It took me more than 30 to do 6...
Well done titiogepi!
AI for all games is here!
Chess
playstrategy.org/kN0rwhvTAYlR
International Draughts
playstrategy.org/vojgZM0F9yB9
Lines of Action
playstrategy.org/FUaBDhQIsQuQ
Shogi
playstrategy.org/O0P4lCYDe4hO
Xiangqi
playstrategy.org/7DWzDXFzn6AK
Othello
playstrategy.org/Tucjo3l0TO6C
Oware
playstrategy.org/k7zkwivejt2A
Dear programmers! Your words are not true! You claim that the BOT (PS-Greedy-Two-Move) sees for 2 moves. Check out the antidraughts game:
playstrategy.org/PIWi44Em/p2
In this game, pay attention to the last 2 moves!!! Why did the BOT play 51 9-13? According to you, he should have seen the end of the party.
He should have played 51 20-25 and the game goes on.
I ask you to correct this error in the game of the bot.
<Comment deleted by user>
<Comment deleted by user>
Hello @marahonov_dima
Thanks for taking the time to post in the forum about this game.
At move 51, the PS-Greedy-Two-Move has only 6 possible moves. 2 result in making a King, something that the bot sees as bad in the game of Antidraughts. This leaves us with 4 moves, of the other men. For each of these moves, PS-Greedy-Two-Move sees an equal evaluation when looking 2 moves deep - whichever move it picks, white cannot capture any of its pieces. The fact that it sets up a capture for itself on move 52 isn't seen, the bot would need to look 3 moves deep to identify this as a bad move.
To be clear, the two moves PS-Greedy-Two-Move looks at are its own move and the opponents move. Nothing more.
Congratulations to @komacchin for being the second user to complete the PS-Greedy-Two-Move challenge! komacchin needed only 10 games to win at the 7 variants. Can anyone beat that?
Hello @statmatt
Thank you for the detailed explanation. I just misunderstood the name "2 moves".
However, in order to play antidraughts, the bot needs to see for 3 turns. It sounds sad, but I will say: the PS-Greedy-Two-Move will not play well in an antidraughts game. Please consider creating the PS-Greedy-Three -Move (-antidraughts).
And someone answer the question: is PS-Greedy-Two-Move a regular program or is it an artificial intelligence that learns from its mistakes?
<Comment deleted by user>
playstrategy.org/blog/Y5H-OxEAACsA6wze/ai-for-all-games-is-here
it's an AI
Not sure about how it works ( my above comments were wrong, I have removed them )
PS-Greedy-Two-Move uses a greedy algorithm with weightings based on piece value, checks, piece movement and other factors varying between games to evaluate each possible move. The algorithm looks two moves deep, i.e. it picks the best move after considering its own moves and what the opponent could respond with. It will take your unprotected pieces!
From this I presume that is it