G3-F TrueSkill μ=30.1 ± 3.3, 72.9% wins, median score 5,250 on combined-comp1 slice (n=98 canonical games)

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• G3.1-FL TrueSkill μ=28.0 ± 2.9, 44.9% win ratefinding0.896
  Second-best LLM, competitive with TrackerAgent.
• G2.5-FL TrueSkill μ=20.5±2.9finding0.822
  Gemini 2.5 Flash Lite, lowest TrueSkill rating
• TrackerAgent TrueSkill μ=28.7±3.6 in 98 canonical gamesfinding0.816
  second-highest TrueSkill rating
• G3-F win rate=72.9% in 98 canonical gamesfinding0.814
  Gemini 3 Flash won nearly 3/4 of its games
• DS-v3.2 TrueSkill μ=23.9±2.7finding0.812
  DeepSeek v3.2 TrueSkill rating
• Gemini 3 Flash TrueSkill µ = 30.1 ± 3.3 in 98 canonical gamesfinding0.808
  The Bayesian skill posterior mean for G3-F is 30.1 with 3σ confidence interval ±3.3.
• TrackerAgent TrueSkill μ=28.7 ± 3.6, 53.6% win ratefinding0.807
  Best code agent outperforming six of seven LLMs.
• Sonnet 4.5 TrueSkill μ=26.4 ± 4.9 (n=14, 35.7% win rate)finding0.802
  Mid-field performance with larger uncertainty due to small sample.