Movei00_8_438 vs. Novag Jade QR odds

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

LucenaTheLucid
Posts: 197
Joined: Mon Jul 13, 2009 2:16 am

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by LucenaTheLucid »

After a long day Movei elects for Kd3

[d] 6k1/5ppp/1p2p3/3pPP2/3NrP2/3K4/1P5P/8 b - - 0 2

This experiment begs the question Steve...just how high an ELO can over come rook odds against Prissy?
Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by Steve B »

LucenaTheLucid wrote:After a long day Movei elects for Kd3

Jade having a multitude of pawn exchanges to choose from opts for ..
32..exf5
LucenaTheLucid wrote:
This experiment begs the question Steve...just how high an ELO can over come rook odds against Prissy?
An Interesting Question...
another Question ...
can one of my machines rated 2000+..draw or defeat a 3000+ rated engine at Rook odds..does the R =1000 ELO theory ..hold at higher Elos ??

So many questions..So little time Regards
Steve
Vinvin
Posts: 5297
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by Vinvin »

One big paramaters of the problems is : "do the engine is aware how to play odds games ?" or "Is the engine able to handle a game when he is behind at material ?" or : keep a lot of material, close the lines, don't exchange pieces too quickly, recognize a lost endgame (yes, the game is lost in the starting position :-/ ) , ...
LucenaTheLucid wrote:This experiment begs the question Steve...just how high an ELO can over come rook odds against Prissy?
Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by Steve B »

Vinvin wrote:One big paramaters of the problems is : "do the engine is aware how to play odds games ?" or "Is the engine able to handle a game when he is behind at material ?" or : keep a lot of material, close the lines, don't exchange pieces too quickly, recognize a lost endgame (yes, the game is lost in the starting position :-/ ) , ...
LucenaTheLucid wrote:This experiment begs the question Steve...just how high an ELO can over come rook odds against Prissy?
i guess no engine programmers will attempt to program engines to play odds games like this
it takes alot of effort and there is very little interest in odds games
even as common and widely accepted chess variant as Chess 960 does not see a huge amount of interest from leading programmers today although there does seem to be some movement in this area as of late ..it took awhile though
the main focus in CC was always ELO..the higher the better..always was that way and always will be i guess
even when the dedicated's ruled the commercial CC market the emphasis was always on cold ..hard ..ELO


one thing to note is that i played about half a dozen games with 1700 Dedicated Units against a 2700+ engine at minor piece odds and the engine won every game..
it seems the R does the trick

perhaps minor piece +P odds would also overcome 1000 Elo but the decision of which P to remove creates a whole different scenario in an odds match which is a bit beyond the scope of my experiment

Regards
Steve
Vinvin
Posts: 5297
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by Vinvin »

Steve B wrote:
Vinvin wrote:One big paramaters of the problems is : "do the engine is aware how to play odds games ?" or "Is the engine able to handle a game when he is behind at material ?" or : keep a lot of material, close the lines, don't exchange pieces too quickly, recognize a lost endgame (yes, the game is lost in the starting position :-/ ) , ...
LucenaTheLucid wrote:This experiment begs the question Steve...just how high an ELO can over come rook odds against Prissy?
i guess no engine programmers will attempt to program engines to play odds games like this...
Rybka 3 does (I suppose Rybka 4 too ...)
Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by Steve B »

Vinvin wrote:
Steve B wrote:
Vinvin wrote:One big paramaters of the problems is : "do the engine is aware how to play odds games ?" or "Is the engine able to handle a game when he is behind at material ?" or : keep a lot of material, close the lines, don't exchange pieces too quickly, recognize a lost endgame (yes, the game is lost in the starting position :-/ ) , ...
LucenaTheLucid wrote:This experiment begs the question Steve...just how high an ELO can over come rook odds against Prissy?
i guess no engine programmers will attempt to program engines to play odds games like this...
Rybka 3 does (I suppose Rybka 4 too ...)

Thanks
i knew Rybka could play 960 but i didn't realize it was also programmed for odds games
i suppose the Rybka's are rated at or near 3000 Elo?
perhaps one day a future match...
a dedicated unit rated 2000+ v Rybka 3/4 at R odds


Regards
Steve
lkaufman
Posts: 6258
Joined: Sun Jan 10, 2010 6:15 am
Location: Maryland USA
Full name: Larry Kaufman

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by lkaufman »

Rybka 4 is not especially designed for odds games, but the parameters allow you to value computer pieces more than opponent pieces, so if these values are well chosen it will make the engine avoid even trades to an appropriate extent and thus will make it play much more suitably for odds games.
LucenaTheLucid
Posts: 197
Joined: Mon Jul 13, 2009 2:16 am

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by LucenaTheLucid »

Movei plays Nxf5:

[d] 6k1/5ppp/1p6/3pPN2/4rP2/3K4/1P5P/8 b - - 0 3
Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by Steve B »

LucenaTheLucid wrote:Movei plays Nxf5:
Jade.. a bit concerned with a possible check from Movei's Pesky N plays..
33..Kf8

Overly Cautious Regards
Steve
LucenaTheLucid
Posts: 197
Joined: Mon Jul 13, 2009 2:16 am

Re: Movei00_8_438 vs. Saitek Prisma QR odds

Post by LucenaTheLucid »

Movei in response plays Ne3

[d] 5k2/5ppp/1p6/3pP3/4rP2/3KN3/1P5P/8 b - - 0 2