Comments/Ratings for a Single Item

Gary Gifford  wrote [2008-04-19] 
'Hi Rich... Thanks for the update, I stand corrected.'

No, you do not. In their first match in London in 1834, De Labourdonnais defeated McDonnell, scoring 16 wins, 5 losses and 4 draws. The players alternated colors after each decisive game. De Labourdonnais played with the White pieces in the first four games, because games 1,2,3 were draws.

But in Kieseritzky-Horwitz, London 1846 they alternated colors in the 'modern' way, unaffected by the drawn game (number 11). Final score was 7.5-4.5 The fact that a variety of systems have been used in international tournaments from 1851 to 1866, would not change my opinion on the question before us, even if I knew what that question was.

Bronstein is the original person to propose a time delay between moves,
which is the basis of the 'Bronstein clock'.  Fischer decided to take
the clock and add time, and then to the countdown.

Gary, your approach would be to use a Bronstein clock set to a very short
time delay as a way to do overtime?

A possible solution keeping the way I propose is to use a 3 or 5 second time delay in the tie-breaker should either player have less than 5 minutes on their clock.... something like that. Some experimenting can be done. There can even be a few different options.

As far as a last round thing, most players I know like to know where they stand throughout an event. So settling the score during each round is far more desirable than wrapping up the mystery after all rounds are done.

Given a choice between the 2 types of events... draw or drawless, I still might be inclined to choose the draw one. I think below GM level draws are typically not an issue and players do not seem to mind hard fought out draws. But I certainly would not object to the other (drawless) system.

If a game has say 1 minute vs 1 1/2 minutes left, then I think it is a bit
absurd to have this be the tiebreaker.  How about the same amount of time
is added to both sides, so the player who has the least amount of time
would have a minimum of 5 minutes, while their opponent gets 5 more
minutes added to their clock?

Of course, you could lump all the time from all the ties into a final
match, if it is needed, and give a reasonable amount of time for that.

If I am playing in this system and I see we are heading for a draw, then I better a) offer the draw to start the new game or b) move fast to have enough time for the tie-breaker.

Hey Gary, since this is your own creation, maybe we can go with it, with
some minor modifications :-).  Also, I don't think Bruce created it.  It
looks like the club did.  It looks like he commented on it though.

So, maybe IAGO can use your method, if it would work well, for live
tournaments :-).

I do agree with Bruce's comments that there is often not enough time to
be meaningful.  Perhaps you add 5 minutes back on the clock to what is
left, then 4, then 3, and so on.  Maybe even players can bid, as I
suggested on time to pick sides.  The idea of bidding for time adds more
gamesmanship outside of the actual game, which is used to captivate people
who don't play normally.

http://www.chess.bc.ca/team.html

A quote from that site:' ... tested an anti-draw rule, in which each player had two hours at the start of the first game. If the first game was drawn, each player received 1/3 of a point, then a second game was played with colours reversed, using the remaining time. This continued until one player won, with the winner getting the other 1/3 of a point. '

So, I guess the way I recalled it deviated from reality. I gave someone a full-point, but the method was awarding 1/3 points? I think I prefer my all or nothing way as an anti-draw system.

Excuse me, when writing an old suggestion how to reduce tactical draws
(where material does not force a draw): 

Instead of only accepting or refuting a draw the opponent player should
have the right to simply change the sides (because the suggestor obviously
is convinced that would be no disadvantage). A draw should be possible
only after at least one such a change of sides and a minimum of five moves
following it.

Anyone have a link to Harper's suggestion where he goes into it?

An alternate approach I was laying out involved using a scoring system for
game conditions, like SETS.  This system could be used as a secondary
system for tie-breaking purposes, if need be.

My take is that just about ANY changes, no matter how minor and deviating
from the way it is now, is going to be argued against.  It stands a chance
of being buried with the reason, 'Well, I don't think this issue is THE
problem.'  The end result is that if it could be part of the answer, it
gets buried.  If something faces multiple issues, it is possible that
nothing changes or is fixed, because one issue alone doesn't resolve
anything.

For example, assume the following situation:

(1) players have 60 minutes each / that is a 2 hour game.

(2) at move 40 the game ends in a draw by stalemate

(3) player A used 35 minutes; player B used 40 minutes

(4) The tie-break game is set up and starts with 25 minutes for A and with 20 minutes for B.

(5) That game ends in a draw after 50 moves by perpetual check

(6) player A has 8 minutes left. Player B has 5 minutes left

(7) The crowd gathers around for the fast tie-break game.

(8) Player B gets in trouble and loses on time.

The 1-0 result is obtained within the 2 hours alloted for the round. While other players played 1 game in their 2 hours, some played 2 and some played 3. But 2 hour rounds remained 2 hours and there were no 1/2 - 1/2 results posted.

I do like Bruce's approach also.  However, it is done after EVERY draw,
and not just done if the end result of the entire tournament was a draw. 
What I had suggested is a modification of what they did for the Chess960
tournament last year.  I am suggesting players have a chance to bid time
to see what side they play.  If there is a case of there being one last
game to decide, then this would be the best approach for that,
particularly when dealing with variants where the favored side is
unknown.

There is a secondary pragmatic reason for looking for something else
besides what Bruce Harper wants.  In my attempts to have Seirawan Chess
involved with IAGO, the response was very negative.  The last email with
him, going out at least in peaceful terms, was to wait on everything.  The
end result is that I don't see him wanting the variant community to use
anything he is working on, until he gives the word.  I am assuming his
methods are his.

How about this instead, that borrows a bit from him, but is unique? 
During the match, track how much time was left for each draw.  If there is
a tie score after it is done, add up all the times left and play two games,
one white and one black, with the remaining time that was left. Count these
scores.  If still a tie, then do one last sudden death blitz, as I have
descibed.  In other words, rather than do an overtime after every draw,
just do it at the end, if necessary.

Both players would certainly want to avoid a draw. Of course, if one player had 40 minutes left and another had 5 minutes left, then the one with 40 could certainly play for a draw as he'd have a nice time advantage in the tie-break game. A flat time might be acceptable for correspondence games. For on-line real-time games however, there is often server lag and 6 minutes seems too quick.

I like Harper's approach.  However, it is not likely for awhile to have
anything with Bruce Harper in IAGO, unfortunately.  This being said, can I
run this proposal by people?  It is mean to be a variation on the blitz at
the end of the tournament as a tie breaker.

The final game, in event of the scores being even, would consist of a
blitz game.  The amount of time that both players have at start is six
minutes.  By some method (randomly or otherwise), one player is picked to
enter in a bid on how much time they will give up from their six minutes
to pick their color.  They can choose to pass, and allow their opponent to
pick the side with six minutes, or they can choose, in 15 or 30 second
increments, bid to give up time (to get the choice to pick a side.  From
then on, the time the player picking a side chooses, would keep decreasing
by the bids, until someone passes.  At that point, that player would pick
what side they will play, and have that much time to work with.

Then you do the normal blitz game, with this time adjustment.  This
approach does add a bit of brinksmanship to the game, but is also mainly
used as a balancing mechanism across just about all variants, particularly
those where one side or another may have an advantage, but players are
unable to tell for sure which side does.  This bidding method could also
be used for picking a game also.  Winner of the bid gets that much time,
and the game of their choice, and side of their choice.  Their opponent
gets the other.  You could also bid turns this way with draw rights going
to the opponent of the person bidding.

Draws have to be lesser issue even in regular played-out Mad Queen. It is
GMs' problem not ours. I still personally place Computer problem at the
top. Actually, toward opposite goal of increasing Draw possibilities, Draws could have the same number of variants -- millions -- as CVs themselves. Draw by agreement, 3-fold repetition, 50-Move no pawn
move or capture, insufficient material. Draw by Stalemate, Bare King. Draw by reaching set number of moves: 100, 80, 125, 75. Draw by (and here the
new variants come in, to dodge or postpone other issues, keeping our
comfort zones of wild imagination sans action): (1) time control
parameters, e.g. 4 hours becomes a Draw (2) repetition of opening
repertoire, say 20 moves and over having been played by others before (3)
lack of beauty, i.e. uninteresting lines determined by judge-panel at move 20 (4) Instead of being prohibited, Mures' Braves' point two permits
achieving Draw by placing King a Knight's move away from King. (There follows our regular more-thought-out blog-Comment.)

I like Harper's approach.  However, it might end up being that you might
as well award the win to a player who has the most time on their clock.

Regarding the 3-1-0 flaw see:

http://chessbase.com/newsdetail.asp?newsid=4209

I am again thinking that since we are below the GM level, it might be best to keep the 1-0-1/2 system.

There is a Bruce Harper proposal that I like a lot. If there is a draw, a new game is played using the remaining clock times. If that game is drawn, a new game is played using remaining clock times, etc. Finally someone will win, even if by time default. That is great for over-the-board... but many who like to get the most out of their clock time would likely not like this. It would likely tend to speed chess up so that, in the event of a draw, a player would have some descent time for the next possible game. Harper's system doesn't seem meaningful to correspondence games.

To avoid rehashing stuff, read following, especially the article list at
the end.

http://chessbase.com/newsdetail.asp?newsid=4553

The question is how to incentive players to play for win in all
circumstances? After all, a draw in pursuit of a win is no shame, but to
not try at all is a letdown for all.

For variant creators the question is what kind of rules encourages players
to play for win?

I had discussed the idea to have a blitz game as a tie-breaker with some,
and even more 'Chess purists' were against the idea.  In a tournament
for Chess960, where the current world champ took on someone else, they
used blitz as a tie-breaker.  I am in favor of that, or whatever else
would work, that would resolve the tie at the end.  I am not sure why
anyone is in favor of the current system that is less than 150 years old
personally.  It was adapted when chess had been turned into a competitive
sport in the middle of the 19th century (as documented here), and hadn't
been changed since then.  The current system that produces in the NY
Times: 'Kasparov retains title on a draw' (this was from around 1990),
one has to ask if this is a good thing or not.  

I would also go with the FIFA scoring also of 3-1-0, unless you want to
give draw advantage.  I believe the issue is from a SPORTS perspective,
not the game, even if chess is producing 60%+ draws.

By the way, how about this for a short system of running a tournament? 
Players alternate playing black and white until someone has won a game as
black and a game as white?  Perhaps set it for so many games, and use the
tiebreaker system of blitz to resolve who the winner is.  

I know there are purists who complain about blitz, but one can argue about
whether or not time control distorts chess anyhow.  In order to have chess
as a competitive sport, time control becomes needed.

So, are people here in favor of 3-1-0 for tournament scoring, and the use
of blitz as a tie breaker?  You would use a coin toss to decide which side
plays what.  I am up for this.  I am up for anything people will play that
works, which doesn't produce headlines like: 'Kasparov regains title on
a draw'.  The issue is whether or not such a headline helps or hurts the
growth of a game.  It isn't that it is the answer, or a big part of the
answer, but whether or not changing it would help more than it hurts.  If
this sounds good, maybe the CV community can start to use this for live
tournaments.

I agree that draws aren't a problem. They aren't that frequent at
subgrandmaster level and grandmasters can decide for themselves what suits
them best. But for the record, I would suggest the following solution to
this non-problem. 
In the Yugoslav Soccer Championship, draws led to overtime and penalties,
but the penalties were shot before the overtime. Hence, the players knew
whose interest it was to attack. Following that idea, how about a blitz
with inverted colors giving half a point to the winner and a third of a
point to the loser should the real game end in a draw?

So now I looked at this site:

http://en.wikipedia.org/wiki/Draw_(chess)

which addresses the draw issue. I don't mind #6 under the section entitled: 'Grandmaster draw problem.' Basically, it has been used for soccer (by FIFA) and is this:

'3 points for a win, 1 point for a draw, and 0 points for a loss. This system discourages draws since they would only be worth 2/3 of their current value.' Unlike BAP (mentioned below) there is no color bias.

The BAP system is a bit comical to me as some players have very keen Black Defensive Systems and would rather play Black anyway. That is how I was when I went to the World Open in 1980. I had a keen French Defense and therefore loved playing from the Black side of the board. I had no losses with the French, and just 1 draw with it. In my opinion, the BAP system is biased. Also, such a system would be bad in final rounds of a tournament.

At my level of play draws have yet to be an issue, even in correspondence games where players have several days to think.

If I needed to pick a anti-draw system, FIFA's 3-1-0 seems best to me.

Gary, to address what I wrote early, I found out the first time a draw was
given 1/2 point in a tournament match.  It was in 1867:

http://www.logicalchess.com/info/history/1800-1899.html
1867.09.01 1st time draws count as 1/2 point - Dundee International.

http://www.geocities.com/SiliconValley/Lab/7378/steinitz.htm
In September, 1867 Steinitz took second place in the Dundee International
in England (won by Neumann). This was the first tournament in which draws
were not replayed, but counted as a half a point.


In the past, if a game was drawn, they replayed it.  So, your comment
about, 'playing chess as it had been played for hundreds of years'
isn't exactly true for this invention, which is less than 150 years old. 
And, my take is this invention is wearing out of gas, and producing too
many draws.

In light of this, can you please defend that this is the optimal way chess
should be played, if the invention is less than 150 years old?  Can anyone
here?

The 1/2 - 1/2 for a draw I believe is a mid-18th centuries addition to
chess.  Well, at least the time control is.  So, it isn't like it has
been around a lot.

Anyhow, let me give you a headline here.  Tell me if this makes chess more
or less appealing to people who would consider it:
With a Draw, Kasparov Keeps Title

http://query.nytimes.com/gst/fullpage.html?res=9C0CE3DD1138F934A15751C1A966958260

Why should a defending champion keep their title with a draw?  Gary,
please explain that to me.  And answer me if this is more advantageous to
getting people interested in chess.  I ask it this way, as the issues with
chess aren't simply draw.  But, probably a bunch of major and minor
things.  If it is a list of things, then why not consider ALL things that
may of had issues associated with them.