Page 9 of 11

Re: To Larry Kaufman

Posted: Sat Mar 12, 2016 10:03 pm
by Nordlandia
Here's is a similar challange:

NDE 1986 (2nd Prize, Israel Ring 1986)

[d]5k2/1n6/1p3K2/P7/1P2N3/p1P2B2/8/2b5 w - - 0 0

1 a6 a2 2 axb7 Bf4 3 Ng5! Bb8! (a1Q 4 Ne6+ Kg8! 5 Nxf4 Qxc3+ 6 Kf5 and Black can't win; 3 . . . Bd6 4 Ne6+ Kg8 5 Nc7! Bxc7 6 Bc4+ draws. So the Bishop crosses the critical square c7, which however White can exploit too:) 4 Nh7+! Ke8 5 Bc6+ Kd8 6 Ng5 Kc7 7 b5! a1Q (Kd6!? 8 Ne4+ Kc7 9 Ng5 repeating) 8 Nf7! positional draw.

[d]1b6/1Pk2N2/1pB2K2/1P6/8/2P5/8/q7 b - - 0 7

Re: To Larry Kaufman

Posted: Sat Mar 12, 2016 10:19 pm
by Jesse Gersenson
Larry, is this new term applied during analysis mode?

Re: To Larry Kaufman

Posted: Sat Mar 12, 2016 11:51 pm
by TShackel
lkaufman wrote:The new term cannot be set differently for middlegame or endgame, but I think there is no need. If a move allows the opponent more checks or attacks on our king, a higher setting will avoid the move. In other words, it's symmetric; it treats both sides the same.
We could be off a few percent in the optimum setting, and it may also be that the optimum setting is quite different for long time control games than for the levels at which we can test. I suppose if you set it for something like 110% of our recommendation the elo loss will be pretty small, probably single digit, so if that makes Komodo aggressive enough for you you can use that all the time (except for official tests of course). But a really high setting like 150% would cost a lot of elo. Pawns do have value!
Thanks for the tips as to how high the value could go with only a loss of a few elo. That might be advantageous for some type of contempt concept as well, that against weaker engines maybe use a higher value.

But nevertheless, it is clear to me you are listening to your subscribers for new UCI parameters while always focusing on elo gain. And you are very innovative with novel, new programming ideas. Congrats.

Sincerely,

Tim.

Re: To Larry Kaufman

Posted: Sun Mar 13, 2016 2:55 am
by mjlef
I ran the position with a few values for the new parameter, and yes, with a high enough value it finds the move very quickly.

Re: To Larry Kaufman

Posted: Sun Mar 13, 2016 4:45 am
by lkaufman
Jesse Gersenson wrote:Larry, is this new term applied during analysis mode?
I don't know any reason it wouldn't be valid for analysis mode, though I'll have to test the final version to confirm that this is actually so.

Re: To Larry Kaufman

Posted: Sun Mar 13, 2016 4:48 am
by lkaufman
Nordlandia wrote:Someone suggested adjustable 50-move rule for certain positions that is usually a draw, e.g Behting Study.

For example 25-move rule may be adequate in that particular position.

Is this possible to add in Komodo parameters?

https://chessprogramming.wikispaces.com/Behting+Study
We could add it, but the problem is that some GUIs have a limit on the number of UCI options, and we are near that limit now I think for some GUIs. I think it is not the most useful option we could add.

Re: To Larry Kaufman

Posted: Sun Mar 13, 2016 2:40 pm
by Nordlandia
lkaufman wrote:
Nordlandia wrote:Someone suggested adjustable 50-move rule for certain positions that is usually a draw, e.g Behting Study.

For example 25-move rule may be adequate in that particular position.

Is this possible to add in Komodo parameters?

https://chessprogramming.wikispaces.com/Behting+Study
We could add it, but the problem is that some GUIs have a limit on the number of UCI options, and we are near that limit now I think for some GUIs. I think it is not the most useful option we could add.
The difference will be clear 0.00 evaluation against +3.00 or even 4 in such scenarios. Maybe not the most useful parameter but fine for certain studies as mentioned above.

Re: To Larry Kaufman

Posted: Mon Mar 14, 2016 12:41 am
by carldaman
lkaufman wrote:
Nordlandia wrote:Someone suggested adjustable 50-move rule for certain positions that is usually a draw, e.g Behting Study.

For example 25-move rule may be adequate in that particular position.

Is this possible to add in Komodo parameters?

https://chessprogramming.wikispaces.com/Behting+Study
We could add it, but the problem is that some GUIs have a limit on the number of UCI options, and we are near that limit now I think for some GUIs. I think it is not the most useful option we could add.
Hi Larry,

Thanks for always trying to make Komodo a better analysis partner. I was wondering about the UCI option limit - what exactly is the limit? I remember seeing a lot of parameters in other UCI engines (Fruit comes to mind), and I did not think Komodo came anywhere close to having that many parameters. Or, are options and parameters a 'different animal' altogether?

Thanks,
CL

Re: To Larry Kaufman

Posted: Mon Mar 14, 2016 2:13 am
by lkaufman
carldaman wrote:
lkaufman wrote:
Nordlandia wrote:Someone suggested adjustable 50-move rule for certain positions that is usually a draw, e.g Behting Study.

For example 25-move rule may be adequate in that particular position.

Is this possible to add in Komodo parameters?

https://chessprogramming.wikispaces.com/Behting+Study
We could add it, but the problem is that some GUIs have a limit on the number of UCI options, and we are near that limit now I think for some GUIs. I think it is not the most useful option we could add.
Hi Larry,

Thanks for always trying to make Komodo a better analysis partner. I was wondering about the UCI option limit - what exactly is the limit? I remember seeing a lot of parameters in other UCI engines (Fruit comes to mind), and I did not think Komodo came anywhere close to having that many parameters. Or, are options and parameters a 'different animal' altogether?

Thanks,
CL
The limit is set by the GUI, so each is different. We have no way to know what the minimum value is for all known GUIs, so we just limit it to below any GUIs we actually test. We could probably add one or two more safely, but it's not worth taking the risk for an option that would be rarely used.

Re: To Larry Kaufman

Posted: Mon Mar 14, 2016 6:49 am
by Jesse Gersenson
lkaufman wrote:The limit is set by the GUI, so each is different. We have no way to know what the minimum value is for all known GUIs, so we just limit it to below any GUIs we actually test. We could probably add one or two more safely, but it's not worth taking the risk for an option that would be rarely used.
Why not off-load parameters into a configuration file?