Rybka 4 vs Houdini 1.02 - Long TC Match - LIVE BROADCAST

Discussion of computer chess matches and engine tournaments.

Moderators: Harvey Williamson, bob, hgm

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
S.Taylor
Posts: 8481
Joined: Thu Mar 09, 2006 2:25 am
Location: Jerusalem Israel

Re: the next match

Post by S.Taylor » Mon Jul 05, 2010 12:36 pm

Now Houdini reverts to its self-destruct tendencies, move 15.....Ng3 (game 43)

pacifist

Re: the next match

Post by pacifist » Mon Jul 05, 2010 1:49 pm

Rybka should have won game 38 but the evaluation misleaded it and it slipped into a fortress with exchange up. Humans players tend to avoid this but the computer doesn't "see" it until it's too late and too many pawns have been exchanged :(.

S.Taylor
Posts: 8481
Joined: Thu Mar 09, 2006 2:25 am
Location: Jerusalem Israel

Re: the next match

Post by S.Taylor » Mon Jul 05, 2010 2:56 pm

S.Taylor wrote:Now Houdini reverts to its self-destruct tendencies, move 15.....Ng3 (game 43)
(At 43.Nd1)
This wild wild game, in which Houdini grabbed and tried to utilize his initiative, is now spent, and is no more, and if he gets a draw, he should be happy.
Houdini may well have prevailed against any other engine or human (opinions?), and if he gets a draw vs R4, and was black, then i think he had good value for his money (riskiness).
It is interesting to speculate if there could have been better options, than (what is seen from what followed) 15....Ng3

(disclaimer: I'm really really sorry, that i speak too much on this thread. I hope no one minds. I only wish there were more discussion of the games. I don't usually view any games that are in proccess, as i do not normally see them with the technology setup i have. But here, i view them [but not with an analyzing engine beside it])
Last edited by S.Taylor on Mon Jul 05, 2010 3:06 pm, edited 1 time in total.

pacifist

Re: the next match

Post by pacifist » Mon Jul 05, 2010 3:04 pm

It was interesting to make a draw but I think black would have very good attacking chances on the kingside. Exchange Rook + pawn vs 2 knights is ok since black had possibilities of mining the queenside with a5 and b6. It seems to me Houdini played very well from there and is now assured the half-point. Still I would have preferred a kingside attack :)

Martin Thoresen
Posts: 1833
Joined: Wed Jun 21, 2006 10:07 pm

Re: the next match

Post by Martin Thoresen » Mon Jul 05, 2010 3:46 pm

S.Taylor wrote: (disclaimer: I'm really really sorry, that i speak too much on this thread. I hope no one minds. I only wish there were more discussion of the games. I don't usually view any games that are in proccess, as i do not normally see them with the technology setup i have. But here, i view them [but not with an analyzing engine beside it])
I am very happy that you make all these comments!

:)

S.Taylor
Posts: 8481
Joined: Thu Mar 09, 2006 2:25 am
Location: Jerusalem Israel

Re: the next match

Post by S.Taylor » Mon Jul 05, 2010 7:14 pm

Thanks, Martin.


I feel already now, that game 44 can be written off as a draw. I'm willing to be challenged, I wish i was wrong.

(black has well equalized at move 20, maybe a little more, but white has no problem to keep it accurate and draw).

pacifist

Re: the next match

Post by pacifist » Mon Jul 05, 2010 7:26 pm

White is better due to queen side extra pawn since in the king side there are doubled pawns for black. If black can exchange a doubled pawn or is helped by some tactics he can maybe draw. Black has to work hard to achieve a draw here I think.

14...Qf5 was a dubius move for black (at least for a human understanding) :)

S.Taylor
Posts: 8481
Joined: Thu Mar 09, 2006 2:25 am
Location: Jerusalem Israel

Re: the next match

Post by S.Taylor » Mon Jul 05, 2010 10:13 pm

pacifist wrote:White is better due to queen side extra pawn since in the king side there are doubled pawns for black. If black can exchange a doubled pawn or is helped by some tactics he can maybe draw. Black has to work hard to achieve a draw here I think.

14...Qf5 was a dubius move for black (at least for a human understanding) :)
To "work hard" is no object for these computers.
I also don't understand how Rybka allowed itself to play 14...Qf5, but a few moves later, when a set of rooks were exchanged, i felt that it had benefitted from the sequence, and had a good game, and needn't worry about those technicalities.

And NOW!
Houdini MUST win the remaining 4 games to win.

Martin Thoresen
Posts: 1833
Joined: Wed Jun 21, 2006 10:07 pm

Re: the next match

Post by Martin Thoresen » Mon Jul 05, 2010 10:22 pm

S.Taylor wrote: And NOW!
Houdini MUST win the remaining 4 games to win.
That, my friend is not gonna happen. :D

pacifist

Re: the next match

Post by pacifist » Mon Jul 05, 2010 10:42 pm

S.Taylor wrote: I also don't understand how Rybka allowed itself to play 14...Qf5, but a few moves later, when a set of rooks were exchanged, i felt that it had benefitted from the sequence, and had a good game, and needn't worry about those technicalities.
Well 16.g3 gave Rybka something to bite with a possible f4 break. Same as in the exchange variation of the Spanish when black pushes c5 to c4 when white plays b3 before being able to block by c4. In the spanidh black has the 2 bishops as compensation also. Here white should play g3 only if f4 can follow . Anyway, there is a difference between knowing the plan and being able to follow it :). Tactics have a great impact and should be taken in consideration at every point :).

Post Reply