Rustam Kasimdzhanov proposal

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
Don
Posts: 5106
Joined: Tue Apr 29, 2008 2:27 pm

Rustam Kasimdzhanov proposal

Post by Don » Tue Jun 04, 2013 12:07 pm

This is an old proposal dating back to 2011, but I think a very good one.

http://en.chessbase.com/home/TabId/211/PostId/4007387

Rustam advocates eliminating draws completely from chess by using a tie-break system similar to what is done in tennis.

I would love to see some computer chess tournaments set up using that system.

Don
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.

Uri Blass
Posts: 9111
Joined: Wed Mar 08, 2006 11:37 pm
Location: Tel-Aviv Israel

Re: Rustam Kasimdzhanov proposal

Post by Uri Blass » Tue Jun 04, 2013 1:05 pm

I do not like it because it means chess is more about blitz and players who are relatively worse at blitz are going to lose rating.

I think that it may be more interesting with no draw but another rule.

If no side can give checkmate then the side who made the last capture wins.
If no side made a capture then black wins.

the repetition rule and the fifty move rule and the stalemate rule remain (the first 2 in order to prevent games that are too long) but one of the players can claim a win by them and not a draw.

User avatar
Don
Posts: 5106
Joined: Tue Apr 29, 2008 2:27 pm

Re: Rustam Kasimdzhanov proposal

Post by Don » Tue Jun 04, 2013 1:48 pm

Uri Blass wrote:I do not like it because it means chess is more about blitz and players who are relatively worse at blitz are going to lose rating.
My thinking is that this is a good thing though. If you are to be considered a top notch champion you should be well rounded. If you cannot think on your feet it's a weakness.

To use Rustam's analogy to tennis, if you follow the game, you will know that there are 3 basic surfaces that tennis is played on, grass, clay and hard courts. Each surface favors a certain type of player. Grass favors the aggressive serve and volley player and clay favors the player who likes to stand back and keep the ball in play. However to be considered one of the top players you must be able to play all surfaces well. The greats of tennis could win on all surfaces.

So I see no problem with incorporating SOME faster chess into the equation - and the players will adapt. It might favor some players slightly over others but so what? Who is to say that it should be just the one way?

I think that it may be more interesting with no draw but another rule.

If no side can give checkmate then the side who made the last capture wins.
If no side made a capture then black wins.
The problem with rules like that is that they actually change the game - we don't want to invent a completely new game which is not really chess. Also, the rules you propose are not in the spirit of the game because making the last captures is almost a random rule.

the repetition rule and the fifty move rule and the stalemate rule remain (the first 2 in order to prevent games that are too long) but one of the players can claim a win by them and not a draw.
Rules for eliminating or reducing draws have been around for decades and none of them seem to be very helpful or else change the game very drastically and would get tremendous resistance.

One rule that I always believed would make great sense is to consider a stalemate a win. I have had people tell me that would have very little impact on the draw rate, but I think it would have a non-trivial impact, even if not a huge one. But it would overturn theory and make all current chess literature obsolete.

In Go you are not allowed to repeat a position - it's consider illegal. A rule like that would prevent draws by repetition but I think a lot of players would have a difficult time determining (in the long maneuvering drawn endings) when a position is going to repeat and it would be difficult to prove it without going over the game. But I think this rule and the stalemate rule effectively would eliminate most draws. All that would be remaining are the 50 move draws.
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.

lkaufman
Posts: 4884
Joined: Sun Jan 10, 2010 5:15 am
Location: Maryland USA
Contact:

Re: Rustam Kasimdzhanov proposal

Post by lkaufman » Tue Jun 04, 2013 2:16 pm

Don wrote:
Uri Blass wrote:I do not like it because it means chess is more about blitz and players who are relatively worse at blitz are going to lose rating.
My thinking is that this is a good thing though. If you are to be considered a top notch champion you should be well rounded. If you cannot think on your feet it's a weakness.

To use Rustam's analogy to tennis, if you follow the game, you will know that there are 3 basic surfaces that tennis is played on, grass, clay and hard courts. Each surface favors a certain type of player. Grass favors the aggressive serve and volley player and clay favors the player who likes to stand back and keep the ball in play. However to be considered one of the top players you must be able to play all surfaces well. The greats of tennis could win on all surfaces.

So I see no problem with incorporating SOME faster chess into the equation - and the players will adapt. It might favor some players slightly over others but so what? Who is to say that it should be just the one way?

I think that it may be more interesting with no draw but another rule.

If no side can give checkmate then the side who made the last capture wins.
If no side made a capture then black wins.
The problem with rules like that is that they actually change the game - we don't want to invent a completely new game which is not really chess. Also, the rules you propose are not in the spirit of the game because making the last captures is almost a random rule.

the repetition rule and the fifty move rule and the stalemate rule remain (the first 2 in order to prevent games that are too long) but one of the players can claim a win by them and not a draw.
Rules for eliminating or reducing draws have been around for decades and none of them seem to be very helpful or else change the game very drastically and would get tremendous resistance.

One rule that I always believed would make great sense is to consider a stalemate a win. I have had people tell me that would have very little impact on the draw rate, but I think it would have a non-trivial impact, even if not a huge one. But it would overturn theory and make all current chess literature obsolete.

In Go you are not allowed to repeat a position - it's consider illegal. A rule like that would prevent draws by repetition but I think a lot of players would have a difficult time determining (in the long maneuvering drawn endings) when a position is going to repeat and it would be difficult to prove it without going over the game. But I think this rule and the stalemate rule effectively would eliminate most draws. All that would be remaining are the 50 move draws.
Making stalemate a win is logical and would help a bit to reduce draws. It would not change opening theory significantly but of course would radically change endgame theory. A less drastic version that would have a much better chance of being accepted is to count stalemate as 3/4 of a point or so. As for repetitions, banning all of them is not very practical, but banning repetitions due to perpetual check is easy to do and is done in all the Asian chess variants. This would do more than the stalemate rule to reduce draws. It would of course impact opening theory somewhat. Yet another helpful rule is to restore the bare king rule of old, which is that a player left with only a king loses unless he also bares the enemy king immediately. These three rules together with forbidding draws by agreement before move 40 might knock out nearly half the draws in GM play without radically altering the game.

User avatar
Don
Posts: 5106
Joined: Tue Apr 29, 2008 2:27 pm

Re: Rustam Kasimdzhanov proposal

Post by Don » Tue Jun 04, 2013 2:48 pm

lkaufman wrote:
Don wrote:
Uri Blass wrote:I do not like it because it means chess is more about blitz and players who are relatively worse at blitz are going to lose rating.
My thinking is that this is a good thing though. If you are to be considered a top notch champion you should be well rounded. If you cannot think on your feet it's a weakness.

To use Rustam's analogy to tennis, if you follow the game, you will know that there are 3 basic surfaces that tennis is played on, grass, clay and hard courts. Each surface favors a certain type of player. Grass favors the aggressive serve and volley player and clay favors the player who likes to stand back and keep the ball in play. However to be considered one of the top players you must be able to play all surfaces well. The greats of tennis could win on all surfaces.

So I see no problem with incorporating SOME faster chess into the equation - and the players will adapt. It might favor some players slightly over others but so what? Who is to say that it should be just the one way?

I think that it may be more interesting with no draw but another rule.

If no side can give checkmate then the side who made the last capture wins.
If no side made a capture then black wins.
The problem with rules like that is that they actually change the game - we don't want to invent a completely new game which is not really chess. Also, the rules you propose are not in the spirit of the game because making the last captures is almost a random rule.

the repetition rule and the fifty move rule and the stalemate rule remain (the first 2 in order to prevent games that are too long) but one of the players can claim a win by them and not a draw.
Rules for eliminating or reducing draws have been around for decades and none of them seem to be very helpful or else change the game very drastically and would get tremendous resistance.

One rule that I always believed would make great sense is to consider a stalemate a win. I have had people tell me that would have very little impact on the draw rate, but I think it would have a non-trivial impact, even if not a huge one. But it would overturn theory and make all current chess literature obsolete.

In Go you are not allowed to repeat a position - it's consider illegal. A rule like that would prevent draws by repetition but I think a lot of players would have a difficult time determining (in the long maneuvering drawn endings) when a position is going to repeat and it would be difficult to prove it without going over the game. But I think this rule and the stalemate rule effectively would eliminate most draws. All that would be remaining are the 50 move draws.
Making stalemate a win is logical and would help a bit to reduce draws. It would not change opening theory significantly but of course would radically change endgame theory. A less drastic version that would have a much better chance of being accepted is to count stalemate as 3/4 of a point or so. As for repetitions, banning all of them is not very practical, but banning repetitions due to perpetual check is easy to do and is done in all the Asian chess variants. This would do more than the stalemate rule to reduce draws. It would of course impact opening theory somewhat. Yet another helpful rule is to restore the bare king rule of old, which is that a player left with only a king loses unless he also bares the enemy king immediately. These three rules together with forbidding draws by agreement before move 40 might knock out nearly half the draws in GM play without radically altering the game.
I think all of these rules could easily be incorporated into computer chess though and it would be a way to check the draw reduction rate of the rules.

The last I checked, Komodo draws about 70% of the games against itself when using a time control of 1m + 1s fischer. With even more time this number goes up and with about half that amount of time it is about 65% draws.

In just a very few years we are looking at mostly draws in computer chess.

Don
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.

tpetzke
Posts: 686
Joined: Thu Mar 03, 2011 3:57 pm
Location: Germany
Contact:

Re: Rustam Kasimdzhanov proposal

Post by tpetzke » Tue Jun 04, 2013 3:35 pm

A tennis match is much easier to understand to the average viewer than a high quality chess game. I think this makes it more attractive because people can follow and talk about it. More action and sometimes the participation of well known female players help certainly too.

Thomas...

User avatar
Don
Posts: 5106
Joined: Tue Apr 29, 2008 2:27 pm

Re: Rustam Kasimdzhanov proposal

Post by Don » Tue Jun 04, 2013 4:32 pm

tpetzke wrote:A tennis match is much easier to understand to the average viewer than a high quality chess game. I think this makes it more attractive because people can follow and talk about it. More action and sometimes the participation of well known female players help certainly too.

Thomas...
I actually once imagined scoring a chess match as if it were tennis. The player who has the serve is "white", so the first "game" (in the tennis sense) would be a win by two match where the "server" always gets white. This would give the server a nice advantage, just as in tennis. Of course in the second "game", the other player takes the black pieces until one player wins by 2 points.

So you can play it identically to tennis, whoever is "serving" takes white.

Because of draws and the fact that a tennis point take very little time compared to a single game of chess, I would consider 30-0 to be a win for the server so that you are only required to win by 2 points period. In tennis you have to have to win by 2 AND win at least 4 points. That does make it a bit more exciting because the server (or his opponent) can win the first 3 points and you still have a chance to come from behind.

The other rule I would probably impose (for the sake of patience) is something seen in tennis clubs where time is limited. After a score of duece (at least 6 points played and you are tied) the next point wins the game. For chess with draws it would have to be that after this you only have to win by 1. Maybe even better is that you only have to win by 1 point period (but starting with the second "point")

You could also say that a match consists of only 4 games instead of 6 and you could use the same tennis tie break rule at 4-4 or 6-6 or a shortened version of it.

I know, it doesn't quite make sense in chess, but it's just a mental exercise. The thing about how tennis is scored however is that it's incredibly exciting, you can have numerous critical points through the entire match. For example the "break of serve" is a serious threat but then "breaking back" is terribly exciting to the spectators. Tennis has a way of packing many "mini-matches" into a single match with tons of drama. The only element that might be missing is having the computers scream at the judges and throw temper tantrums.
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.

Albert Silver
Posts: 2975
Joined: Wed Mar 08, 2006 8:57 pm
Location: Rio de Janeiro, Brazil

Re: Rustam Kasimdzhanov proposal

Post by Albert Silver » Tue Jun 04, 2013 5:20 pm

Don wrote:This is an old proposal dating back to 2011, but I think a very good one.

http://en.chessbase.com/home/TabId/211/PostId/4007387

Rustam advocates eliminating draws completely from chess by using a tie-break system similar to what is done in tennis.

I would love to see some computer chess tournaments set up using that system.

Don
I think it is a bad idea.

Uri's comment on favoring blitz players is valid. Comparing 5 minute games with 6 hour games is ridiculous. It is like taking a marathon runner and demanding he prove he is 'well-rounded' by doing 100-meter sprints.

Anyhow, imposing artificial conditions to force a win makes even less sense to me as I think the game is ultimately a draw with best play. I do not claim we are anywhere near that, but I do think forcing a result where there is none mathematically is a mistake.

Albert
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."

User avatar
Don
Posts: 5106
Joined: Tue Apr 29, 2008 2:27 pm

Re: Rustam Kasimdzhanov proposal

Post by Don » Tue Jun 04, 2013 6:40 pm

Albert Silver wrote:
Don wrote:This is an old proposal dating back to 2011, but I think a very good one.

http://en.chessbase.com/home/TabId/211/PostId/4007387

Rustam advocates eliminating draws completely from chess by using a tie-break system similar to what is done in tennis.

I would love to see some computer chess tournaments set up using that system.

Don
I think it is a bad idea.

Uri's comment on favoring blitz players is valid. Comparing 5 minute games with 6 hour games is ridiculous. It is like taking a marathon runner and demanding he prove he is 'well-rounded' by doing 100-meter sprints.
I understand the analogy, but it's a poor one. I saw a study once and I'm looking for it as a reference and cannot find it now, but the conclusion was that there was little difference in blitz chess ELO compared to long time control ELO for most players. Yes, I know that 95% of all players will swear up and down that they play horrible 5 minute chess but that is probably because it's true. But what is relevant is how you play compared to other players and that is more or less in line with your ELO rating even though everyone has anecdotes that prove they suck at fast chess. I know that I felt the same and ended up in this group of players at the club that played speed chess all day long and considered themselves specialists at speed chess, you know the type, coffeehouse gambits with a constant supply of crap moves. We were all rated about the same and I didn't like speed chess and wasn't good at it but as it turned out I did better than my rating.

I think almost anyone will agree that sprinters and marathoners have little in common though, I know a lot about this as I used to be a devoted hobby runner. I remember Bill Rodgers saying he couldn't run a 60 second 440 but every high school has kids that can beat 60 seconds. If Bill Rodgers could be compared to a Grandmaster this would be saying that a Grandmaster cannot be expected to beat high school chess players at speed chess.

After my speed chess experience I talked to the local master about this and it was his opinion and his experience that if you are good at one you are good at the other and vice versa and that these imagined disparities do not exist or at least they are really rare.

Having said all of that, you must weigh the potential benefits of any given proposal. This does create some logistical problems but I wouldn't be so quick to dismiss it. I guarantee you that if this is how it had been done for the past 20 years there would be an uproar if we considered changing it what we do now! It's just conservative human nature. But his points really resonated with me. Here is one comment he made:

Still, in a well-organized tournament, top players, getting up to go to their hotel rooms after a ten minute draw, do not add attractiveness to chess.


He actually proposed a real and workable solution that doesn't change the game at all.

Anyhow, imposing artificial conditions to force a win makes even less sense to me as I think the game is ultimately a draw with best play. I do not claim we are anywhere near that, but I do think forcing a result where there is none mathematically is a mistake.

Albert
There is nothing artificial about this. It's not a proposal to change the rules of the game but it maintains the game in it's exact form. You play each game with the same rules. How you structure a tournament is completely arbitrary ANYWAY, is a round robin "artificial" because it's not a swiss? Is a swiss artificial because it's not a match? Who decided what is "artificial" except conservative people who don't like change? (I'm not referring to you on this by the way, I'm just laying out why nothing ever happens in games that have bureaucratic organizations that control them.)
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.

dchoman
Posts: 171
Joined: Wed Dec 28, 2011 7:44 pm
Location: United States

Re: Rustam Kasimdzhanov proposal

Post by dchoman » Tue Jun 04, 2013 8:31 pm

An interesting variation on this is the idea of "matchplay" where a fixed time control is used to play a sequential series of games (if the first game is drawn), each game uses the player's time remaining from the previous games. First won game wins the 'match'.

http://www.uschess.org/content/view/11446/643

Post Reply