Guy Macon http://www.guymacon.com/ wrote:
Does anyone have any proposals for reducing the number of draws
in tic-tac-toe?

Don't play?


Dave.

--
David Richerby Revolting Apple (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a tasty fruit but it'll turn your
stomach!

Guy Macon http://www.guymacon.com/ wrote:
In Test Cricket, roughly 30% of matches are drawn. A toss taking
place with no ball being bowled is fairly rare, so the change would
be, I am guessing, 1% or less.

There have been approximately 1850 Tests. My guess is that not more
than one or two of those could have been affected by the rule change.
The toss is not made until conditions are good enough to play: in
particular, if it rains solidly for the whole five days, the toss will
not be made. So, for the toss to be made but no cricket played, the
weather would have to become good enough to look like play would start
and then deteriorate again between the toss and the scheduled start
time, which is about fifteen minutes, I think.


Dave.

--
David Richerby Homicidal Disgusting Cat (TM): it's
www.chiark.greenend.org.uk/~davidr/ like a cuddly pet but it'll turn your
stomach and it wants to kill you!

David Kane wrote:
"David Richerby" wrote:
Soccer is one of the most popular sports on the planet and ties are
common. Cricket is more popular in India alone than most sports are
worldwide: draws[1] are common. Tie-breaks are only used in knock-
out competitions in both sports.

Ties are nowhere near as common in soccer as chess, yet most soccer
leagues use a scoring system which acts as a disincentive. (Win = 3
* tie) The World Cup has never ended with co-champions. Instead it
uses one of the most artificial tiebreaks in sports, PKs, which have
very little connection to play in the field. Why? Because some games
have to have a winner, and the rules of soccer are such that field
play would take a long time to produce it. PKs are barely soccer,
but they are dramatic!

The World Cup Final is the culmination of a knock-out tournament. As
I said, tie-breaks are used in knock-outs. In addition, the penalty
shoot-out in soccer is only used after a period of extra time.

Likewise, chess uses tie-breaks to decide the winners of knock-out
competitions.

The chance of two basketball teams making the same score is low,
just because the scores are so large; rugby and American football
scores are relatively large numbers. Cricket scores are huge
numbers and there are almost no ties (as distinct from draws).
Golf scores are large numbers.

You've picked some poor examples. Almost all of those activities
with "points" as opposed to analog measurements, can and do end in
ties

They can end in ties but they're not very common, for the reasons I
outlined. They're much less common than tied soccer games, for
example.

I'd be curious if you could name a single sport that has gone the
other direction -where rules which produced decisive results were
amended in order to produce more "well-played" ties?

I don't see the relevance of this. Nobody is proposing to change the
rules of chess to increase the number of draws.

And chess did make rule changes that increased draws! That's the
point. We need to recognize that fact and deal with it.

Which rule changes would those be?


Dave.

--
David Richerby Technicolor Love Toy (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a fun child's toy that you can share
with someone special but it's in
realistic colour!

zdrakec wrote:
Actually, of all the suggestions I offered, the "stalemate=loss" was
the one I thought the most reasonable. [...] it seems to me that the
player who has stalemated his opponent has satisfied the basic
object of the game: to place his opponent in a situation from which
his king cannot avoid being captured.

You misunderstand. The goal of chess is to checkmate the enemy king.
The fact that stalemate is a draw proves that the goal of the game is
not to force the capture the enemy king.


Dave.

--
David Richerby Swiss Monk (TM): it's like a man of
www.chiark.greenend.org.uk/~davidr/ God but it's made in Switzerland!

On Dec 8, 6:39 am, "M Winther" wrote:
Den 2007-12-08 08:01:42 skrev David Kane :







"M Winther" wrote in messagenewspt2z7c9hb3bzrao@kalroten...
Den 2007-12-07 07:10:55 skrev David Kane :

"zdrakec" wrote in message
...
Hullo all:

Giving some thought to how draws happen over the board, it occurs to
me that many of the rules in place that allow draws to happen are
quite arbitrary, and have little or nothing to do with the actual
movement of the pieces. With that in mind, the following rule changes
strike me as perhaps interesting:
1. Draws may not be agreed.
2. Three-fold repetition shall be illegal. That is, the player having
the move may not make a move that repeats the position for the third
time. If he has no other legal move, we may regard this as a form of
stalemate, perhaps. I am not completely sure, but I think that this
rule would do away with perpetual check. By analogy, I refer to the ko
rule in Go, which essentially prevents the same position from
appearing on the board more than once (with some rather esoteric
exceptions).
3. There shall be no 50-move limit. This becomes more practical than
it once was, in an age where clocks can have a delay, and can add a
time increment per move.
4. Since the object of the game is to place the opponent's king in a
position from which it cannot avoid being captured, I suggest that
stalemate should be a loss for the stalemated player. In principal,
the player who has stalemated his opponent has accomplished the
primary goal of the game. Of course, this has been suggested many
times before.

I am not certain, but that would seem to leave lack of mating material
as the sole way to draw a game. Naturally, there can still be abuse,
but it would, I think, tend to be obvious...

Actually, if you disallow the current drawing methods, make repetition
illegal, and define an inability to make a legal move as losing stalemate,
then there would never by *any* draws! (Even K vs. K would be
decisive because positions would eventually repeat.) Of course it wouldn't be
chess exactly since endgame theory would change dramatically, and
it wouldn't objectively be a good game (because it could take quite
a long time to exhaust all possible board positions), but it strikes me
as a very interesting intellectual exercize.

.

You can't be quite serious. An alternative is to change the properties
of the pawn a little, as in "Improved Chess"
http://hem.passagen.se/melki9/improvedchess.htm
or Reformed Chess
http://hem.passagen.se/melki9/reformedchess.htm
or perhaps change the properties of the Rook a little, thus to increase
winning chances
http://hem.passagen.se/melki9/warlockchess.htm

My view is that making statelemate a win and (possibly) disallowing
repetition would be an objectively superior game to chess. But I do not
favor those changes. Chess has a lot of history behind it and changing
the fundamental rules of the game should not be done
lightly. On the other hand, trying different tournament
formats, scoring, etc., that is changing the chess meta-game, *should*
be experimented with - frequently and radically.

My point is that the the referenced variants do *not* change the
fundamental rules of chess much. Strategy and tactics remain largely
the same. A pawn has a little extra capability if (1) it has passed
the middle line and (2) is blocked. This is not such a huge difference
in termes of rules, but it makes a huge difference on the score table.
Draws will be much fewer will endgames will be much less drawish.

A big problem about orthochess is the drawishness of endgames
"All rook endings are drawn", said Tartakower(?). It is not so in
Chinese Chess and Korean Chess. Addressing the drawishness
of orthochess implies reducing drawishness of the endgame.

Mats- Hide quoted text -

- Show quoted text -

Actually, you can change how the pieces move all you want...it still
would not change the result of a game the players have decided will be
a draw after, say, 20 moves or so....so long as draws may be
agreed....

Regards,
zdrakec

David Kane wrote:

Guy Macon wrote...

We do know that slight differences in search depth and pruning
make a huge difference in computer vs. computer games, and far
less difference in computer vs. human games.

What is your source?

A while back I asked whether the above was true (it seems
reasonable that one program seeing just a bit deeper would
be able to avoid various bad situations) and Dr. Hyatt said
that his experience was that it is true. It should be fairly
easy to find in the archives.

That alone is enough to explain why computer vs. computer
matches have fewer draws.

Not even close. You might be able to put together an argument that
it contributes in some way, but you have yet to do so.

Imagine factor X, that has the attribute of multiplying any
difference in skill between players with beards, but not
between bearded and shaven players.

It is trivial to show from the statistics that matches between
very strong and very weak players end up in draws less often
than matches between evenly matched players.

All other things being equal, if factor X exists, it will
reduce the comparative number of draws among Hassidic Jews,
Amish, and Muslims, and will increase the comparative number
of draws among Mormons and active military personel.


--
Guy Macon
http://www.guymacon.com/

David Richerby wrote:

Guy Macon http://www.guymacon.com/ wrote:

In Test Cricket, roughly 30% of matches are drawn. A toss taking
place with no ball being bowled is fairly rare, so the change would
be, I am guessing, 1% or less.

There have been approximately 1850 Tests. My guess is that not more
than one or two of those could have been affected by the rule change.
The toss is not made until conditions are good enough to play: in
particular, if it rains solidly for the whole five days, the toss will
not be made. So, for the toss to be made but no cricket played, the
weather would have to become good enough to look like play would start
and then deteriorate again between the toss and the scheduled start
time, which is about fifteen minutes, I think.

Good point. Looks like the difference is 0.1% or less.

--
Guy Macon
http://www.guymacon.com/

"Guy Macon" http://www.guymacon.com/ wrote in message
...



David Kane wrote:

Guy Macon wrote...

We do know that slight differences in search depth and pruning
make a huge difference in computer vs. computer games, and far
less difference in computer vs. human games.

What is your source?

A while back I asked whether the above was true (it seems
reasonable that one program seeing just a bit deeper would
be able to avoid various bad situations) and Dr. Hyatt said
that his experience was that it is true. It should be fairly
easy to find in the archives.

Hardly very convincing, this.

That alone is enough to explain why computer vs. computer
matches have fewer draws.

Not even close. You might be able to put together an argument that
it contributes in some way, but you have yet to do so.

Imagine factor X, that has the attribute of multiplying any
difference in skill between players with beards, but not
between bearded and shaven players.

It is trivial to show from the statistics that matches between
very strong and very weak players end up in draws less often
than matches between evenly matched players.

All other things being equal, if factor X exists, it will
reduce the comparative number of draws among Hassidic Jews,
Amish, and Muslims, and will increase the comparative number
of draws among Mormons and active military personel.

Still not there. The draw rate between programs *of the same strength*
is less than between humans of the same strength, even at ratings
higher than those of humans.

"Guy Macon" http://www.guymacon.com/ wrote in message
...



David Richerby wrote:

Guy Macon http://www.guymacon.com/ wrote:

In Test Cricket, roughly 30% of matches are drawn. A toss taking
place with no ball being bowled is fairly rare, so the change would
be, I am guessing, 1% or less.

There have been approximately 1850 Tests. My guess is that not more
than one or two of those could have been affected by the rule change.
The toss is not made until conditions are good enough to play: in
particular, if it rains solidly for the whole five days, the toss will
not be made. So, for the toss to be made but no cricket played, the
weather would have to become good enough to look like play would start
and then deteriorate again between the toss and the scheduled start
time, which is about fifteen minutes, I think.

Good point. Looks like the difference is 0.1% or less.


Or to put it another way, we still don't have an example
of rule changes made with a purpose of increasing the
number of ties.

Why do you suppose that is?

"David Richerby" wrote in message
...
David Kane wrote:
"David Richerby" wrote:
Soccer is one of the most popular sports on the planet and ties are
common. Cricket is more popular in India alone than most sports are
worldwide: draws[1] are common. Tie-breaks are only used in knock-
out competitions in both sports.

Ties are nowhere near as common in soccer as chess, yet most soccer
leagues use a scoring system which acts as a disincentive. (Win = 3
* tie) The World Cup has never ended with co-champions. Instead it
uses one of the most artificial tiebreaks in sports, PKs, which have
very little connection to play in the field. Why? Because some games
have to have a winner, and the rules of soccer are such that field
play would take a long time to produce it. PKs are barely soccer,
but they are dramatic!

The World Cup Final is the culmination of a knock-out tournament. As
I said, tie-breaks are used in knock-outs. In addition, the penalty
shoot-out in soccer is only used after a period of extra time.

But you haven't answered why? Why doesn't soccer just
declare World Cup co-champions? The shoot out may be dramatic,
but it's little more than a mockery of the sport. Why don't those fans
just appreciate those ties?

Likewise, chess uses tie-breaks to decide the winners of knock-out
competitions.

The chance of two basketball teams making the same score is low,
just because the scores are so large; rugby and American football
scores are relatively large numbers. Cricket scores are huge
numbers and there are almost no ties (as distinct from draws).
Golf scores are large numbers.

You've picked some poor examples. Almost all of those activities
with "points" as opposed to analog measurements, can and do end in
ties

They can end in ties but they're not very common, for the reasons I
outlined. They're much less common than tied soccer games, for
example.

And even though fairly rare, those sports*still* go to the bother of breaking
those ties, with overtimes, playoff rounds etc. Why?



I'd be curious if you could name a single sport that has gone the
other direction -where rules which produced decisive results were
amended in order to produce more "well-played" ties?

I don't see the relevance of this. Nobody is proposing to change the
rules of chess to increase the number of draws.

And chess did make rule changes that increased draws! That's the
point. We need to recognize that fact and deal with it.

Which rule changes would those be?

In 1867, chess introduced the practice of counting draws as half a win.
At the time, I doubt anyone would have predicted that that change would lead
to the absurd drawfests that plague today's high-end game. But they have.