Thanks for the mobility info ed!

[Steelman]

First: Ed, thanks a lot for getting back to your programmers page and adding that stuff on mobility. Now that you have started I check the forum every so often to see if you have added anything else. This stuff is great!
I can't even imagine the hundreds of hours of testing, trying this, trying that, adjusting this, tweaking a value here, replacing a section of code with something else, and so on you have tried over the years when you was developing rebel! And the hundreds even thousands of test and tournament games you must have played to come up with these ideas.
This weekend I'm going to try coding some of the mobility ideas into my eval function, run some test games, and see what happens.

Speaking for myself and I am sure others again Thanks! I'm looking forward for some more when you find the time.

Second thing: I hope you can get to read this post before it disappears!

......Ron P.

[WinPooh]

Can you please post direct link to this paper? I can't find MOBILITY section at http://members.home.nl/matador/chess840.htm#MOBILITY

[Uri Blass]

First: Ed, thanks a lot for getting back to your programmers page and adding that stuff on mobility. Now that you have started I check the forum every so often to see if you have added anything else. This stuff is great!
I can't even imagine the hundreds of hours of testing, trying this, trying that, adjusting this, tweaking a value here, replacing a section of code with something else, and so on you have tried over the years when you was developing rebel! And the hundreds even thousands of test and tournament games you must have played to come up with these ideas.
This weekend I'm going to try coding some of the mobility ideas into my eval function, run some test games, and see what happens.

Speaking for myself and I am sure others again Thanks! I'm looking forward for some more when you find the time.

Second thing: I hope you can get to read this post before it disappears!

......Ron P.

I also thank Ed for the new material.

My opinion is that prodeo like many other engines has too much "knowledge" and it may be possible to earn rating in prodeo simply by having less knowledge.

Movei seems to be almost in the same playing strength as prodeo based on the CCRL 40/4 list and maybe the latest private version that was not tested by CCRL is even better than prodeo.

I feel that prodeo has more knowledge than movei based on reading the page so my guess is that simply part of the knowledge in prodeo is counterproductive(considering the fact that movei still does not use hash for pruning I doubt if movei has a superior search to prodeo).

Note that even in case that evey single change is productive it is possible that you have too much knowledge.

It is possible that adding knowledge A gives 5 elo improvement
adding knowledge B gives another 5 elo improvement
adding knowledge C gives another 5 elo improvement and after doing all these deleting knowledge A gives another 5 elo improvement.

Uri

[Uri Blass]

Can you please post direct link to this paper? I can't find MOBILITY section at http://members.home.nl/matador/chess840.htm#MOBILITY

I do not know about direct link but you can find mobility in the following link:
http://www.top-5000.nl/authors/rebel/chess840.htm

You only need to click 68 times on page down in order to find mobility

Edit:I found a direct link

http://www.top-5000.nl/authors/rebel/ch ... m#MOBILITY

[ed]

Can you please post direct link to this paper? I can't find MOBILITY section at http://members.home.nl/matador/chess840.htm#MOBILITY

I do not know about direct link but you can find mobility in the following link:
http://www.top-5000.nl/authors/rebel/chess840.htm

You only need to click 68 times on page down in order to find mobility

On the right part you find the link menu plus the new added stuff

Ed

[WinPooh]

Thank you, it looks like my old bookmark in browser was obsolete.
BTW, is it possible to update PDF version of the document (current one is 3 years old, and is finished at "King safety" topic)?

[ed]

First: Ed, thanks a lot for getting back to your programmers page and adding that stuff on mobility. Now that you have started I check the forum every so often to see if you have added anything else. This stuff is great!
I can't even imagine the hundreds of hours of testing, trying this, trying that, adjusting this, tweaking a value here, replacing a section of code with something else, and so on you have tried over the years when you was developing rebel! And the hundreds even thousands of test and tournament games you must have played to come up with these ideas.
This weekend I'm going to try coding some of the mobility ideas into my eval function, run some test games, and see what happens.

Speaking for myself and I am sure others again Thanks! I'm looking forward for some more when you find the time.

Second thing: I hope you can get to read this post before it disappears!

......Ron P.

Thx Ron. Having earned a lot of money for 20 years from you guys I am happy to do something in return. And it gives me pleasure it is appreciated.

Ed

[YL84]

Hi,
There are some few sites describing the basics of chess programming (alpha-beta, quiescence,...). They permit to create a decent chess engine. But your website is the most detailed about the tip and tricks of programmation, which may help to create a strong chess
engine. Though I have still some years before I arrive to that point!
Thanks Ed!