MemberMarch 10, 2022 at 1:07 am
Thanks for Fail Faster and other cool stuff you do for the board game design world.
I’m experimenting with computer based playtesting. What are your thoughts on the potential for using AI to do massive amount of playtesting? My current thoughts is that it could be useful for getting quantitative data (how many rounds are 95 percent of the games, how often does the first player win, what happens if I change these numbers…) and also, if machine learning is used, find potential dominant strategies.
It would absolutely not replace human playtesting (since games are not about numbers, ultimately). But I think AI playtesting could speed up the process and get better games. Maybe.
What do you think?