Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Kilobyte’s Gambit

vole.wtf

Post reply on HN
Print view View original

131–140 of 158 posts

Re: The Kilobyte’s Gambit

#132

Post by drdaeman »

Thanks, I hate it. I'm losing every single game (I think I tried like about 30-ish times). If that's how 12-years old are playing them I should probably apply for disability benefits. My naive approach by trying to define safe squares to move pieces to and naive capture goals by exchanging figures of lesser value is not working against this algorithm, I guess I'll probably have better chance if I'd just run the same logic on paper.

Re: The Kilobyte’s Gambit

#133

Post by kuroguro »

schoen wrote: post #99

Sharing other people's amazement that such a tiny amount of code plays so well (I'm an inexperienced, infrequent player and it beat me once and drew once), I was pondering that there's no absolute reason to think that small programs inherently must play games poorly. For many abstract strategy games, you could write an explicit game tree search in a relatively small amount of code and it would, in principle, play per…

Somewhat related: I was blown away when I first heard the argument that intelligence is basically compression.

http://prize.hutter1.net/hfaq.htm#compai

Haven't done any more digging into this but it did give me something to ponder about.

Re: The Kilobyte’s Gambit

#134

Post by Closi »

drdaeman wrote: post #132

Thanks, I hate it. I'm losing every single game (I think I tried like about 30-ish times). If that's how 12-years old are playing them I should probably apply for disability benefits. My naive approach by trying to define safe squares to move pieces to and naive capture goals by exchanging figures of lesser value is not working against this algorithm, I guess I'll probably have better chance if I'd just run the same…

You are using something like the same logic as the algorithm itself and describing how a newcomer to the game would play chess. I think this is what they mean by "12 year old". In fact the algorithm probably beats this strategy as your approach probably doesn't identify 2 move 'tactics'.

If you fancy learning more about the game take a look at tactics, and maybe try a tactics trainer. The next level of play is about understanding things like forks and discovered attacks where you aren't assuming that the other player is just going to make huge blunders that give you pieces.

Re: The Kilobyte’s Gambit

#137

Post by jkonline »

sickygnar wrote: post #111
haskal wrote: post #97

Earlier quoted context omitted.

If you play a defensive setup like push D, E pawns one step ahead, fianchetto both the bishops, try to castle early, the computer just flings all the pawns forward, even from the king side: https://imgur.com/a/Nt4x7TP Anti-computer tactics in chess: https://en.wikipedia.org/wiki/Anti-computer_tactics EDIT: GothamChess used a similar system to defeat a bot: https://www.youtube.com/watch?v=l5MD6hn5PgI (from 1:59)

the almighty hippopotomus. love playing it against humans too (with limited success)

"Hip... Hip hop... Hip hop Anonymous?"

Your comment reminded me of one of my favorite scenes from the great movie, Big Daddy.

https://youtu.be/KnbL56_S9rQ

Re: The Kilobyte’s Gambit

#140

Post by EnockNitti »

A very interesting piece if code.

But I hate the graphic, lost several games due to me confusing pieces.

I used to play some when I was as kid, and some more when the kids were small. Has no idea of my rating.

If I remove the games I lost due to me confusing pieces I am up 5-1 and 1 draw.

I think I have found out its weaknesses.

1. In situation with pending exchange of several piece the 4 ply algorithm can't check the exchange-chain to its end and therefore makes mistakes.

2. Don't care very much for its pawns, in 2 endgames I was up 2 pawns that I was enabled to promote.

3 As other has pointed out, weak at defending the king in the beginning. 2 wins to what I call "extended" fools mate.

Post reply on HN