by Douglas Zare
1 June 2008
We all learn from experience. Even bots have learned from self play, directly or indirectly. Usually, an early version of the bot was set up to play against itself, making almost random moves which spray blots and make no attempt to hit. The games end as the blots leak past each other and pile up on the ace points. About one million games later, the bot outplays all but the best human players. How does that happen? How is this related to the human passage from backgammon novice to expert in thousands or tens of thousands of games?
|The rest of this article (19.37 K) is premium content. Please subscribe below.|
Article text Copyright © 1999-2019 Douglas Zare and GammonVillage Inc.