| |
Main
Date: 02 Oct 2008 11:06:48
From: Quadibloc
Subject: Rules for a World Championship Match
|
The question of what would be a good set of rules for a World
Championship match has been raised here.
On the assumption that the rules should be crafted to meet these
objectives -
The victory should preferably be by means of a player obtaining wins
instead of simply defensively forcing draw after draw,
The match should consist of up to a fixed maximum number of games,
The champion should be given only the minimum possible advantage as
required to ensure that the preceding condition can be met,
then, it seems to me fairly easy to construct a set of rules
consistent with these objectives.
Here is such a set, based on the assumption that the fixed maximum
length will be 30 games rather than 24:
A player wins the match if:
That player is the first to win six games.
That player is the first to earn 15 1/2 points, where a loss is 0, a
win is 1, and a draw is 1/2, following normal Chess scoring
conventions.
In addition, at the conclusion of game 30, if each player has 15
points, the outcome of the match will be decided as follows:
If one player has more victories as Black than the other player, the
player with more victories as Black will win.
Otherwise, the champion will retain his title.
One can change "15 1/2", "15", and "30" to "12 1/2", "12", and "24"
for a maximum 24-game match. And the first condition, the first to win
six games, can be changed to whatever number appears reasonable,
whether bigger or smaller.
Perhaps there are other victory conditions that would make more sense,
such as "the first player to win three more games than his opponent
has won", although one would wish to avoid conditions that exclude the
possibility of an exciting come-from-behind victory.
John Savard
|
|
| |
Date: 03 Oct 2008 13:51:47
From: William Hyde
Subject: Re: Rules for a World Championship Match
|
On Oct 2, 7:16=A0pm, "[email protected]" <[email protected] > wrote:
>
> Kasparov=92s argument makes sense, yet Fischer=92s point is still valid.
> Only wins should count. This way, a champion can=92t cling to the title
> by playing for draws.
How many world championship matches have there been in which the
champion clung to the title by playing for draws? Sure, there were
many draws in Petrosian's title matches, but that was his style, draw
often, win occasionally, lose almost never.
William Hyde
|
| |
Date: 03 Oct 2008 04:55:54
From: Quadibloc
Subject: Re: Rules for a World Championship Match
|
On Oct 3, 1:38 am, Offramp <[email protected] > wrote:
> The first to 6 wins, but if at game 24, or at any point after that,
> one player leads by three points (3-0, 4-1, 5-2), he is the winner.
But that isn't _absolutely guaranteed_ not to go on forever (except by
the fact that neither player will _live_ forever, of course). That's
why I think the old-style rules have to be kept as a component for any
scheme that is practical enough to be accepted.
The idea is that the victory by almost all draws can't be totally
abolished, but at least make it so they'll have to *work* for it.
John Savard
|
| |
Date: 03 Oct 2008 00:38:02
From: Offramp
Subject: Re: Rules for a World Championship Match
|
On Oct 2, 7:06 pm, Quadibloc <[email protected] > wrote:
The first to 6 wins, but if at game 24, or at any point after that,
one player leads by three points (3-0, 4-1, 5-2), he is the winner.
|
| |
Date: 02 Oct 2008 19:25:15
From: Quadibloc
Subject: Re: Rules for a World Championship Match
|
On Oct 2, 5:39 pm, Mike Murray <[email protected] > wrote:
> On Thu, 2 Oct 2008 16:16:32 -0700 (PDT), "[email protected]"
>
>
>
> <[email protected]> wrote:
> >THIS CRAZY WORLD OF CHESS by GM Larry Evans (page 14)
> >When Kasparov finally dethroned Karpov in 1986, he promptly struck a
> >blow for chess justice by voluntarily renouncing the rematch clause.
> >But he didn=92t strike the second blow. The format for future title
> >matches returned to the best-of-24 games with draws counting. In 1987
> >he barely saved his title against Karpov on a 12-12 tie. Kasparov gave
> >two reasons for sticking with this system at a symposium we both
> >attended in Madrid:
> >1. Since he had to overcome draw odds when he was the underdog, he saw
> >no reason why the challenger shouldn=92t have to vault the same
> >obstacle.
> >2. Organizers must have a definite budget and solid dates when they
> >book a playing hall, which isn=92t possible in an open-ended match.
> >Kasparov=92s argument makes sense, yet Fischer=92s point is still valid.
> >Only wins should count. This way, a champion can=92t cling to the title
> >by playing for draws.
>
> I don't know if you remember the late Angus Pitt, but in the
> mid-1970s, I had arguments with him about this.
>
> In arguing for x wins, I focused on the possibility of the leader in a
> limited-game match playing for draws to run out the clock.
>
> Angus' counter argument, which turned out to be prophetic, was that
> between two almost evenly matched very strong players, the person
> behind in the match could start playing for draws in hopes of
> exhausting his opponent, and the match would never end.
My suggestion would have the virtue of solving that problem. Try to
have the match settled by x wins, for x fairly small... but also make
the match limited to N games, for N fairly large. So the attempt is
made to make it "for x wins", but a counterbalancing limit is also in
effect. Of course, since the stronger player can now still play really
hard for draws, my suggestion may be too weak to be a real change. But
if it's the best change in the direction of "x wins" that is
practical, it's still worth making, isn't it?
And, of course, there's my other suggestion of "Dynamic Scoring" as a
sort of _komidashi_ for Chess to *really* solve the draw problem. It
may not be a cure-all, but I *think* it at least suggests a new kind
of thing to explore as an option which, if done right, has a chance to
do better than the other proposals we've had up till now.
John Savard
|
| |
Date: 02 Oct 2008 19:19:45
From: Quadibloc
Subject: Re: Rules for a World Championship Match
|
On Oct 2, 5:13 pm, "Wlodzimierz Holsztynski (Wlod)"
<[email protected] > wrote:
> For this reasons
> the rule should be that none of the players is allowed
> to fall behind the other in a game by more than,
> say, 90 minutes.
The Fischer clock, while it isn't quite the same as that, since it
doesn't allow the time per move to be increased by the two players,
still moves a bit towards that ideal in perhaps an easier-to-
understand way.
Since, of course, your suggestion means the two players together could
take a lot of time, though, this is a type of time control that is
unlikely to be accepted for tournaments.
John Savard
|
| |
Date: 02 Oct 2008 19:18:45
From: Quadibloc
Subject: Re: Rules for a World Championship Match
|
On Oct 2, 5:16 pm, "[email protected]" <[email protected] > quoted, in
part:
> 2. Organizers must have a definite budget and solid dates when they
> book a playing hall, which isn=92t possible in an open-ended match.
>
> Kasparov=92s argument makes sense, yet Fischer=92s point is still valid.
> Only wins should count. This way, a champion can=92t cling to the title
> by playing for draws.
And these are the conditions I'm trying to meet. Letting only wins
count means an open-ended match, so one can't totally eliminate the
draw advantage to the champion.
Hence, I decrease it - by combining a "last resort" rule, where the
traditional style of rule applies, but for a long, although finite,
match of 30 games instead of 24 - with an only wins count rule that
should make for a shorter match unless there are an overwhelming
number of draws.
The defect might be that it's too small a move in the direction of
what Fischer wanted, but attempts to make a *bigger* move run into the
problem of making the match open-ended. Which has been tried, and
ended up by causing bad problems. So why not just do what we can until
we can come up with a better idea?
John Savard
|
| |
Date: 02 Oct 2008 16:16:32
From: [email protected]
Subject: Re: Rules for a World Championship Match
|
THIS CRAZY WORLD OF CHESS by GM Larry Evans (page 14)
When Kasparov finally dethroned Karpov in 1986, he promptly struck a
blow for chess justice by voluntarily renouncing the rematch clause.
But he didn=92t strike the second blow. The format for future title
matches returned to the best-of-24 games with draws counting. In 1987
he barely saved his title against Karpov on a 12-12 tie. Kasparov gave
two reasons for sticking with this system at a symposium we both
attended in Madrid:
1. Since he had to overcome draw odds when he was the underdog, he saw
no reason why the challenger shouldn=92t have to vault the same
obstacle.
2. Organizers must have a definite budget and solid dates when they
book a playing hall, which isn=92t possible in an open-ended match.
Kasparov=92s argument makes sense, yet Fischer=92s point is still valid.
Only wins should count. This way, a champion can=92t cling to the title
by playing for draws.
Quadibloc wrote:
> The question of what would be a good set of rules for a World
> Championship match has been raised here.
>
> On the assumption that the rules should be crafted to meet these
> objectives -
>
> The victory should preferably be by means of a player obtaining wins
> instead of simply defensively forcing draw after draw,
>
> The match should consist of up to a fixed maximum number of games,
>
> The champion should be given only the minimum possible advantage as
> required to ensure that the preceding condition can be met,
>
> then, it seems to me fairly easy to construct a set of rules
> consistent with these objectives.
>
> Here is such a set, based on the assumption that the fixed maximum
> length will be 30 games rather than 24:
>
> A player wins the match if:
>
> That player is the first to win six games.
>
> That player is the first to earn 15 1/2 points, where a loss is 0, a
> win is 1, and a draw is 1/2, following normal Chess scoring
> conventions.
>
> In addition, at the conclusion of game 30, if each player has 15
> points, the outcome of the match will be decided as follows:
>
> If one player has more victories as Black than the other player, the
> player with more victories as Black will win.
>
> Otherwise, the champion will retain his title.
>
> One can change "15 1/2", "15", and "30" to "12 1/2", "12", and "24"
> for a maximum 24-game match. And the first condition, the first to win
> six games, can be changed to whatever number appears reasonable,
> whether bigger or smaller.
>
> Perhaps there are other victory conditions that would make more sense,
> such as "the first player to win three more games than his opponent
> has won", although one would wish to avoid conditions that exclude the
> possibility of an exciting come-from-behind victory.
>
> John Savard
|
| | |
Date: 02 Oct 2008 16:39:35
From: Mike Murray
Subject: Re: Rules for a World Championship Match
|
On Thu, 2 Oct 2008 16:16:32 -0700 (PDT), "[email protected]"
<[email protected] > wrote:
>THIS CRAZY WORLD OF CHESS by GM Larry Evans (page 14)
>When Kasparov finally dethroned Karpov in 1986, he promptly struck a
>blow for chess justice by voluntarily renouncing the rematch clause.
>But he didn�t strike the second blow. The format for future title
>matches returned to the best-of-24 games with draws counting. In 1987
>he barely saved his title against Karpov on a 12-12 tie. Kasparov gave
>two reasons for sticking with this system at a symposium we both
>attended in Madrid:
>1. Since he had to overcome draw odds when he was the underdog, he saw
>no reason why the challenger shouldn�t have to vault the same
>obstacle.
>2. Organizers must have a definite budget and solid dates when they
>book a playing hall, which isn�t possible in an open-ended match.
>Kasparov�s argument makes sense, yet Fischer�s point is still valid.
>Only wins should count. This way, a champion can�t cling to the title
>by playing for draws.
I don't know if you remember the late Angus Pitt, but in the
mid-1970s, I had arguments with him about this.
In arguing for x wins, I focused on the possibility of the leader in a
limited-game match playing for draws to run out the clock.
Angus' counter argument, which turned out to be prophetic, was that
between two almost evenly matched very strong players, the person
behind in the match could start playing for draws in hopes of
exhausting his opponent, and the match would never end.
|
| |
Date: 02 Oct 2008 16:13:54
From: Wlodzimierz Holsztynski (Wlod)
Subject: Re: Rules for a World Championship Match
|
On Oct 2, 11:06 am, Quadibloc <[email protected] >
(John Savard) wrote:
> In addition, at the conclusion of game 30, if each player has 15
> points, the outcome of the match will be decided as follows:
>
> If one player has more victories as Black than the other player, the
> player with more victories as Black will win.
This rule doesn't make any sense. I know, there is a common
confusion about this simple issue. Only, if one player played
more as black, and the score is equal, and only then s/he
should be considered a tinge better, microscopically -- in
a match you may have at the most one black game more than
white games, and it is not worth the draw odds.
As I've written in the past, the most important
should be that one wins fiorst of all by pure chess
skills, and not for other reasons. This ideal is hard
to approximate. We don't want matches decided by
clock handling. That's why any electronic hardware
assistance should be welcome. We don't want an accidental
result, hence we want long matches. But then, we don't
want the endurance to be the main factor.
I think that the main change should be the time
limit. It should depend on the depth of the game,
as perceived by the two opponents. For this reasons
the rule should be that none of the players is allowed
to fall behind the other in a game by more than,
say, 90 minutes. On the top of it, the players may
have a reasonable allotment (or perhaps 90 minute
difference limit is just fine).
Now players will spend more time on complex games,
and less on the simple games. They will not forfeit
on time when they are both seriously immersed in the
consecutive positions.
I have proposed this in the past. Perhaps this time
someone will pay attention.
Regards,
Wlod
|
| |
Date: 02 Oct 2008 13:23:07
From: Quadibloc
Subject: Re: Rules for a World Championship Match
|
On Oct 2, 1:19=A0pm, William Hyde <[email protected] > wrote:
> What you suggest consists essentially of the old FIDE rules, with two
> modifications which won't do much harm and may do some good.
> I'm not terribly keen on the six game part of the rules. =A0Applying
> that to the old FIDE matches would truncate several of them. =A0With the
> higher proportion of draws today it might not matter, I suppose.
Well, what I'm trying to do - and to achieve that, "six" will have to
be changed to some other number, perhaps - is to set up rules where
the "first to win N games" is the dominant part of the definition of
what it takes to win, and the win by the older FIDE rule is set for a
longer match - which is why I said 30 games instead of 24 - as simply
a fallback position.
So the idea is that draws won't count... unless not counting draws
would make the match go on forever. If you don't want the net effect
to be that matches will be shorter than formerly, then "six" would be
replaced by a higher number - but the maximum length would have to be
more than 24 games so that the outcome is usually determined by wins
alone, and the parts of the rules that count draws are seldom invoked.
John Savard
|
| |
Date: 02 Oct 2008 12:19:24
From: William Hyde
Subject: Re: Rules for a World Championship Match
|
On Oct 2, 2:06=A0pm, Quadibloc <[email protected] > wrote:
> The question of what would be a good set of rules for a World
> Championship match has been raised here.
[..]
What you suggest consists essentially of the old FIDE rules, with two
modifications which won't do much harm and may do some good.
And since the old FIDE rules produced better matches than anything
that has come along since (IMNSHO, of course) I'd be happy to see
these rules in place. Twenty four games should be the minimum
scheduled match length, though.
I'm not terribly keen on the six game part of the rules. Applying
that to the old FIDE matches would truncate several of them. With the
higher proportion of draws today it might not matter, I suppose.
Mind you, now we need a champion and a challenger, both accepted by
some reasonable fraction of the chess public.
I'm a bit out of touch. Who is/are the real/alleged world/fide
champions/first-among-equals today?
William Hyde
|
| | |
Date: 02 Oct 2008 19:54:34
From: Ed Gaillard
Subject: Re: Rules for a World Championship Match
|
In article <89ff4e5e-bedd-4a03-8f46-9f49d2f7af71@y71g2000hsa.googlegroups.com >,
William Hyde <[email protected] > wrote:
>
>I'm a bit out of touch. Who is/are the real/alleged world/fide
>champions/first-among-equals today?
Anand. I think almost everyone agrees on that. There may be a few
diehards who don't think Kramnik can have lost the title without
losing a match (Anand won the title in a tourrnament in 2007), but I
think even Kramnik is not among them. In any case, Anand is about to
play a title match with Kramnik, at the end of which there should be
no dispute at all.
-ed g.
--
Caissa have mercy on a miserable patzer: http://altergoniff.blogspot.com
|
|
