Reply
 
LinkBack Thread Tools Display Modes
  #1   Report Post  
Old February 29th 04, 05:36 PM
Chris Kantack
 
Posts: n/a
Default analyses mode: evaluation values question

I'm currently playing around with ChessGenius on the Palm. ChessGenius has
an "Analyze game" option. It's like being in 2 (human) player mode but as
you move the pieces, ChessGenius continuously calculates and shows what
lines it's favoring.

It also displays the overall numeric value of the position. I've always
understood this value is to be in "# of pawns" with positive values favoring
white and a negative value meaning black is better.

ChessGenius doesn't seem to work this way. The value of the position flips
from positive to negative and back depending on which side is to move. I
even played an extremely lopsided game just to make sure this was what was
happening. White was up a full 12 points (or pawns). But when it is black
to move, the value shown was a -12.

I believe this is incorrect. The value should remain positive as long as
white is better and should only go negative if black is better. It should
not matter who's turn it is to move.

Am I correct in this assumption? Is ChessGenius in error here?
--
Chris Kantack
http://home.earthlink.net/~kantack/lcdchess/home.htm
(remove the "nospampls" from my email address if replying by email)
  #2   Report Post  
Old February 29th 04, 05:44 PM
Linares_Lulls
 
Posts: n/a
Default analyses mode: evaluation values question

What difference does it make as long as you know what side the program is
favouring? If you want to be a real nitpick for ergonomics.......learn how
to sling some code and write your own program.


"Chris Kantack" wrote in message
k.net...
I'm currently playing around with ChessGenius on the Palm. ChessGenius

has
an "Analyze game" option. It's like being in 2 (human) player mode but

as
you move the pieces, ChessGenius continuously calculates and shows what
lines it's favoring.

It also displays the overall numeric value of the position. I've always
understood this value is to be in "# of pawns" with positive values

favoring
white and a negative value meaning black is better.

ChessGenius doesn't seem to work this way. The value of the position

flips
from positive to negative and back depending on which side is to move. I
even played an extremely lopsided game just to make sure this was what was
happening. White was up a full 12 points (or pawns). But when it is

black
to move, the value shown was a -12.

I believe this is incorrect. The value should remain positive as long as
white is better and should only go negative if black is better. It

should
not matter who's turn it is to move.

Am I correct in this assumption? Is ChessGenius in error here?
--
Chris Kantack
http://home.earthlink.net/~kantack/lcdchess/home.htm
(remove the "nospampls" from my email address if replying by email)



  #3   Report Post  
Old February 29th 04, 06:38 PM
David Richerby
 
Posts: n/a
Default analyses mode: evaluation values question

Chris Kantack wrote:
ChessGenius doesn't seem to work this way. The value of the position
flips from positive to negative and back depending on which side is to
move. I even played an extremely lopsided game just to make sure this
was what was happening. White was up a full 12 points (or pawns). But
when it is black to move, the value shown was a -12.

I believe this is incorrect. The value should remain positive as long
as white is better and should only go negative if black is better. It
should not matter who's turn it is to move.

Am I correct in this assumption? Is ChessGenius in error here?


Many pieces of software use the convention that the score is always given
from White's point of view (i.e., a positive score means that White is
ahead; a negative score, that Black is ahead). However, there's nothing
at all wrong in giving the score from the point of view of the player who
has the move, as ChessGenius seems to be doing. It's just a different
convention that's no more or less right than the other.

My chess program also uses scores relative to the player who has the
move. Since all the data structures are also set up this way, it means
that neither the evaluation function nor the minimax algorithm needs to
know which colour has the move which, in fact, turns minimax into
`maximax'.


Dave.

--
David Richerby Solar-Powered Clock (TM): it's like a
www.chiark.greenend.org.uk/~davidr/ clock but it doesn't work in the dark!
  #4   Report Post  
Old March 1st 04, 02:25 AM
Vincent Diepeveen
 
Posts: n/a
Default analyses mode: evaluation values question

Majority of programs until a few years ago showed their values from the side
to move.

Only informator always showed it from white viewpoint.

non chess players when you say: "up 2 pawns" they *will* assume egoistic
score usually.

So always taking a look from the white side is not a natural thing to do.

My own program prints: "white up 12.0 pawns"

or "black up 12.0 pawns"

that's clearer. I was not the first to do this, Bruce Moreland i saw it do
in 1999 already (1997 i didn't take a look to what he was doing), so i don't
deserve credits for this system, but Bruce does.

Best regards,
Vinent.


"Chris Kantack" wrote in message
k.net...
I'm currently playing around with ChessGenius on the Palm. ChessGenius

has
an "Analyze game" option. It's like being in 2 (human) player mode but

as
you move the pieces, ChessGenius continuously calculates and shows what
lines it's favoring.

It also displays the overall numeric value of the position. I've always
understood this value is to be in "# of pawns" with positive values

favoring
white and a negative value meaning black is better.

ChessGenius doesn't seem to work this way. The value of the position

flips
from positive to negative and back depending on which side is to move. I
even played an extremely lopsided game just to make sure this was what was
happening. White was up a full 12 points (or pawns). But when it is

black
to move, the value shown was a -12.

I believe this is incorrect. The value should remain positive as long as
white is better and should only go negative if black is better. It

should
not matter who's turn it is to move.

Am I correct in this assumption? Is ChessGenius in error here?
--
Chris Kantack
http://home.earthlink.net/~kantack/lcdchess/home.htm
(remove the "nospampls" from my email address if replying by email)



  #5   Report Post  
Old March 1st 04, 04:16 AM
Chris Kantack
 
Posts: n/a
Default analyses mode: evaluation values question

In article ,
says...
Chris Kantack wrote:
ChessGenius doesn't seem to work this way. The value of the position
flips from positive to negative and back depending on which side is to
move. I even played an extremely lopsided game just to make sure this
was what was happening. White was up a full 12 points (or pawns). But
when it is black to move, the value shown was a -12.

I believe this is incorrect. The value should remain positive as long
as white is better and should only go negative if black is better. It
should not matter who's turn it is to move.

Am I correct in this assumption? Is ChessGenius in error here?


Many pieces of software use the convention that the score is always given
from White's point of view (i.e., a positive score means that White is
ahead; a negative score, that Black is ahead). However, there's nothing
at all wrong in giving the score from the point of view of the player who
has the move, as ChessGenius seems to be doing. It's just a different
convention that's no more or less right than the other.

My chess program also uses scores relative to the player who has the
move. Since all the data structures are also set up this way, it means
that neither the evaluation function nor the minimax algorithm needs to
know which colour has the move which, in fact, turns minimax into
`maximax'.


Dave.


Dave,
Thanks for the follow-up. That is why I asked the question. I had not
enough experience to know if what ChessGenius did was abnormal or not. It
appears from other postings that there is no set standard.

Thanks,
Chris
--
Chris Kantack
http://home.earthlink.net/~kantack/lcdchess/home.htm
(remove the "nospampls" from my email address if replying by email)


  #8   Report Post  
Old March 5th 04, 07:21 PM
Mark
 
Posts: n/a
Default analyses mode: evaluation values question

Jason, if you would have thought for just a moment before you posted you
would have realized that he is currently writing a review that will be
presented when it is complete. His original posts were not the review.

But - you probably already knew that didn't you?

Mark


"Linares_Lulls" wrote in message
news:0Yy0c.653697$X%[email protected]

"Chris Kantack" wrote in message
k.net...
In article pyp0c.646484$X%[email protected],
says...
What difference does it make as long as you know what side the program

is
favouring? If you want to be a real nitpick for ergonomics.......learn

how
to sling some code and write your own program.


Thanks but I'm not interested in writing my own program. What I am

doing
is writing a formal review of ChessGenius and was just curious as to

what
was considered standard. (Apparently there is no standard on displaying
overall score.)



You weren't "reviewing" anything. You were whining about the fact that you
need to have your hand held when using chess software





  #9   Report Post  
Old March 6th 04, 06:07 AM
Linares__Lulls
 
Posts: n/a
Default analyses mode: evaluation values question


"Mark" wrote in message
news:[email protected]_s54...
Jason, if you would have thought for just a moment before you posted you
would have realized that he is currently writing a review that will be
presented when it is complete. His original posts were not the review.

But - you probably already knew that didn't you?

Mark




Mark, you should really go back and finish your junior high so you don't
keep embarassing yourself like the uneducated jackass you are. I knew he was
writing a review. The guy said he was writing a review. There was never any
debate as to whether or not he is/was writing a review. If you got your head
out of your dumb uneducated ass long enough to get some fresh air before
posting your drivel, you would see that the fact he is writing a review has
nothing at all to do with my post. The guys comment was:

"I believe this is incorrect. The value should remain positive as long as
white is better and should only go negative if black is better. It should
not matter who's turn it is to move."

That is what the issue was about.

Jason Repa





"Linares_Lulls" wrote in message
news:0Yy0c.653697$X%[email protected]

"Chris Kantack" wrote in message
k.net...
In article pyp0c.646484$X%[email protected],
says...
What difference does it make as long as you know what side the

program
is
favouring? If you want to be a real nitpick for

ergonomics.......learn
how
to sling some code and write your own program.


Thanks but I'm not interested in writing my own program. What I am

doing
is writing a formal review of ChessGenius and was just curious as to

what
was considered standard. (Apparently there is no standard on

displaying
overall score.)



You weren't "reviewing" anything. You were whining about the fact that

you
need to have your hand held when using chess software







Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On



All times are GMT +1. The time now is 01:02 AM.

Powered by vBulletin® Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
Copyright 2004-2019 ChessBanter.
The comments are property of their posters.
 

About Us

"It's about Chess"

 

Copyright © 2017