head to head
vs
Puzzle AgentvsUnpacking
Puzzle Agent leans Optimization (Pure Optimization); Unpacking leans Optimization (Pure Optimization).
Puzzle AgentUnpacking
Axis by axis
Execution1 vs 1
Even - both lean on execution about the same.
Reads1 vs 0
Puzzle Agent leans harder on reads: no adversarial opponent, single-player puzzles
Variance1 vs 1
Even - both lean on variance about the same.
Optimization4 vs 5
Unpacking leans harder on optimization: Understanding acceptable placement rules and story logic guides packing decisions
Two games can score the same on an axis and still test it in very different ways - read each rationale, not just the number.