Back to Blog
Target stratego game5/16/2023 ![]() And it’s a zero-sum game, so any gain by one player represents a loss of the same magnitude for their opponent. It’s a game of bluff and tactics, of information gathering and subtle manoeuvring. ![]() ![]() ![]() Stratego is a turn-based, capture-the-flag game. Our paper shows how DeepNash can be applied in situations of uncertainty and successfully balance outcomes to help solve complex problems. In pursuit of our mission of solving intelligence to advance science and benefit humanity, we need to build advanced AI systems that can operate in complex, real-world situations with limited information of other agents and people. The value of mastering Stratego goes beyond gaming. For this reason, DeepNash goes far beyond game tree search altogether. It also means that a very successful AI technique called “game tree search”, previously used to master many games of perfect information, is not sufficiently scalable for Stratego. This complexity has meant that other AI-based Stratego systems have struggled to get beyond amateur level. Unlike chess and Go, Stratego is a game of imperfect information: players cannot directly observe the identities of their opponent's pieces. So hard, in fact, that DeepNash has reached an all-time top-three ranking among human experts on the world’s biggest online Stratego platform, Gravon.īoard games have historically been a measure of progress in the field of AI, allowing us to study how humans and machines develop and execute strategies in a controlled environment. Its play style converges to a Nash equilibrium, which means its play is very hard for an opponent to exploit. Published in Science, we present DeepNash, an AI agent that learned the game from scratch to a human expert level by playing against itself.ĭeepNash uses a novel approach, based on game theory and model-free deep reinforcement learning. Stratego, the classic board game that’s more complex than chess and Go, and craftier than poker, has now been mastered. Game-playing artificial intelligence (AI) systems have advanced to a new frontier. Pieces are not allowed to move there.DeepNash learns to play Stratego from scratch by combining game theory and model-free deep RL Within the board there are two obstacles of 2 x 2 squares each. The Stratego board consists of 10 x 10 squares. When you capture the flag of your opponent you win the game. A bomb defeats every piece that tries to attack it, except the miner. Bombs lose when they are defused by a miner. However, when the marshal attacks the spy, the spy loses. When the spy attacks the marshal, the spy defeats the higher ranked marshal. Exception to the rule of the higher rank winning is the spy. Pieces may not move onto a square already occupied by another piece without attacking. If the engaging pieces are of equal rank, they are both removed. The weaker piece is removed from the board, and the stronger piece is moved into the place formerly occupied by the weaker piece. ![]() If a piece is moved onto a square occupied by an opposing piece, their identities are revealed. Only the scout can move over multiple empty squares per turn. Pieces move 1 square per turn, horizontally or vertically. In addition to those ranked pieces you can use bombs to protect your flag. Pieces have a rank and represent individual officers and soldiers in an army. To capture the flag you use your army of 40 pieces. Stratego is a game in which you need to capture the flag of your opponent while defending your own flag. MAKE SURE TO HAVE THE HOST CLICK THE "DEAL DECK" BUTTON UNTIL THE DECK IS GONE!. ![]()
0 Comments
Read More
Leave a Reply. |