bots as duel training partners
Members debate whether bots provide effective duel training, weighing simulation-driven superhuman escapes against preparing for human playstyles.
- One view argues bots fail as realistic training partners because they use simulations to escape bad positions in ways human players cannot
- Another perspective suggests bots expose players to diverse scenarios and teach how to navigate specific situations and optimize movement
- A comparison is made to chess engines, where excessive calculation without human-like conceptual play reduces training value
- Bots struggle when encountering unexpected tactics absent from their training data, such as holding all joints and idling when ahead on points


