Benign Pegasusround 6

Hypothesis: Round 5 gave the club's best result so far (+49, 1494 to 1543). Gamma 0.997 together with the flee/space/danger drill raised defender quality: snapshot_win_rate_defender rose to 0.375, training episode length rose to 203, defender corner_time_frac fell to 0.01 and defender mean_distance rose to 144. These changes have had only one round of training. The value function needs time to adapt to the longer horizon, and the evasion habits need another drill cycle before match prep consolidates them. Changing several things now would make it impossible to tell which change helped. So this round keeps hparams, opponents and the drill unchanged, and lets a second identical round show whether the gains continue, especially against the strong clubs. The open problems are defender vs Classy Cat (0/100) and Caiman (14), and attacker vs Caiman (29), greedy (37) and Magpie (38).

Expected signal: Training: snapshot_win_rate_defender stays at 0.3+, and episode_len_mean stays near or above 200. Round-robin: defender wins vs Caiman rise from 14 to 20+ and vs greedy from 26 to 30+. Defender vs Classy Cat stays low (0-10). Attacker vs Classy Cat holds at about 80+ and vs Roadrunner and Pronghorn at 75+. Attacker vs Caiman, greedy and Magpie holds at 30-40. Net Elo +10 to +30. If Elo falls or Cat and Caiman results don't move, next round targets attacker pressure against wall-pinned defenders and stalemates with a new drill.

Diversity: The club keeps its distinct style: a high-speed, mobile game in both roles. Its defender spends little time on walls and almost none in corners (wall 0.50, corner 0.01, against a league average of 0.66 and 0.18), and it trains on a long horizon (gamma 0.997). This sets it apart from the corner-parking clubs in the league.