David Kane wrote:
"Harold Buck" wrote in message
...
In article ,
"David Kane" wrote:


Your claim that endgame theory is "simplified" is
unproven. It *would* be radically changed. What is the
outcome of KB or KN vs k in the event that k wins the
50 move tiebreak? Are there symmetrical material
situations (KXY vs kxy) in which one side can force the other to
capture first? I'll concede that it's *possible* that endgame
theory would be simpler, but there haven't been any coherent
arguments to that effect yet. To argue that a a specfic
type of position is harder in one set of rules than another
proves absolutely nothing.

If you really have solved all of endgame theory
in the proposed game and figured out how it will
influence the middle game, by all means I'd be
interested to hear a detailed assessment. It's a
fascinating intellectual exercize if nothing else.


You are correct that there is no proof that there would not be
interesting endgames under the new rules. But the fact that entire rich
classes of endgames are reduced to "win" or "loss' with no subsequent
play is telling. For example, there is suddenly no strategy at all to
K+P vs. K. It *seems* unlikely that enough "new" endgame stuff would
arise in the new rules to offset all of the strategy that was rendered
useless by the rule change. Again, no proof, but take an endgame book
and look at how many pages you could rip out under the new rules, and
that's what the new rules would have to replace just to stay even with
the richness of current endgame theory.

Another example, K vs K is a non-position in chess
but a forced "mate" in no-draw chess. Of course,
forcing "mate" may take 1000's of moves, which means that
if we want to maximize the richness of endgame theory
we would have to abolish the 50-move rule.

So if it's theory you want, no-draw chess
can supply massive amounts of it.

David,

I understand what you're getting at, but I still find it hard to
believe that the modified chess described by the OP strikes you as
anything remotely like a well-designed game. The obvious problem with
it is the radical discontinuity between the goals you're trying to
achieve in a winnable game and the goals the rules force you to pursue
if the game is not winnable, in the traditional sense.

Recall that the whole point of this, presumably, was to make chess more
exciting to the general public. But now you're picturing a situation
where in an endgame we'd now consider a dead draw, the players shuffle
around their pieces for hundreds of moves, each trying to get the other
to repeat a position. A thrilling prospect!

Note that one of the skills needed for this final phase of the game
(which we might call "the legal phase") is meticulous accounting. It
could be, as you suggest, that rook versus bishop will turn out to be a
theoretical win in several hundred moves. Still, rather than resign, it
might be worth playing out to the bitter end for the losing side,
because who knows, the player with the rook might screw up and
inadvertently repeat, turning a win into a loss...

Larry T.

"Larry Tapper" wrote in message
ups.com...

David Kane wrote:
"Harold Buck" wrote in message
...
In article ,
"David Kane" wrote:


Your claim that endgame theory is "simplified" is
unproven. It *would* be radically changed. What is the
outcome of KB or KN vs k in the event that k wins the
50 move tiebreak? Are there symmetrical material
situations (KXY vs kxy) in which one side can force the other to
capture first? I'll concede that it's *possible* that endgame
theory would be simpler, but there haven't been any coherent
arguments to that effect yet. To argue that a a specfic
type of position is harder in one set of rules than another
proves absolutely nothing.

If you really have solved all of endgame theory
in the proposed game and figured out how it will
influence the middle game, by all means I'd be
interested to hear a detailed assessment. It's a
fascinating intellectual exercize if nothing else.


You are correct that there is no proof that there would not be
interesting endgames under the new rules. But the fact that entire rich
classes of endgames are reduced to "win" or "loss' with no subsequent
play is telling. For example, there is suddenly no strategy at all to
K+P vs. K. It *seems* unlikely that enough "new" endgame stuff would
arise in the new rules to offset all of the strategy that was rendered
useless by the rule change. Again, no proof, but take an endgame book
and look at how many pages you could rip out under the new rules, and
that's what the new rules would have to replace just to stay even with
the richness of current endgame theory.

Another example, K vs K is a non-position in chess
but a forced "mate" in no-draw chess. Of course,
forcing "mate" may take 1000's of moves, which means that
if we want to maximize the richness of endgame theory
we would have to abolish the 50-move rule.

So if it's theory you want, no-draw chess
can supply massive amounts of it.

David,

I understand what you're getting at, but I still find it hard to
believe that the modified chess described by the OP strikes you as
anything remotely like a well-designed game. The obvious problem with
it is the radical discontinuity between the goals you're trying to
achieve in a winnable game and the goals the rules force you to pursue
if the game is not winnable, in the traditional sense.

The part that rubs me the wrong way is the
50-move tiebreak. That strikes me as arbitrary and
"unchesslike". In another post, I suggested different tiebreak
possibilities, as well as leaving that a draw. I do find the "no
repetitions" and "stalemate = loss" intrinsically appealling
and perfectly consistent with chess' spirit.

Obviously, the "mathematical" version of the rules
that I proposed (no 50-move rule) is not a practical
game, but I think it is interesting to discuss. (I''m waiting
for someone to generate the K vs K tablebase)


Recall that the whole point of this, presumably, was to make chess more
exciting to the general public. But now you're picturing a situation
where in an endgame we'd now consider a dead draw, the players shuffle
around their pieces for hundreds of moves, each trying to get the other
to repeat a position. A thrilling prospect!

The poster did not state his motivations. I took
it as an interesting thought experiment.

I consider myself a traditionalist and would not favor
any changes to the basic rules of chess. Arbitrary as chess'
rules are, there is nothing wrong with having a preference for
something arbitrary. I don't feel ashamed that I
like chess even though other games are objectively
superior by certain measures.

However, the "anti" arguments have ranged from the
utterly moronic "Chess becomes roulette" to the almost
certainly false and definitely unproven
"endgames become trivial". Bad and poorly thought
out arguments should not be given a pass just
because they seek to preserve the status quo.
They should be evaluated on their merits or lack
thereof.

Note that one of the skills needed for this final phase of the game
(which we might call "the legal phase") is meticulous accounting. It
could be, as you suggest, that rook versus bishop will turn out to be a
theoretical win in several hundred moves. Still, rather than resign, it
might be worth playing out to the bitter end for the losing side,
because who knows, the player with the rook might screw up and
inadvertently repeat, turning a win into a loss...

That would depend on whether the rules would allow
him to repeat. In analogy with checkmate, a player
wouldn't be able to repeat. If a player found himself
in a position where he was *forced* to repeat, he would
lose. It would definitely give new meaning to the
phrase "playing out the ending"

On Fri, 08 Dec 2006 20:03:13 GMT, Ron
wrote:

More to the point, an ending like K+B v K is a win for different players
depending on who made the last capture. EG, if the side with a king just
captured an enemy pawn, he wins.

And similarly (which I pointed out before), if you get to a K+Q vs. K,
you don't have to know how to checkmate if you made the last capture!
You win anyway.
--
Replace you know what by j to email

David Kane wrote:
wrote in message
2) Broken Rule: You will have situations where one move, from position
A, resulted in position X, and a couple moves later a *different* move,
from position B (very slightly different from A), resulted in the same
position X. Both moves, from their respective positions, were
objectively the best move (this can certainly happen). Therefore, you
are not allowing a player to play the *best* move in a position.

This objection doesn't make much sense. It's arbitrary as to whether
allow repetitions. e.g. Go generally disallows them (though the details are
complicated). What it does is turn easy draws into complicated decisive
games. e.g. consider a White pawn on c4 and Black pawn on c5. In chess
these positions are draws unless one King is far away. But in no-draw
chess, it all comes down to the King placement. It's not chess, but it's
a very interesting game of skill that looks alot like chess.

The rule is broken because of the forced nature of it. Why should I be
forced to make a sub-optimal move, because of an 'arbitrary' rule? Is
not chess the pursuit of perfection?


3) Why? If I am black and I make a capture of the last pawn resulting
in an endgame of K+R vs. K+B, with me having the Bishop, I can *win*
the game by hanging on for 50 moves. That makes no sense.

So the question becomes whether you misplayed the position when pawns were
on the board? Could you have used your Rook to ensure that the last
capture was yours?

Probably not, as I stated I had the bishop and captured the final pawn
on the board. No matter, you are penalizing one side or the other
unfairly for the sake of avoiding the most likely outcome. It simply
does not make any sense.


Chess is not broken. What is broken is allowing GM draws. Restrict this
practice and you have interesting, fighting chess again, via the
Sofia-like rules.

Logically the Sofia rules are seriously flawed because they
do nothing to address the fact that playing for a draw is
an optimal strategy in a high-percentage of GM situations.

Or, reworded, a draw is the most likely outcome in most GM situations.

So instead of playing optimally, players have to
"pretend" they are playing sub-optimally.

So, we decide to penalize them instead and force them to play
suboptimally with a set of arbitrary rules...

It would be kind of like
instituting a requirement in the Masters that you have to hit at
least 3 balls into the woods every round. Viewers love to watch
golfers in trouble. But if golf were to institute such an idiotic
policy, the good golfers would, of course, get their requirement
out of the way by aiming for the easiest woods on the course.

That analogy is so far off...I don't even see the connection at all.

Like I said, "sofia-like". Not saying they are perfect - they are not -
but they do attempt to address some of the problems.

Look at it this way: When you and I play (joe class players all of us),
we usually play to the end. At *our* level, the game is not broken
because of the draw characteristics or rules. The fact is, GM's will
and do take advantage of the sponsors and play these short draws to
conserve energy, et., or for whatever reason. We simply need to remove
those reasons or put into force a framework to disallow that type of
draw, WITHOUT CHANGING THE UNDERLYING GAME ITSELF.

These *ludicrous* draw rules do not accomplish that.

On 8 Dec 2006 13:48:54 -0800, "
wrote:

Like I said, "sofia-like". Not saying they are perfect - they are not -
but they do attempt to address some of the problems.

From 1962 to 1964 FIDE had a rule against no draws by agreement before
move 30, but it didn't work.
--
Replace you know what by j to email

In article .com,
" wrote:

So, we decide to penalize them instead and force them to play
suboptimally with a set of arbitrary rules...

Something similar came up in a soccer newsgroup not too long ago - the
idea that you could reduce draws by, after a certain amount of time,
giving the win to the team who'd won the most corner kicks.

The logic of this was that an attacking team tends to get more corners,
so you're rewarding attacking play. But, of course, it fails for the
same reasons these changes to chess do:

It changes the objective. At a certain point, soccer teams would stop
trying to score and start trying to win corners.

Similarly, by removing checkmate as the ultimate objective of a chess
game, these changes would have drastic and, I suspect, severely negative
impacts on the way the game is played.

-Ron

On Sat, 09 Dec 2006 01:04:32 GMT, Ron
wrote:

Something similar came up in a soccer newsgroup not too long ago - the
idea that you could reduce draws by, after a certain amount of time,
giving the win to the team who'd won the most corner kicks.

There is (or was) something similar in American high school football.
If a decisive result was needed, and the score was tied, the team that
achieved the most first downs was the winner. If that was tied, I
think it went to something like total yardage gained or something like
that.
--
Replace you know what by j to email

Ron wrote:
In article .com,
" wrote:

So, we decide to penalize them instead and force them to play
suboptimally with a set of arbitrary rules...

Something similar came up in a soccer newsgroup not too long ago - the
idea that you could reduce draws by, after a certain amount of time,
giving the win to the team who'd won the most corner kicks.

If memory serves, at one time that was the rule in collegiate soccer.


The logic of this was that an attacking team tends to get more corners,
so you're rewarding attacking play. But, of course, it fails for the
same reasons these changes to chess do:

It changes the objective. At a certain point, soccer teams would stop
trying to score and start trying to win corners.

Not really as bad as it seems - forcing corners is an excellent way to
attempt to score.

Also, by creating an incentive for the defense to do something *other*
than give up a corner, you open up play just a bit, and increase the
probability of a goal.


Similarly, by removing checkmate as the ultimate objective of a chess
game, these changes would have drastic and, I suspect, severely negative
impacts on the way the game is played.

But, to be clear...for *chess* I agree. Moving the goalposts
significantly changes the goal.


-Ron


--
Kenneth Sloan
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/sloan/

In article
,
Ron wrote:


3) ...Why? If I am black and I make a capture of the last pawn resulting
in an endgame of K+R vs. K+B, with me having the Bishop, I can *win*
the game by hanging on for 50 moves. That makes no sense.

More to the point, an ending like K+B v K is a win for different players
depending on who made the last capture. EG, if the side with a king just
captured an enemy pawn, he wins. But if the last move was the side with
the bishop capturing an enemy pawn with his bishop, then he wins.


Or if the last move was 1. exd8=B!

--Harold Buck


"Hubris always wins in the end. The Greeks taught us that."

-Homer J. Simpson

In article ,
Jud McCranie wrote:

On 8 Dec 2006 13:48:54 -0800, "
wrote:

Like I said, "sofia-like". Not saying they are perfect - they are not -
but they do attempt to address some of the problems.

From 1962 to 1964 FIDE had a rule against no draws by agreement before
move 30, but it didn't work.


Because people will flout the rule and agree to a draw at move 29? :-)

--Harold Buck


"Hubris always wins in the end. The Greeks taught us that."

-Homer J. Simpson