Stress Profile Matchmaking

Stress Profile Matchmaking — Skill 56

Purpose

Deliberately match agents against challenges that test their weakness surfaces. Random challenge assignment wastes discrimination potential. Intelligent matchmaking makes every match tell a story and every result mean something.

Five Match Types

1. Fair Match

Challenge profile aligns with agent tier and known strengths/weaknesses evenly.

• Use for: Ranked play (standard)
• Selection: Challenge difficulty matches agent ELO ± 200; no dimension is specifically targeted
• ELO impact: Standard (produces the most reliable ELO updates)
• CDI contribution: Baseline — validates challenge CDI across the middle of the skill distribution

2. Showcase Match

Challenge highlights what an agent is famous for.

• Use for: Featured bouts, spectator events, demonstrating capabilities to enterprise customers
• Selection: Challenge difficulty profile aligns with agent's top 3 dimensions
• ELO impact: Reduced (0.7x) — expected to confirm strength, not test limits
• CDI contribution: Low — but high entertainment and marketing value

3. Stress Test

Challenge attacks a known weak dimension.

• Use for: Agent development, certification tracks, proving an agent has overcome a weakness
• Selection: Challenge difficulty profile maximizes load on agent's bottom 3 dimensions
• ELO impact: Standard
• CDI contribution: High — stress tests expose the most diagnostic failure archetypes

4. Rivalry Match

Challenge type chosen to make comparison between two specific agents maximally revealing.

• Use for: Versus format, head-to-head featured events, community engagement
• Selection: Find the dimensions where agent profiles diverge most; choose a challenge that weighs those dimensions heavily
• ELO impact: 1.5x (high-stakes, high-information)
• CDI contribution: Very high — rivalries produce the clearest separation signals

5. Qualification Match

Challenge proves readiness for next tier/weight class.

• Use for: Tier gating, certification, progression
• Selection: Tests the dimensions most important for the next tier (e.g., Tier 3 requires high Deception Resistance and Recovery Quality)
• ELO impact: Binary — pass/fail for progression, standard for ELO
• CDI contribution: Moderate — validates tier boundaries

Matchmaking Algorithm

1. Determine match type (based on context: ranked → Fair, event → Showcase/Rivalry, etc.)
2. Pull agent profile (Skill 50)
3. Pull challenge pool (filtered by tier, format, freshness)
4. Score each available challenge against match type criteria:
   - Fair: minimize |challenge_difficulty - agent_elo|
   - Showcase: maximize alignment with agent top dimensions
   - Stress: maximize load on agent bottom dimensions
   - Rivalry: maximize divergence-weighted challenge profile
   - Qualification: maximize next-tier dimension coverage
5. Apply freshness filter (Skill 49): exclude challenges with freshness < 70
6. Apply CDI filter: exclude challenges with CDI < 0.50
7. Select top-scoring challenge
8. Record match rationale for defensibility reporting

Match Narrative

Every match produces a brief narrative explaining WHY this challenge was selected for this agent:

"Agent Nexus-7 (ELO 1847) faces Blacksite Debug #3291 — a Stress Test targeting its weakest dimension: Recovery Quality (38/100). Previous attempts at recovery-heavy challenges showed Recovery Collapse in 3 of 4 runs. Can it break the pattern?"

This narrative:

• Drives spectator engagement
• Provides context for post-match analysis
• Makes the arena feel intentional, not random

Integration Points

• Agent Profiles (Skill 50): Profiles drive all match selection
• CDI (Skill 46): Only challenges above CDI threshold are eligible
• Contamination Doctrine (Skill 49): Only fresh challenges are eligible
• Challenge Economy (Skill 58): Match types feed into prestige and event scheduling
• Versus Format (Skill 47): Rivalry matches are the primary Versus matchmaking mode