OSDev.org

I don't think that's just Americans, Schol. Most people think even the rules of chess are too complicated to learn and they never even try.

Hm... I'd like to see some statistics on that. Ad hoc, I'd say about 95% of the people I know personally could sit down for a spontaneous casual match. Most of them will have next to no experience, be fuzzy about how exactly castling works, and might not have heard about taking a pawn en passant (something I am fuzzy about the specifics of...). They will probably play rather badly, and probably prefer to play something else, but they would be able to play the game without mucking up. The rules are pretty simple, it's just that playing good is hard.

We actually played a LARP once where the forces of good and evil ended up as figures on a chess board, and two figures meeting on the same square would duel it out in a ring, the winner remaining on the board. (Who still remembers the computer game Archon?) All present were familiar enough with the rules to cast quite anxious glances around about whom they might meet in the ring soon, while the two "stewards" pondered their next move...

I think Solar is right, I know the rules just fine, and I think most people do. But I'd never call myself a chess player because even the basics of chess strategy are a total mystery.

StudlyCaps wrote:I think Solar is right, I know the rules just fine, and I think most people do. But I'd never call myself a chess player because even the basics of chess strategy are a total mystery.

I think, the basics are covered by lots of practice. It's a skill.

Hi,

StudlyCaps wrote:I think Solar is right, I know the rules just fine, and I think most people do. But I'd never call myself a chess player because even the basics of chess strategy are a total mystery.

For chess, "strategy" is mostly basic game theory. Start by analysing the board and giving it a rating (e.g. maybe based on how defended/exposed your pieces are, how defended/exposed your opponents pieces are, how many of your opponents pieces can be attacked, how many of your pieces opponents can attacked, and possibly with an underlying "this type of piece is worth this many points" system). Then for each move you can make, see if it improves the board rating, recursively (e.g. for each move you can make, for each response opponent can make, for each move you'd be able to make after that, ..). Finally make the move that improves the board rating the most.

Of course when computers do it it's typically "brute force search with optimisations"; and when humans do it they use a lot more estimation, assumption and "gut feeling".

The main differences between beginners and experienced chess players is "depth of recursion" and how accurate their estimations/assumptions/feelings are.


Cheers,

Brendan

Brendan wrote:Hi,
For chess, "strategy" is mostly basic game theory. Start by analysing the board and giving it a rating (e.g. maybe based on how defended/exposed your pieces are, how defended/exposed your opponents pieces are, how many of your opponents pieces can be attacked, how many of your pieces opponents can attacked, and possibly with an underlying "this type of piece is worth this many points" system). Then for each move you can make, see if it improves the board rating, recursively (e.g. for each move you can make, for each response opponent can make, for each move you'd be able to make after that, ..). Finally make the move that improves the board rating the most.

Of course when computers do it it's typically "brute force search with optimisations";

I think the new "generation" of AI chess algorithms are a little different. Rather than pre-defining the "weight" of the pieces and the squares, I think they look at the board more like a bitmap image, and break it down into patterns, kind of like a facial-recognition, or object-recognition algorithm.

Here's a high level overview: https://www.sciencealert.com/it-took-4- ... -alphazero

But essentially the bottom line is that the new AI designs do not assign any "values" at all. They let the AI come up with it's own solutions. And the game that it's playing doesn't really matter. It would learn chess and checkers just as well, with no code changes. (Other than maybe the rules of the game.)

Hi,

SpyderTL wrote:

Brendan wrote:For chess, "strategy" is mostly basic game theory. Start by analysing the board and giving it a rating (e.g. maybe based on how defended/exposed your pieces are, how defended/exposed your opponents pieces are, how many of your opponents pieces can be attacked, how many of your pieces opponents can attacked, and possibly with an underlying "this type of piece is worth this many points" system). Then for each move you can make, see if it improves the board rating, recursively (e.g. for each move you can make, for each response opponent can make, for each move you'd be able to make after that, ..). Finally make the move that improves the board rating the most.

Of course when computers do it it's typically "brute force search with optimisations";

I think the new "generation" of AI chess algorithms are a little different. Rather than pre-defining the "weight" of the pieces and the squares, I think they look at the board more like a bitmap image, and break it down into patterns, kind of like a facial-recognition, or object-recognition algorithm.

Here's a high level overview: https://www.sciencealert.com/it-took-4- ... -alphazero

But essentially the bottom line is that the new AI designs do not assign any "values" at all. They let the AI come up with it's own solutions. And the game that it's playing doesn't really matter. It would learn chess and checkers just as well, with no code changes. (Other than maybe the rules of the game.)

From the paper (linked to from the article you mentioned); it looks like its almost exactly the same as what I described except for the addition of a neural network to auto-guess various parameters (the weights of the pieces, etc); mostly by playing lots of games against itself and while using a neural network to tweak a formula used to calculate board ratings.

Note that for computer chess engines the goal should be to lose about 75% of the time (e.g. by estimating the opponents skill and adjusting the AI's "search depth" to suit). It's far too easy for stupid people to write chess engines that ruin the game for humans (guaranteed failure is not something most humans enjoy). With this in mind, except for a little publicity ("Our unenjoyable failure beat their unenjoyable failure!") Google's research has no practical value.


Cheers,

Brendan

Hi,
Having had many American mentors and teachers, I would not utter anything bad against them as some of the other posters. There are lot of brilliant American chess players with Bobby Fisher and Paul Morphy being dominating World Champions in their prime.

The main differences between beginners and experienced chess players is "depth of recursion" and how accurate their estimations/assumptions/feelings are.

This is an extremely simplistic assessment. What makes strong chess players "strong" would be an active field of research because no one really knows it. Even with lot of games or even a lifetime of play many people do not break their class. It is lot of hard work to improve at chess after you have hit a plateau ( just like music ). You also discount game psychology. Some people are simply naturally strong in this area. Young American chess player Daniel Naroditsky has done some good work in this area. I am talking mainly about human aspects of chess above.
--Thomas

Brendan wrote: Note that for computer chess engines the goal should be to lose about 75% of the time (e.g. by estimating the opponents skill and adjusting the AI's "search depth" to suit). It's far too easy for stupid people to write chess engines that ruin the game for humans (guaranteed failure is not something most humans enjoy). With this in mind, except for a little publicity ("Our unenjoyable failure beat their unenjoyable failure!") Google's research has no practical value.

As far as a "game" can have practical value..

My impression is that the introduction of these modern AI opponents has changed (or may change) the way that people (grand masters, etc.) look at the game, itself. For instance, maybe the board "values" would be best based on some criteria that no one has ever used before, like finding least number of possible win solutions for your opponent, rather than the most (or quickest) win solutions for the player. But by playing millions of games a day, and analyzing the results, you may start to see patterns emerge that are completely unexpected, like maybe having less pieces actually improves your overall outcome.

But it's more exiting (slash terrifying) when you start to apply this to other problem domains, like military strategies, or traffic patterns.

SpyderTL wrote:But it's more exiting (slash terrifying) when you start to apply this to other problem domains, like military strategies, or traffic patterns.

People's online or real-world behavioral patterns. Feeding directly into insurance rates, chances of employment, or attention by the police / intelligence agencies.

All that's not tomorrow, that's yesterday. A coworker of mine is positively happy that his insurance is offering him lower (starting...) rates in exchange for being allowed to track his behavior in traffic, feeding the data into undisclosed algorithms to figure out what his real rates should be. Of course he's conviced that it will be others that will have to pay more... (because of what, exactly?)

I started my chess blog, http://sandmanchess.com where i have some games in which I trolled titled players online ( basically i scored upset wins over strong opponents ). I have lost bunch of tournaments recently, I am trying to pick myself up again. I am finally engaged to someone. That with full time job means it is practically curtains to my chess career. Nevertheless, i will try to post interesting chess content in my website.

--Thomas

"Chess like love, is infectious at any age"
- Salo Flohr

"Chess, like love, like music, has the power to make men happy"
- Tarrasch.

Proceed with your quest Thomas. Your love for the game matters more than anything else.

Chess is my favorite game my Grand father taught me how to play chess.