"CeeBee" wrote in message
.6.84...
"David Kane" wrote in rec.games.chess.misc:



The issue is a software tool using available
information to assess playing strength.

No, this is not the issue. The issue is the way to reach a correct
assessment of chess strength, disconnected from actual play.


The moves could be drawn from the exact games which are used to determine
the
rating based upon performance only. It's not at all disconnected from actual
play.
It uses all of the information about the actual play, not just the result.


Many programs do have a rating function. The Fritz GUI rates you on basis
of results against a "rated" Fritz opponent.


But this is inefficient compared to the proposed alternative because you
are throwing away all of the information about the game except the result.



If results of a chess game correlate with the quality of
the moves played in the game, then analysis of the moves can be used
to predict results. For a fixed number of games, this will be more
accurate than a prediction based on results alone.

It's a useful _bypass_. But this is not what's the discussion is about.
The discussion is not about the useability of such a bypass, but about the
question if such a system is _better_ than the current system, which isn't
the case.


You haven't defined what you mean by better. I would say that the rating
that
has the best predictive value of future results is better. Suppose a player
plays in
a tournament.

A. He achieves a perfomance rating for the event. This means that there is a
probability
that his true strength lies within a range (which can be very large if the
number of games
in the tournament is small).
B. Now imagine that the moves of the games are analyzed and compared to
some standard, so that a profile of his move quality is determined. This
would
be compared to information for other players and a range analogous to that
in A is developed.

My bet is that B could produce a much smaller range and therefore predict
future results better.

Bottom line of the problem (or misunderstanding) is that chess is a
competitive sport. The opponent is not one of the many factors, she or
he's the determining factor. And as long as this is not the basis of the
calculation, every result of another calculation will always be a
substitute for the rating you get based on play against opponents, like in
the Elo-rating system.

The only way that analyzing the strength of moves would not work is if there
is something other than the quality of the moves a player makes that
determines his results on average.
This is not impossible in principle (e.g. a player could have such bad
hygiene that his opponents can't
concentrate when they play him) but it seems far-fetched to me.

The existing "performance only" rating is irrelevant for most players
because they
simply don't play rated games or don't play enough of them to be
statistically significant. Consider
scholastic players, whose true strength changes quickly but play a half
dozen rated games once or twice a year.

"CeeBee" wrote in message
. 6.84...
"David Kane" wrote in rec.games.chess.misc:

But anyone who
plays chess better than
we do and looked at the moves of the games would easily see which of
us was stronger. For example, he could just count the frequency of
errors at 1-ply or 2-ply.

That is true, yet it is not relevant for the current discussion.

I did't doubt that one is able to see strength differences in actual play,
I tried to explain that looking at actual moves is _not a better way_ to
determine _absolute strength_.


I don't think that anyone was proposing eliminating the result-based
rating system.

Again: it wasn't discussed. The original poster suggested adding a
_better_ rating system with "more absolute values". With that he makes two
mistakes: it suggests incorrectly that the current system gives
"inaccurate absolute values". The current Elo system might give
inaccurate values, but _no_ absolute values whatsoever; secondly he
supposes that the validation of individual moves can give a useful
official rating to even players who have never played a game before, which
is also not true, because it might tell you something (very limited) about
chess understanding, but doesn't tell you anything about _absolute_ chess
strenght in actual play.



This is very true. I have known some very strong casual chess players - even
2000
strength - who fold up completely with serious / tournament chess due to
nerves and other factors. There are those who cannot handle a loss in a
tournament
and go to pieces. Time management is also a key part of chess skills.
Then there is the competitive nature of the person involved.

There is a lot more to chess strength then the strength of individual moves.

Regards

snip

"David Kane" wrote in message
news:joJib.554559$Oz4.504403@rwcrnsc54...

"CeeBee" wrote in message
.6.84...
"David Kane" wrote in rec.games.chess.misc:



The issue is a software tool using available
information to assess playing strength.

No, this is not the issue. The issue is the way to reach a correct
assessment of chess strength, disconnected from actual play.


The moves could be drawn from the exact games which are used to determine
the
rating based upon performance only. It's not at all disconnected from
actual
play.
It uses all of the information about the actual play, not just the result.


Many programs do have a rating function. The Fritz GUI rates you on
basis
of results against a "rated" Fritz opponent.


But this is inefficient compared to the proposed alternative because you
are throwing away all of the information about the game except the result.



If results of a chess game correlate with the quality of
the moves played in the game, then analysis of the moves can be used
to predict results. For a fixed number of games, this will be more
accurate than a prediction based on results alone.

It's a useful _bypass_. But this is not what's the discussion is about.
The discussion is not about the useability of such a bypass, but about
the
question if such a system is _better_ than the current system, which
isn't
the case.


You haven't defined what you mean by better. I would say that the rating
that
has the best predictive value of future results is better. Suppose a
player
plays in
a tournament.

A. He achieves a perfomance rating for the event. This means that there is
a
probability
that his true strength lies within a range (which can be very large if the
number of games
in the tournament is small).
B. Now imagine that the moves of the games are analyzed and compared to
some standard, so that a profile of his move quality is determined. This
would
be compared to information for other players and a range analogous to that
in A is developed.

My bet is that B could produce a much smaller range and therefore predict
future results better.

Bottom line of the problem (or misunderstanding) is that chess is a
competitive sport. The opponent is not one of the many factors, she or
he's the determining factor. And as long as this is not the basis of the
calculation, every result of another calculation will always be a
substitute for the rating you get based on play against opponents, like
in
the Elo-rating system.

The only way that analyzing the strength of moves would not work is if
there
is something other than the quality of the moves a player makes that
determines his results on average.
This is not impossible in principle (e.g. a player could have such bad
hygiene that his opponents can't
concentrate when they play him) but it seems far-fetched to me.

The existing "performance only" rating is irrelevant for most players
because they
simply don't play rated games or don't play enough of them to be
statistically significant. Consider
scholastic players, whose true strength changes quickly but play a half
dozen rated games once or twice a year.



I really have no idea where this 'standard' of moves is going to come from.

Computers are better then all humans at simple tactics but lag considerably
behind on positional , long range planning and endgames.

A comp evaluation of 0.6 for move A does not mean it is better then 0.2 for
move B.

A computer could give an idea of tactical strength by analysing games but in
no
way could this be translated in to an ELO.

All this of course ignores the competitive side chess. There is much more to
chess then good moves.

How are you going to rate player A who makes better moves then B but then
loses
on time or becomes nervous at the thought of winning the tournament.


Regards

"Terry" wrote in message
...

Again: it wasn't discussed. The original poster suggested adding a
_better_ rating system with "more absolute values". With that he makes
two
mistakes: it suggests incorrectly that the current system gives
"inaccurate absolute values". The current Elo system might give
inaccurate values, but _no_ absolute values whatsoever; secondly he
supposes that the validation of individual moves can give a useful
official rating to even players who have never played a game before,
which
is also not true, because it might tell you something (very limited)
about
chess understanding, but doesn't tell you anything about _absolute_
chess
strenght in actual play.



This is very true. I have known some very strong casual chess players -
even
2000
strength - who fold up completely with serious / tournament chess due to
nerves and other factors. There are those who cannot handle a loss in a
tournament
and go to pieces. Time management is also a key part of chess skills.
Then there is the competitive nature of the person involved.

You seem to be saying that these factors affect the results of games upon
which the ELO is based (I agree) but that they mysteriously do so *without*
affecting the moves played in the game.

There is a lot more to chess strength then the strength of individual
moves.

Such as? What are the significant factors that affect results but not moves?

Regards

snip

"David Kane" wrote in rec.games.chess.misc:


You haven't defined what you mean by better.


I haven't defined anything. The original poster has. I have responded to the
original poster who stated he wanted to improve the current rating system.
Please read all the messages to get an indication of what the discussion is
about.



--
CeeBee


Uxbridge: "By God, sir, I've lost my leg!"
Wellington: "By God, sir, so you have!"


Google CeeBee @ www.geocities.com/ceebee_2

what I think will happen in the not too distant futu

many different rating systems based on the moves will be developed.
Each game will be given a rating and a weighting number
for both players.There will be white-ratings and black-ratings.
Several chess-software companies, chess journals, chess-websites
will have their own rating system and these will be competing with
each other.
It _is_ possible to rate a rating system by checking how reliable
the predicted outcome of future games is. See e.g. Sonas' article
at chessbase.de

The old ELO-rating system will be maintained for some time and
for historical reasons, but finally they will adapt one
of the other systems, since the advantages become too apparant.

Other sports, like tennis,boxing, etc. will follow.
In tennis e.g. the advantage to rate each single point is easy
to see and easy to implement.

Tournaments will be played and rated, where the players don't play
against each other, but get chesspositions from the computer
and then have to find a good move in that position.


I don't understand CeeBee's "_absolute_ rating-" argument,
why move-rating can't work. Does anyone here understand it ?


Guenter

"Sterten" wrote in message
...


what I think will happen in the not too distant futu

many different rating systems based on the moves will be developed.
Each game will be given a rating and a weighting number
for both players.There will be white-ratings and black-ratings.
Several chess-software companies, chess journals, chess-websites
will have their own rating system and these will be competing with
each other.
It _is_ possible to rate a rating system by checking how reliable
the predicted outcome of future games is. See e.g. Sonas' article
at chessbase.de

The old ELO-rating system will be maintained for some time and
for historical reasons, but finally they will adapt one
of the other systems, since the advantages become too apparant.

Other sports, like tennis,boxing, etc. will follow.
In tennis e.g. the advantage to rate each single point is easy
to see and easy to implement.

Tournaments will be played and rated, where the players don't play
against each other, but get chesspositions from the computer
and then have to find a good move in that position.

These already exist and are called problem solving tournies.
You are missing the competitive nature of chess. I and most
chess players want to play a game of chess against
other chessplayers. I fully understand CeeBee's point of view on
absolute rating, but feel he has fully explained it.

I still do not understand who / what is going to rate individual moves.
Computers are not up to it and are unlikely to be in forseeable future.

Do you actually play chess and have you ever played in a chess
tournament ?. You show a lack of understandind of chess.

Regards

"Sterten" wrote in message
...


what I think will happen in the not too distant futu

many different rating systems based on the moves will be developed.
Each game will be given a rating and a weighting number
for both players.There will be white-ratings and black-ratings.
Several chess-software companies, chess journals, chess-websites
will have their own rating system and these will be competing with
each other.
It _is_ possible to rate a rating system by checking how reliable
the predicted outcome of future games is. See e.g. Sonas' article
at chessbase.de

The old ELO-rating system will be maintained for some time and
for historical reasons, but finally they will adapt one
of the other systems, since the advantages become too apparant.

Other sports, like tennis,boxing, etc. will follow.
In tennis e.g. the advantage to rate each single point is easy
to see and easy to implement.

Tournaments will be played and rated, where the players don't play
against each other, but get chesspositions from the computer
and then have to find a good move in that position.

These already exist and are called problem solving tournies.

they should also be used on grandmaster level and co-used for
their rating.

You are missing the competitive nature of chess. I and most
chess players want to play a game of chess against
other chessplayers.

you still can do it. But other forms of chess-competition
should get a chance too.

I fully understand CeeBee's point of view on
absolute rating, but feel he has fully explained it.

and you agree to his conclusion that any sort of
move-rating-system is doomed to fail ?

I still do not understand who / what is going to rate individual moves.
Computers are not up to it and are unlikely to be in forseeable future.

they already do give position evaluations of chesspositions
in "pawn units", you can watch this during the game on most standard
chess programs

Do you actually play chess and have you ever played in a chess
tournament ?. You show a lack of understandind of chess.

no / yes / sorry if I did-that was not intended
http://chessmetrics.com/PL/PL38915.htm

Guenter Stertenbrink

"Sterten" wrote in message
...
"Sterten" wrote in message
...


what I think will happen in the not too distant futu

many different rating systems based on the moves will be developed.
Each game will be given a rating and a weighting number
for both players.There will be white-ratings and black-ratings.
Several chess-software companies, chess journals, chess-websites
will have their own rating system and these will be competing with
each other.
It _is_ possible to rate a rating system by checking how reliable
the predicted outcome of future games is. See e.g. Sonas' article
at chessbase.de

The old ELO-rating system will be maintained for some time and
for historical reasons, but finally they will adapt one
of the other systems, since the advantages become too apparant.

Other sports, like tennis,boxing, etc. will follow.
In tennis e.g. the advantage to rate each single point is easy
to see and easy to implement.

Tournaments will be played and rated, where the players don't play
against each other, but get chesspositions from the computer
and then have to find a good move in that position.

These already exist and are called problem solving tournies.

they should also be used on grandmaster level and co-used for
their rating.

You are missing the competitive nature of chess. I and most
chess players want to play a game of chess against
other chessplayers.

you still can do it. But other forms of chess-competition
should get a chance too.

I fully understand CeeBee's point of view on
absolute rating, but feel he has fully explained it.

and you agree to his conclusion that any sort of
move-rating-system is doomed to fail ?

I cannot see that a move-rating-system will work but keep
an open mind.

I still do not understand who / what is going to rate individual moves.
Computers are not up to it and are unlikely to be in forseeable future.

they already do give position evaluations of chesspositions
in "pawn units", you can watch this during the game on most standard
chess programs


You have missed my previous points about computer programs.

Their strength is only tactics. Their endgames , positional and long term
planning is very poor. At my level of approx 2200 they could only assess
my tactics. This is not a good basis for assessing my rating.

Most computers will beat the best humans purely on tactics if they look
deep enough. Us humans cannot evaluate millions of positions per sec for
material, so we use other methods of playing chess like the positional side,
long range planning , postional concepts and pattern recognition etc.
Computers are completely inadequate for assessing the human method
of playing chess.

Do you actually play chess and have you ever played in a chess
tournament ?. You show a lack of understandind of chess.

no / yes / sorry if I did-that was not intended
http://chessmetrics.com/PL/PL38915.htm


Yes ok - I see you are a similar strength to myself.

Do you understand how computers play chess ?

I am sorry about my last paragrapth in my last post.
It was very rude.

Regards

Terry

You have missed my previous points about computer programs.

Their strength is only tactics. Their endgames , positional and long term
planning is very poor. At my level of approx 2200 they could only assess
my tactics. This is not a good basis for assessing my rating.

Most computers will beat the best humans purely on tactics if they look
deep enough.

yet, they beat 'em. And that's (still) the only thing ,
what the rating system is about.

Us humans cannot evaluate millions of positions per sec for
material, so we use other methods of playing chess like the positional side,
long range planning , postional concepts and pattern recognition etc.
Computers are completely inadequate for assessing the human method
of playing chess.

you can hardly argue that computers are so weak positionally when
they have 2800 ELO. And this almost corresponds to their strength
when playing against human opponents. Subtract 20-50 points, if you want.
Sure, computers have their strong parts and their weaknesses, but
then (some) humans have this too. Yet I didn't see people
questioning the ELO of players specialized in tactics or
positional play or closed positions or endgames or
who always play some special openings or such.


Do you understand how computers play chess ?

better than I do understand, how humans play chess


Guenter.