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

Blunder Free Chess – visualize which squares are attacked

taonexus.com

Post reply on HN
Print view View original

31–40 of 52 posts

Re: Blunder Free Chess – visualize which squares are attacked

#31

Post by logicallee »

primitivesuave wrote: post #3

I'm curious as to why you wrote your own chess implementation - seems like this would have been much easier to implement with chessboard.js [1] and chess.js [2]. 1. https://chessboardjs.com/ 2. https://github.com/jhlywa/chess.js/blob/master/README.md

>I'm curious as to why you wrote your own chess implementation

I was investigating whether AI can write a complete game of chess in one shot, I livestreamed that here[1] and the answer is no (conclusion: major rules are there, but buggy), but I was able to get it into a working state after many hours of back and forth.

The original implementation froze whenever a pawn got to the seventh rank, and allowed castling out of and through check (illegal), it was very buggy.

Iterating with the AI, I added tons of stuff like drag to move pieces, mobile support, and, yes, computer play. When that was finished (it is available at: https://taonexus.com/chess.html ) I had it make this blunder-free variant and it got it pretty fast.

[1] https://www.youtube.com/watch?v=AhJ07anVkYk

Re: Blunder Free Chess – visualize which squares are attacked

#32

Post by alabastervlog »

dpc_01234 wrote: post #18

Most basic blunders maybe? In a normal game a piece will be attacked and defended by multiple other pieces, with various values, sometimes with some dependencies between them, and I would consider messing this part a relatively trivial blunder as well.

I've played hundreds of games and I'm not sure a single one hasn't been decided by a "basic blunder", and I mean very basic. I bet the only games in which I've not committed one myself (and only gone on to win if it was overlooked, which they often are) were ones in which I fool's-mated someone.

There's chess-people's experience of the game, then there's how every single other person, including those who play it quite a bit, experiences the game. A "normal game" in the sense of the average of all games played, maybe that features very few basic blunders, just because chess people play so very many games (though huge numbers of people play a lifetime count of like ten to one thousand games, and never move out of basic blunders being pretty common, so I wouldn't bet on that describing an average game, but maybe) but the average person's experience of the game is that it's entirely about basic blunders, like, that's the main thing on which the game turns.

Re: Blunder Free Chess – visualize which squares are attacked

#34

Post by alabastervlog »

lucb1e wrote: post #15
mkoryak wrote: post #4

Earlier quoted context omitted.

They wrote their own because it's more fun to do that, and also because they really wanted me to win. The computer is really really bad

Hey now, I was proud to have won from a chess computer for once!

Glad I'm not the only one who often can't beat even the "easy" computer opponent in chess programs.

The main problem is that I only win by having my extremely-stupid blunders overlooked and then spotting my opponent's extremely-stupid blunder, so if the computer never misses my boneheaded blunders and also almost never commits any exploitable-in-one-or-rarely-two-moves blunders of their own, I can't possibly win, zero chance.

Re: Blunder Free Chess – visualize which squares are attacked

#35

Post by JohnMakin »

dyingkneepad wrote: post #25

False advertisement. I played a single match and the AI very very quickly blundered their queen to my bishop, and later a bunch of other pieces. This bot way is worse than Martin (250) form chess.com. Also, those colors are horrible to see. And it's really hard to see which piece the AI is moving each turn.

I think it's fairly obvious the name is to help the user prevent blunders, not that blunders are literally impossible by both players.

Re: Blunder Free Chess – visualize which squares are attacked

#36

Post by muxxa »

There is a big problem in that the purple is much less visible when it appears on top of your own piece vs. a free square. I would suggest instead to color code it red when one of your own pieces is being attacked. There is no need to show the squares your pieces are attacking IMO (yellow heatmap)

Why not take this further to count up the simple point value of attacking/defending pieces, so e.g. a pawn that is attacked by a pawn but defended by a pawn won't show red, whereas an attacked but undefended knight would show '-3 redness'. Also if this was just briefly 'flashed' after the opponents move, or after you reopen a browser tab, it might serve as a nice reminder of the positional situation without interfering with regular/plain board visualisation, and without making the player reliant on it.

Re: Blunder Free Chess – visualize which squares are attacked

#37

Post by blagie »

logicallee wrote: post #27
dyingkneepad wrote: post #25

False advertisement. I played a single match and the AI very very quickly blundered their queen to my bishop, and later a bunch of other pieces. This bot way is worse than Martin (250) form chess.com. Also, those colors are horrible to see. And it's really hard to see which piece the AI is moving each turn.

>those colors are horrible to see After someone gave me feedback on lichess, I tried a heatmap instead (with green for white and red black), I'm not sure it's an improvement but you can see if you like it any better: https://taonexus.com/blunderfreechess2.html

I can see the board! It's a huge improvement.

One more improvement. There needs to be some way to see the move the computer made. If you blink, a piece is somewhere else, and you have no idea what happened.

I like the concept of overlaying visualizations on chess to help people understand what's going on, but it will take a lot of playing with what to show and how to show it to make it useful rather than annoying.

Re: Blunder Free Chess – visualize which squares are attacked

#38

Post by renox »

mkoryak wrote: post #4
primitivesuave wrote: post #3

I'm curious as to why you wrote your own chess implementation - seems like this would have been much easier to implement with chessboard.js [1] and chess.js [2]. 1. https://chessboardjs.com/ 2. https://github.com/jhlywa/chess.js/blob/master/README.md

They wrote their own because it's more fun to do that, and also because they really wanted me to win. The computer is really really bad

ROFL you're right that the computer is really bad, I won in 7 round which is a first for me (against a computer) and I'm bad at chess.

It reminds of a reversi I made for a school projects: users liked it because the computer was so bad that they could win..

Re: Blunder Free Chess – visualize which squares are attacked

#39

Post by jay_kyburz »

logicallee wrote: post #27
dyingkneepad wrote: post #25

False advertisement. I played a single match and the AI very very quickly blundered their queen to my bishop, and later a bunch of other pieces. This bot way is worse than Martin (250) form chess.com. Also, those colors are horrible to see. And it's really hard to see which piece the AI is moving each turn.

>those colors are horrible to see After someone gave me feedback on lichess, I tried a heatmap instead (with green for white and red black), I'm not sure it's an improvement but you can see if you like it any better: https://taonexus.com/blunderfreechess2.html

Whats the range? 0 to 4 or 5 ? I would try x small dots, or even numbers. You could make black red squares, and white green circles.

Re: Blunder Free Chess – visualize which squares are attacked

#40

Post by zamalek »

ViktorRay wrote: post #5

This is a cool program. What system is it using for the “play with computer”? Is it custom or one of the available chess engines? I wonder if there is a way to alter the difficulty of the computer program.

What could be interesting is playing against a full-strength engine, that is forced to sacrifice a piece every N moves (I initially wrote "blunder," but the top sacrificial Stockfish move would likely still be a viable strategy).
Post reply on HN