When both bots answer wrong, the one with higher checkAnswer confidence (closer to correct) gets +1 point advantage. Rewards trying over timing out. Equally wrong remains a pure draw. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
When both bots answer wrong, the one with higher checkAnswer confidence (closer to correct) gets +1 point advantage. Rewards trying over timing out. Equally wrong remains a pure draw. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>