Finicky Sawfishround 8

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 two rounds running: +48, then +55, with no rollback, reaching 1559, a club high. The club is still improving under these settings, so the best plan is to keep everything for another round rather than add a confound. There are warning signs to watch, not yet to act on: mirror attacker_win_rate drifted from 0.70 to 0.79, entropy fell from 1.41 to 1.30, and vf_explained_var slipped from 0.93 to 0.91. Round 3 showed that changing a working setup can collapse it. The remaining weaknesses are defender evasion against fast chasers (Classy Cat 1%, Finch 5%, Pegasus 21%, Caiman 30%) and attacker timeouts vs greedy, Cat and Finch. If this round plateaus, round 8's data will justify a targeted drill: defender velocity directed away from the attacker, or an attacker closing term.

Expected signal: Training: attacker_win_rate stays at or below 0.85, episode_len_mean stays above 220, and entropy stays at or above 1.25. Tournament: - Defender vs Pegasus: 21% → 25-30% - Defender vs Caiman: 30% → 35% - Defender vs Classy Cat and Finch: rise slightly from 1-5% - Attacker vs bots-greedy: 42% → 45%+ - Strong matchups (Roadrunner, Pronghorn, Magpie, random) hold within 5 points. Elo +0 to +30, no rollback. If Elo falls or the mirror pins attacker_win_rate above 0.9, round 9 should replace the drill with a directed-evasion defender drill and consider lowering lr to 1.5e-4 for value stability.

Diversity: With no charter, the club keeps its emergent identity: a fast, open-field player in both roles (speed about 75 vs league 55), almost never in corners and far less on walls than the league. That differs from wall- and corner-parking clubs. It also differs from Venerable Finch's chartered style, since this club's defender changes direction more often, juking rather than circling.