To analysis.sesse.net owner : disable contempt

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
syzygy
Posts: 4343
Joined: Tue Feb 28, 2012 10:56 pm

Re: To analysis.sesse.net owner : disable contempt

Post by syzygy » Sun Dec 02, 2018 3:44 pm

BeyondCritics wrote:
Sun Dec 02, 2018 3:15 pm
Wiggling occurs, if "contempt=both" is set. This you may call problematic or "wrong", if you like.
http://talkchess.com/forum3/viewtopic.p ... 65#p779660
Syzygy wrote:This is why "Off" should be the default value of Analysis Contempt (and "Both" is simply useless).
And this obviously has nothing to do with the tapering.

The contempt implementation is not the problem. The problem is that contempt is left on in analysis mode by default.

BeyondCritics
Posts: 335
Joined: Sat May 05, 2012 12:48 pm
Location: Bergheim

Re: To analysis.sesse.net owner : disable contempt

Post by BeyondCritics » Sun Dec 02, 2018 3:48 pm

syzygy wrote:
Sun Dec 02, 2018 3:44 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:15 pm
Wiggling occurs, if "contempt=both" is set. This you may call problematic or "wrong", if you like.
http://talkchess.com/forum3/viewtopic.p ... 65#p779660
Syzygy wrote:This is why "Off" should be the default value of Analysis Contempt (and "Both" is simply useless).
And this obviously has nothing to do with the tapering.

The contempt implementation is not the problem. The problem is that contempt is left on in analysis mode.
How do you know what the problem is, without knowing what it really does? I don't get that.

syzygy
Posts: 4343
Joined: Tue Feb 28, 2012 10:56 pm

Re: To analysis.sesse.net owner : disable contempt

Post by syzygy » Sun Dec 02, 2018 3:49 pm

BeyondCritics wrote:
Sun Dec 02, 2018 3:48 pm
syzygy wrote:
Sun Dec 02, 2018 3:44 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:15 pm
Wiggling occurs, if "contempt=both" is set. This you may call problematic or "wrong", if you like.
http://talkchess.com/forum3/viewtopic.p ... 65#p779660
Syzygy wrote:This is why "Off" should be the default value of Analysis Contempt (and "Both" is simply useless).
And this obviously has nothing to do with the tapering.

The contempt implementation is not the problem. The problem is that contempt is left on in analysis mode.
How do you know what the problem is, without knowing what it really does? I don't get that.
Ego hurts badly? I don't know what to make of you....

BeyondCritics
Posts: 335
Joined: Sat May 05, 2012 12:48 pm
Location: Bergheim

Re: To analysis.sesse.net owner : disable contempt

Post by BeyondCritics » Sun Dec 02, 2018 3:54 pm

syzygy wrote:
Sun Dec 02, 2018 3:49 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:48 pm
syzygy wrote:
Sun Dec 02, 2018 3:44 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:15 pm
Wiggling occurs, if "contempt=both" is set. This you may call problematic or "wrong", if you like.
http://talkchess.com/forum3/viewtopic.p ... 65#p779660
Syzygy wrote:This is why "Off" should be the default value of Analysis Contempt (and "Both" is simply useless).
And this obviously has nothing to do with the tapering.

The contempt implementation is not the problem. The problem is that contempt is left on in analysis mode.
How do you know what the problem is, without knowing what it really does? I don't get that.
Ego hurts badly? I don't know what to make of you....
Really? Please reread any posting here from me. If i was wrong i have always admitted it, without forcing others going into exhaustive fightings. And countless times i have flinched, when attacked from someone, who appeared less intelligble, because i knew it is pointless to try to presuade him. So you may take it as compliment.

syzygy
Posts: 4343
Joined: Tue Feb 28, 2012 10:56 pm

Re: To analysis.sesse.net owner : disable contempt

Post by syzygy » Sun Dec 02, 2018 4:07 pm

BeyondCritics wrote:
Sun Dec 02, 2018 3:54 pm
syzygy wrote:
Sun Dec 02, 2018 3:49 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:48 pm
syzygy wrote:
Sun Dec 02, 2018 3:44 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:15 pm
Wiggling occurs, if "contempt=both" is set. This you may call problematic or "wrong", if you like.
http://talkchess.com/forum3/viewtopic.p ... 65#p779660
Syzygy wrote:This is why "Off" should be the default value of Analysis Contempt (and "Both" is simply useless).
And this obviously has nothing to do with the tapering.

The contempt implementation is not the problem. The problem is that contempt is left on in analysis mode.
How do you know what the problem is, without knowing what it really does? I don't get that.
Ego hurts badly? I don't know what to make of you....
Really? Please reread any posting here from me. If i was wrong i have always admitted it, without forcing others going into exhaustive fightings. And countless times i have flinched, when attacked from someone, who appeared less intelligble, because i knew it is pointless to try to presuade him. So you may take it as compliment.
So why did you, after the discussion had basically ended by our agreeing that the problem lies with the "Both" setting, restart this whole thing by suggesting I don't know "what it really does"?

BeyondCritics
Posts: 335
Joined: Sat May 05, 2012 12:48 pm
Location: Bergheim

Re: To analysis.sesse.net owner : disable contempt

Post by BeyondCritics » Sun Dec 02, 2018 4:13 pm

syzygy wrote:
Sun Dec 02, 2018 4:07 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:54 pm
syzygy wrote:
Sun Dec 02, 2018 3:49 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:48 pm
syzygy wrote:
Sun Dec 02, 2018 3:44 pm
BeyondCritics wrote:
Sun Dec 02, 2018 3:15 pm
Wiggling occurs, if "contempt=both" is set. This you may call problematic or "wrong", if you like.
http://talkchess.com/forum3/viewtopic.p ... 65#p779660
Syzygy wrote:This is why "Off" should be the default value of Analysis Contempt (and "Both" is simply useless).
And this obviously has nothing to do with the tapering.

The contempt implementation is not the problem. The problem is that contempt is left on in analysis mode.
How do you know what the problem is, without knowing what it really does? I don't get that.
Ego hurts badly? I don't know what to make of you....
Really? Please reread any posting here from me. If i was wrong i have always admitted it, without forcing others going into exhaustive fightings. And countless times i have flinched, when attacked from someone, who appeared less intelligble, because i knew it is pointless to try to presuade him. So you may take it as compliment.
So why did you, after the discussion had basically ended by our agreeing that the problem lies with the "Both" setting, restart this whole thing by suggesting I don't know "what it really does"?
E.g because you suggested to subtract the raw contempt value from the root value. Do you still want to do that?
Last edited by BeyondCritics on Sun Dec 02, 2018 4:20 pm, edited 1 time in total.

syzygy
Posts: 4343
Joined: Tue Feb 28, 2012 10:56 pm

Re: To analysis.sesse.net owner : disable contempt

Post by syzygy » Sun Dec 02, 2018 4:19 pm

BeyondCritics wrote:
Sun Dec 02, 2018 4:13 pm
E.g because you suggested to subtract the raw contempt value from the root value. Do you still want to do that?
What are you talking about?

Even in my first explanation, I clearly pointed out that the contempt bonus was added "throughout the tree".

That the contempt bonus is tapered is true, but irrelevant to the discussion. The important thing is that the bonus is awarded to white in all tree nodes (if white is to move at the root). I strongly suspect that your initial understanding was that its sign depended on the side to move at the internal tree node. But I cannot prove that suspicion because you never actually explained your understanding very well.

(Now you will probably look for some other irrelevant argument, like bonus/penalty, to avoid having to address the real issue.)

BeyondCritics
Posts: 335
Joined: Sat May 05, 2012 12:48 pm
Location: Bergheim

Re: To analysis.sesse.net owner : disable contempt

Post by BeyondCritics » Sun Dec 02, 2018 4:32 pm

syzygy wrote:
Sun Dec 02, 2018 4:19 pm
BeyondCritics wrote:
Sun Dec 02, 2018 4:13 pm
E.g because you suggested to subtract the raw contempt value from the root value. Do you still want to do that?
What are you talking about?

Even in my first explanation, I clearly pointed out that the contempt bonus was added "throughout the tree".

That the contempt bonus is tapered is true, but irrelevant to the discussion. The important thing is that the bonus is awarded to white in all tree nodes (if white is to move at the root). I strongly suspect that your initial understanding was that its sign depended on the side to move at the internal tree node. But I cannot prove that suspicion because you never actually explained your understanding very well.

(Now you will probably look for some other irrelevant argument, like bonus/penalty, to avoid having to address the real issue.)
You wrote:
So if contempt is +0.20 and the objective score is about +1.00 for white, you'll see about +1.20 if white is to move at the root and about +0.80 if black is to move at the root.

Do you still hold that?

syzygy
Posts: 4343
Joined: Tue Feb 28, 2012 10:56 pm

Re: To analysis.sesse.net owner : disable contempt

Post by syzygy » Sun Dec 02, 2018 4:42 pm

BeyondCritics wrote:
Sun Dec 02, 2018 4:32 pm
syzygy wrote:
Sun Dec 02, 2018 4:19 pm
BeyondCritics wrote:
Sun Dec 02, 2018 4:13 pm
E.g because you suggested to subtract the raw contempt value from the root value. Do you still want to do that?
What are you talking about?

Even in my first explanation, I clearly pointed out that the contempt bonus was added "throughout the tree".

That the contempt bonus is tapered is true, but irrelevant to the discussion. The important thing is that the bonus is awarded to white in all tree nodes (if white is to move at the root). I strongly suspect that your initial understanding was that its sign depended on the side to move at the internal tree node. But I cannot prove that suspicion because you never actually explained your understanding very well.

(Now you will probably look for some other irrelevant argument, like bonus/penalty, to avoid having to address the real issue.)
You wrote:
So if contempt is +0.20 and the objective score is about +1.00 for white, you'll see about +1.20 if white is to move at the root and about +0.80 if black is to move at the root.

Do you still hold that?
I still hold that that sentence together with the next sentence clearly and concisely explains the issue without getting too much into details that aren't really important here. Look, you had already explained the tapered thing, why should I have to repeat it again? (I now understand that the discovery of tapered contempt in the SF sources came as a revelation to you. For me it was very old hat, so for me it made a lot of sense to simplify it away, as it was simply unrelated to the discussion. I am sorry that this confused you.)
syzygy wrote:So if contempt is +0.20 and the objective score is about +1.00 for white, you'll see about +1.20 if white is to move at the root and about +0.80 if black is to move at the root. Each time Carlsen or Caruana plays a move, the side to move at the root switches and the score wiggles.
Why do you think I put the stress where I put it?

But do you now at least understand the point I was making?

BeyondCritics
Posts: 335
Joined: Sat May 05, 2012 12:48 pm
Location: Bergheim

Re: To analysis.sesse.net owner : disable contempt

Post by BeyondCritics » Sun Dec 02, 2018 4:58 pm

syzygy wrote:
Sun Dec 02, 2018 4:42 pm
BeyondCritics wrote:
Sun Dec 02, 2018 4:32 pm
syzygy wrote:
Sun Dec 02, 2018 4:19 pm
BeyondCritics wrote:
Sun Dec 02, 2018 4:13 pm
E.g because you suggested to subtract the raw contempt value from the root value. Do you still want to do that?
What are you talking about?

Even in my first explanation, I clearly pointed out that the contempt bonus was added "throughout the tree".

That the contempt bonus is tapered is true, but irrelevant to the discussion. The important thing is that the bonus is awarded to white in all tree nodes (if white is to move at the root). I strongly suspect that your initial understanding was that its sign depended on the side to move at the internal tree node. But I cannot prove that suspicion because you never actually explained your understanding very well.

(Now you will probably look for some other irrelevant argument, like bonus/penalty, to avoid having to address the real issue.)
You wrote:
So if contempt is +0.20 and the objective score is about +1.00 for white, you'll see about +1.20 if white is to move at the root and about +0.80 if black is to move at the root.

Do you still hold that?
I still hold that that sentence together with the next sentence clearly and concisely explains the issue without getting too much into details that aren't really important here. Look, you had already explained the tapered thing, why should I have to repeat it again? (I now understand that the discovery of tapered contempt in the SF sources came as a revelation to you. For me it was very old hat, so for me it made a lot of sense to simplify it away, as it was simply unrelated to the discussion. I am sorry that this confused you.)
syzygy wrote:So if contempt is +0.20 and the objective score is about +1.00 for white, you'll see about +1.20 if white is to move at the root and about +0.80 if black is to move at the root. Each time Carlsen or Caruana plays a move, the side to move at the root switches and the score wiggles.
Why do you think I put the stress where I put it?

But do you now at least understand the point I was making?
So you are referring to the next sentence. Together it reads:
So if contempt is +0.20 and the objective score is about +1.00 for white, you'll see about +1.20 if white is to move at the root and about +0.80 if black is to move at the root. Each time Carlsen or Caruana plays a move, the side to move at the root switches and the score wiggles.

What you seem to be thinking is that the bonus/penalty is applied in each internal node of the tree with a sign that depends on whether white or black is to move at that node.
You are lamenting having been cited incorrectly, the next sentence is relevant. In the next sentence you are talking about your assumptions about my thinking. What does that has to do with issue at hand? I don't get you at all, really. I think every prudent observer should agree, that i have cited you correctly here.
I see that you are start to retort "ad personam" now. A clear sign of running out of arguments.

Post Reply