Underpromotion percentage

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

Moderator: Ras

Tony Thomas

Re: Underpromotion percentage

Post by Tony Thomas »

There are other engines such as queen that doesnt see it, so this isnt a Rybka only problem..
User avatar
Marek Soszynski
Posts: 586
Joined: Wed May 10, 2006 7:28 pm
Location: Birmingham, England

Re: Underpromotion percentage

Post by Marek Soszynski »

Tony Thomas wrote:There are other engines such as queen that doesnt see it, so this isnt a Rybka only problem..
It's interesting that you call it a problem.

Note that Rybka has been deliberately crippled in this way; for Rybka is ruthlessly optimised for game play success, not analysis.
Marek Soszynski
Tony Thomas

Re: Underpromotion percentage

Post by Tony Thomas »

Marek Soszynski wrote:
Tony Thomas wrote:There are other engines such as queen that doesnt see it, so this isnt a Rybka only problem..
It's interesting that you call it a problem.

Note that Rybka has been deliberately crippled in this way; for Rybka is ruthlessly optimised for game play success, not analysis.
If there are other programs such as Symbolic that sees it without any issues, it is indeed a problem.
User avatar
Marek Soszynski
Posts: 586
Joined: Wed May 10, 2006 7:28 pm
Location: Birmingham, England

Re: Underpromotion percentage

Post by Marek Soszynski »

Tony Thomas wrote:If there are other programs such as Symbolic that sees it without any issues, it is indeed a problem.
A problem for those who use Rybka for analysis and assume they can trust it in all cases with no need for other engines.
Marek Soszynski
Tony Thomas

Re: Underpromotion percentage

Post by Tony Thomas »

Marek Soszynski wrote:
Tony Thomas wrote:If there are other programs such as Symbolic that sees it without any issues, it is indeed a problem.
A problem for those who use Rybka for analysis and assume they can trust it in all cases with no need for other engines.
Or a problem for those who expect Rybka to win that position in a tournament..
User avatar
Ovyron
Posts: 4562
Joined: Tue Jul 03, 2007 4:30 am

Re: Underpromotion percentage

Post by Ovyron »

And a problem that Vas just confirmed isn't getting fixed, just like valuing unstoppable pawns higher than a queen.

On the other hand, he's going to fix about 20 other bugs that Rybka 3 had.
Tony Thomas

Re: Underpromotion percentage

Post by Tony Thomas »

Ovyron wrote:And a problem that Vas just confirmed isn't getting fixed, just like valuing unstoppable pawns higher than a queen.

On the other hand, he's going to fix about 20 other bugs that Rybka 3 had.
I guess the only way Vas would fix that position is if Rybka loses a tournament because of that bug. Remember the Knight underpromotion problem with earlier versions of Rybka? He didnt fix it until he experienced the problem twice in one of the ICC tournaments..
User avatar
Ovyron
Posts: 4562
Joined: Tue Jul 03, 2007 4:30 am

Re: Underpromotion percentage

Post by Ovyron »

Indeed. Too bad the chance of an useful Bishop Underpromotion appearing is, very, very low, so low that I think Vas has to play thousands of tourneys before he loses that 0.5 of a score.
Tony Thomas

Re: Underpromotion percentage

Post by Tony Thomas »

Ovyron wrote:Indeed. Too bad the chance of an useful Bishop Underpromotion appearing is, very, very low, so low that I think Vas has to play thousands of tourneys before he loses that 0.5 of a score.
I wish I had a gang of rybka hating bookers at my disposal. They could create a line so good that Rybka would fall for it.. :wink:
User avatar
Ovyron
Posts: 4562
Joined: Tue Jul 03, 2007 4:30 am

Re: Underpromotion percentage

Post by Ovyron »

But you also need a good engine. Junior outbooked Rybka in the last tourney and had a won game, then blew it up, if Hiarcs, Sjeng or another engine played that opening against Rybka it would have been perhaps a win against Rybka. Book wins aren't easy against her.