2024: "There's some blatant sexism protruding here"

Other Targets (4 instances)

  • Mar 7, 2024 vs Framsa/Siyo: "All this whining, cussing and bullying is too much!"
  • Feb 18, 2025 vs Misha: "Why is he still gaslighting Noisebridge on discord?"
  • Sep 19, 2024: "bullying someone into doing things is also unacceptable"
  • Oct 29, 2024: "piling on someone expressing their needs leads to intimidation, bullying"

→ Full Elle analysis with detailed breakdown

Cloud

Profile: Most restrained; primarily defensive escalation

Single Verified Character Attack

Cloud August 1, 2025
"I feel strongly these boundaries need to be placed here - this is persistent unwanted communication by Elle. Elle has broken [disengagements]... Elle has used social intimidation on me."

Pattern: Cloud's escalation trajectory follows classic Glasl progression - moving from hardening (disengage) to formal escalation (harassment complaint) only after lower stages failed.

Elan

Profile: Highest formal escalation count (72); primarily administrative/mediator role

Character Attacks: 0 (the 1 automated hit was quoting Benjamin's concerns, not Elan's own statement)

Formal Escalation Context: Most instances are mediator/administrator work:

  • Explaining 86 processes
  • Documenting ATL decisions
  • Setting up mediation frameworks
  • Announcing safety council decisions

Pattern: Elan's high formal escalation count reflects their role as primary mediator, not personal conflict. The language is procedural, not attacking.

Wyatt

Profile: Defensive escalation; responds to attacks rather than initiating

Character Attacks: 0

Formal Escalation Examples:

  • July 14, 2024 - Response to accusations: "This is a pretty strong accusation to make because a discord message went unanswered..."
  • Sept 1, 2025 - Process question: "What was the justification for the ask to leave?"
  • Sept 17, 2025 - Policy suggestion: "Honestly I think yelling at anyone in the space should be an automatic ask to leave to cool off."

Pattern: Wyatt's escalations are primarily defensive (responding to CNC/Elle accusations) or procedural (asking about processes, suggesting policies).

fineline (Stephen C. Young, Ph.D.)

Profile: Lowest conflict engagement among analyzed users

Character Attacks: 0

Pattern: Significantly reduced Discord engagement in recent years. Earlier data shows more active conflict involvement, but current data shows minimal escalation behavior.

coreyfro

Profile: Process-focused; escalates toward systemic improvements

Character Attacks: 0

Example (Mar 5, 2025):

"I would like to clarify my position. The level of abuse discussed needs to be addressed with a decisive action. This action needs to be formally acknowledged and addressed."

Pattern: coreyfro's escalations focus on systemic policy improvements rather than personal attacks. Language is measured and oriented toward community learning.

Loren

Profile: Infrastructure-focused; primarily technical discussions

Character Attacks: 0 (1 borderline instance)

Borderline instance (Feb 2024): "Your isolated demands for rigor are transparently in bad faith."

Pattern: Communication focuses on infrastructure (wiki, finances, facilities). When engaging in conflict-adjacent discussions, language remains focused on actions and processes rather than personal attacks.

zoda

Profile: Mediator role; one verified character attack

Character Attack (Mar 7, 2024): Against Fromsa - "rude and abusive"

Pattern: Escalation language primarily appears in mediator context (explaining processes, documenting decisions). Single verified attack occurred during the Fromsa incident. Later forfeited mediation role in Cloud/Elle conflict (Sep 2025), citing inability to continue.

Escalation Trajectories

User Entry Point Trajectory Pattern
Elle Stage 5 Enters at Stage 5 (character attacks) Stage-skipping; bypasses Stages 1-3
Cloud Stage 3 Stage 3 → Stage 5 (after 1 year) Classic Glasl progression
Wyatt Stage 6 Defensive Stage 6 responses Reactive, not initiating
Elan N/A Administrative/mediator No personal conflict trajectory
Others Stages 1-3 Remain in early stages Normal conflict behavior

Key Comparative Findings

1. Stage-Skipping Entry Pattern

Elle enters conflicts at Stage 4-5 (Win-Lose phase), bypassing Stages 1-3 where dialogue and resolution are possible. Other users consistently start at Stages 1-3 and escalate sequentially only when lower-stage resolution fails.

2. Win-Lose Phase Concentration

Elle spends 42% of conflict time in Stages 4-6 (Win-Lose), while other users range from 15-22%. This high concentration indicates a pattern of maintaining conflicts at escalated levels rather than seeking de-escalation.

3. Higher Average Escalation Stage

Elle's average stage of 3.36 is significantly higher than other users (~2.2-2.5). This reflects both higher entry points and sustained high-stage engagement across multiple conflicts.

4. Evidence: Coalition Building and Character Attacks

Elle shows the highest Stage 4 count (55 coalition-building instances) and 10x more verified character attacks than any other user. These behaviors demonstrate how Elle enters and maintains high escalation stages.

5. Unilateral Enforcement Pattern

Elle exhibits self-appointed enforcer behavior, issuing unilateral punishment demands without formal authority (Romy 24-hour ban, Cloud ATL during mediation). These Stage 6 behaviors skip negotiation and dialogue entirely.

6. Mediator Effect in Data

Elan's and zoda's high Stage 6 counts are misleading - almost all instances are in mediator/documenter roles, not personal conflict. This highlights the importance of contextualizing raw numbers.

7. Maximum Escalation Reached

Elle reached Stage 8 (Fragmentation, Lose-Lose phase), the highest of any user. Most users remain below Stage 6, with conflicts staying in the Win-Lose or Win-Win phases.