== rounds/memladder-crossover-rotated-2026-09-17/roundA-14sep-binary-27f05a889.jsonl  budget=128  metric=cpu  288 ok leg(s), 0 refused
   arm_order=rotating-by-rep
   load_before min/median/max 3.5/5.2/11.6   load_after 3.6/5.2/11.6

-- rotation, as banked --
   arm    pos0  pos1  pos2
   auto     32    32    32
   fold     32    32    32
   force    32    32    32
   BALANCED: every arm in every position equally often (counts [32])

-- paired per-rep difference, fold minus force (positive = force is cheaper) --
        m  reps   fold med  force med  med diff     med %     sign  p
      160    24       9.44      10.43    -0.980   -10.38%    0/24   0.0000
      192    24      10.62      10.52    +0.100    +0.94%   24/24   0.0000
      224    24      11.82      10.62    +1.195   +10.11%   24/24   0.0000
      256    24      13.01      10.73    +2.270   +17.45%   24/24   0.0000

-- crossover --
   fold wins (significant):  [160]
   ties (sign test n.s.):    none
   force wins (significant): [192, 224, 256]
   => crossover is between m=160 and m=192

-- position effect (arm level and rep level divided out; what running Nth is worth) --
        m          pos0           pos1           pos2
      160        -0.02%         +0.02%         -0.01%   pos0-pos1=-0.04%
      192        -0.00%         +0.02%         -0.05%   pos0-pos1=-0.02%
      224        -0.04%         +0.00%         +0.03%   pos0-pos1=-0.04%
      256        +0.02%         -0.05%         +0.05%   pos0-pos1=+0.06%
   The levels are centred, so the READABLE quantity is the GAP: pos0-pos1 is what
   the first slot cost the arm that held it, and is the term a fixed order banks.
   (a FIXED-order round carries pos0-minus-pos1 as a systematic term it cannot see:
    on this driver until 17 Sep 2026 that term sat entirely on `fold`, which ran first)