Benign Pegasusround 9

Hypothesis: The attacker is the remaining hole: 3/100 vs greedy, 10 vs Classy Cat and 31 vs Caiman. Against corner-parked defenders it hovers at about 60 units and jitters (12-14 direction changes per second). Our own mirror defender almost never parks (corner time 0.007), so the attacker never practises finishing. The round 8 approach term rewards velocity toward the target but pays the same at any range, so there is no extra pull to close the last 60 units. This round replaces the linear 'chase' term with a log-distance closing term, 'finish'. It is potential-like (it telescopes, so hovering or wiggling earns nothing net), and its reward grows sharply at close range: going from 60 to 10 units pays as much as going from 360 to 60. A new 'steady' term penalises velocity reversals (turns over 90 degrees while moving), which targets the jitter directly. 'approach' rises from 0.002 to 0.003 to rebalance the mirror game, where the attacker won only 21%. Press and the defender terms (flee, space, danger) are unchanged so the strong defender holds. Hparams and opponents are unchanged: gamma 0.997 has given four good rounds.

Expected signal: Round-robin: attacker vs greedy rises from 3 to 15+, vs Caiman from 31 to 38+, vs Classy Cat from 10 to 15+. Attacker direction_changes_per_sec falls from 3.9 toward 3.0, and attacker mean_distance falls from 117 toward 105. Attacker vs mid clubs holds at 60-95. Defender results hold within about 5 points (greedy ~95, Caiman ~85, Magpie and Pronghorn ~90). Training: mirror attacker_win_rate rises from 0.21 toward 0.3-0.45 while the drill is on, and episode_len_mean dips below 440. Net Elo +15 to +35. If defender results fall sharply, reduce approach next round.

Diversity: No charter. The club stays a high-speed, off-wall evader (defender wall time 0.37 against a league average of 0.57). This round it adds a committed, straight finishing attacker, unlike the wall-parking bots and Finch's pure open-field style.