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

Reimagining Chess with AlphaZero

cacm.acm.org

Post reply on HN
Print view View original

101–110 of 116 posts

Re: Reimagining Chess with AlphaZero

#101

Post by vikingerik »

tehnub wrote: post #19
discardable_dan wrote: post #18

Earlier quoted context omitted.

I can't tell if this is satire. Draws by insufficient material almost never occur in Grandmaster chess. I have not seen a single grandmaster game end this way in a year of watching top-level chess games daily. If you inspect all of the dozens of draws in this world's Tata Steel, which included ~300 GM-level games, you won't find a single that ends this way.

>If you inspect all of the dozens of draws in this world's Tata Steel, which included ~300 GM-level games, you won't find a single that ends this way. Here's one: https://www.chess.com/events/2022-tata-steel-chess-tournamen... Draws by insufficient material are quite common, but I wouldn't call them the bane of modern chess, or anything close to that.

Well, GMs agree on a draw well before it actually gets down to insufficient material. Many draws-by-agreement would degenerate into that if they kept playing.

Re: Reimagining Chess with AlphaZero

#102

Post by omoikane »

A very different variation is to keep all the same rules, but remove the turn-based nature of chess and allow pieces to be moved simultaneously. One implementation is here: https://kungfuchess.org/

Their twitter account has some videos on what the games look like, for example: https://twitter.com/KungFuChessOrg/status/142929554275706061...

Re: Reimagining Chess with AlphaZero

#104

Post by light_hue_1 »

adgjlsfhk1 wrote: post #26
light_hue_1 wrote: post #22

Earlier quoted context omitted.

Top-level chess leads to draws because of how points are awarded, not because of the rules. Yes, sometimes you just run out of material, but much of the time draws happen because players shut down the game or agree to a draw in a position that is just too dangerous to both parties. Chess rewards draws with far too many points. If you could agree to a draw in tennis, you would see a lot of draws where players want to…

the draw problem isn't entirely human. computers draw something like 80% of games even with openings specifically chosen to reduced the number of draws.

Humans and machines draw for totally different reasons.

Machines don't have any concept of a tournament. Humans go into drawn lines and get rid of tension in order to play the tournament meta-game. Since draws are worth so much, it's often a good idea to just bail out with a draw.

A lot of human positions wouldn't be drawn by machines, they would keep hammering the opponent.

Re: Reimagining Chess with AlphaZero

#105

Post by suyjuris »

If you find chess variants interesting, I would recommend looking at Go [1] as well. It is similar to chess in that it is an ancient turn-based strategy game for two players and with perfect information, but I find it much more pleasing from a game-design perspective. The rules are much simpler, there are no draws, and the starting player does not have an advantage. Also, you can have handicaps so that two players with different playing strengths play a evenly-matched game.

To me, this shows that it is possible to solve most of the commonly cited issues in chess by changing it into a radically different game. But can you do it with only a minute adjustment? I find it difficult to imagine chess without draws, but it is not inconceivable. [2]

[1] https://en.wikipedia.org/wiki/Go_(game)

[2] E.g.: if the game would end in a draw, instead the player who has made more King-moves wins. If that is tied as well, Black wins.

Re: Reimagining Chess with AlphaZero

#106

Post by gundmc »

sorenn111 wrote: post #99
robrenaud wrote: post #94

Is anyone interested in doing this kind of analysis for games that are in development? I am a SWE with 16 years of experience at Google with lots of practical/applied ML. I am about to be funemployed, and I'd love to incorporate an AI player/analysis into the playtest feedback loop of a board/card game in development. I've won the Race for the Galaxy tournament 4 times in 7 attempts at the World Boardgaming Champions…

Start-up/Consulting idea: use ML/RL strategies and tools and help gaming companies improve their AI. In a variety of games, the AI should be a big component of gameplay (all single player games) and ends up being such a liability. I can imagine the litany of issues that prevent this, but I wonder if a company could accumulate some general templates/models/toolkits that could help AI improve for a variety of gaming co…

An AlphaZero type AI for Civ is my absolute dream. A huge problem in these sorts of games is that developers have the computer players cheat to artificially raise the difficulty to compensate for the poor skill and decision making of the AI.

Re: Reimagining Chess with AlphaZero

#107

Post by tehnub »

vikingerik wrote: post #101
tehnub wrote: post #19

Earlier quoted context omitted.

>If you inspect all of the dozens of draws in this world's Tata Steel, which included ~300 GM-level games, you won't find a single that ends this way. Here's one: https://www.chess.com/events/2022-tata-steel-chess-tournamen... Draws by insufficient material are quite common, but I wouldn't call them the bane of modern chess, or anything close to that.

Well, GMs agree on a draw well before it actually gets down to insufficient material. Many draws-by-agreement would degenerate into that if they kept playing.

Well yeah, they usually do that, but draws by insufficient material still happen pretty frequently. Also, king vs. king is not the only way it can end; you'll see king vs. king and knight (or bishop) pop up pretty often too.

Re: Reimagining Chess with AlphaZero

#108

Post by kzrdude »

Andrew_nenakhov wrote: post #91
kzrdude wrote: post #81

Earlier quoted context omitted.

I see. I found the rapid world championship, and at the top level the number of draws is certainly higher. (And it's played with an increment). See this table for example: https://chess-results.com/tnr600852.aspx?lan=6&art=5&flag=30 highlighting ½ with the search gives an idea of the number of draws.

First placed MC had 7+ 1- 5=, doesn't look like a draw death to me.

I agree, this is why I like rapid.

Re: Reimagining Chess with AlphaZero

#109

Post by pests »

londons_explore wrote: post #40
riwsky wrote: post #5

Earlier quoted context omitted.

Software rendering of these rules might lag a bit—but if you have dedicated hardware, it should be possible to make this work.

Help me understand... Why would I need dedicated hardware to render chess moves that are slightly different to regular chess moves...? Your comment makes no sense...

It flew over my head at first too. Subtle joke.

Re: Reimagining Chess with AlphaZero

#110

Post by pests »

istinetz wrote: post #46
pests wrote: post #3

Great read. Interesting variations to the rules. Here's an annotated quote: "The nine changes considered in this study are listed in Table 1. No-castling and No-castling (10) involve a full and partial [not allowed during first 10 turns / 20 plies] restriction on the castling rule. Pawn-one-square, Semi-torpedo [can move 2 squares on 2ed and 3rd ranks], Torpedo [can move 2 squares], Pawn-back [up to your 2ed rank, do…

>Do any chess games support these modes or is these possible to play today? Chess.com currently supports most of the modes in the list - except, I think, semi-torpedo, no-castling (10), and pawn-back. Which, to me, are the least exciting mods anyway.

I found pawn-back to be the most interesting just from the changes to existing chess logic. Pawn-back resulted in the most change in piece-value from classical and during simulations were used 100% of the time.
Post reply on HN