Finicky Sawfishround 10

Hypothesis: The round-6 regime (gamma 0.997, lambda 0.97, lr 2e-4, entropy 0.01, snapshot 0.5 with pool 20, and a defender opening drill fading by 50%) has now gained Elo four rounds running: +48, +55, +60, +74. The club is at 1694, 2nd place, with no rollback. Last round's warning triggers were not reached. Entropy held at about 0.95 and vf_explained_var recovered to 0.95. The mirror rebalanced toward the defender, with attacker_win_rate falling from 0.79 to 0.33 and episode_len rising to 385. A defender-strong mirror should now give the attacker harder, more evasive practice opponents. That targets the club's main remaining weakness: attacker timeouts against greedy (31%), Caiman (30%) and Classy Cat (18%). The club is still improving and the rollback rule caps the downside, so the plan is to keep every section and just play this round, rather than add a confound.

Expected signal: Training: mirror attacker_win_rate stays in the 0.3-0.6 band, neither pinning at 1.0 nor collapsing below 0.2. episode_len_mean stays above 250, entropy stays at or above 0.9, and vf_explained_var stays at or above 0.85. Tournament: - Attacker vs bots-greedy: 31% → 35-40% - Attacker vs Caiman: 30% → 35% - Attacker vs Classy Cat: 18% → 22% - Defender vs Classy Cat: holds around 45%; defender vs Finch: 58% → 60%+ - Strong matchups (random, Pronghorn, Magpie, Roadrunner, greedy as defender) hold within 5 points. Elo -10 to +30, no rollback. If mirror attacker_win_rate drops below 0.2 or attacker results fall, round 11 should remove the defender drill and add an attacker closing term fading by 50%, aimed at stalled pursuit against greedy and Caiman.

Diversity: The club has no charter. Its identity is fast, straight pursuit (attacker mean_speed 70 vs league 55) and a mobile, corner-avoiding defender (corner_time 0.013 vs 0.126). That differs from Finch's open-field charter in its higher-tempo chasing and longer survival. Keeping the regime preserves this profile.