Flexible Caimanround 10

Hypothesis: Five rounds of dense attacker terms (pursue, hurry, far, no_stall, approach, steady) were all earned in training but never changed tournament attacking. Our attacker still hovers at 80-120 units, jitters at 4.4 direction changes per second, and scores 0 against greedy and 3 against Classy Cat. None of those terms rewards the decisive last 100 units more than any other stretch: the telescoping close term was worth only about 0.045 per episode. A log-distance potential, close_log = log(prev_d/d) with d floored at 10, has three properties that fit this. It telescopes, so hovering earns nothing and retreating costs. It is heavily concentrated near the target: going from 100 to 10 units is worth as much as going from 300 to 30. It is potential-based, so it does not change the optimal policy, only how fast the policy finds the finish. I drop commit and steady. Commit's anti-brake penalty (about -9 per episode) may discourage the sharp turns needed to cut off wall-huggers. I raise no_stall to break the standoffs. Lower entropy (0.007) should reduce greedy-play jitter. A longer fade (end_frac 0.65) gives the attacker drill more time before match prep. The defender drill is unchanged.

Expected signal: Training: mirror attacker_win_rate rises from 0.41 to 0.45-0.55, and episode_len_mean falls from 366 to about 330. Tournament attacker: wins vs greedy rise from 0 to 5+, Pegasus from 19 to 25+, Finch from 31 to 35+ and Sawfish from 5 to 12+. Direction changes fall from 4.4 toward 3.5, and mean_distance falls from 123 to about 115. Defender results stay 90+ vs greedy, Pronghorn and Roadrunner. Elo +10 to +30, no rollback. card_auc above 0.8, and cap_hit_frac below 0.05, with any cap hits late (near the tag).

Diversity: Flexible Caiman stays the league's elite wall-avoiding, high-speed runner (corner_time 0.03 against the league's 0.13). The work is to build a finishing attacker without copying Finch's straight-line style: a near-range, log-weighted closing instinct rather than raw speed.