Not logged inRybka Chess Community Forum
Forum RybkaChess Help Search Register Login
Up Topic The Rybka Lounge / Computer Chess / Vas has busted the Kings Gambit- Chessbase
- - By Mark Eldridge (****) [gb] Date 2012-04-02 10:58
Intersting artical.
Parent - - By Werewolf (*****) [gb] Date 2012-04-02 11:15
My hunch would be that THIS is Chessbase's April Fool's Joke.
Parent - - By Werewolf (*****) [gb] Date 2012-04-02 11:17
(And I seriously hope it is an April Fool's joke or we will never see Rybka 5)
Parent - By Fritz202 (**) [us] Date 2012-04-02 11:20
Me to!
Parent - - By Vempele (Silver) [fi] Date 2012-04-02 11:22

> My hunch would be that THIS is Chessbase's April Fool's Joke.


It's dated April 2nd. Or maybe it was just released late and the date is adjusted for time zone...
Parent - - By Dadi Jonsson (Silver) [is] Date 2012-04-02 11:33
...or:

"On March 31 the author of the Rybka program, Vasik Rajlich, and his family moved from Warsaw, Poland to a new appartment in Budapest, Hungary. The next day... Vas, kindly agreed to the following interview".
Parent - - By Nick (****) [gb] Date 2012-04-02 11:34
And the article appeared on the Chessbase site a few minutes before midnight on April 1 (American Samoa).
Parent - By Banned for Life (Gold) Date 2012-04-02 16:01
Vas is late for everything else he does. Why not for April Fools jokes as well? :smile:
Parent - By Mark Eldridge (****) [gb] Date 2012-04-02 12:05
Did not think of that, that does make sense what you said.
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-02 12:31
Assuming this isn't an April Fool's joke this is a total bombshell of an article.

I hope it isn't a joke!
Parent - - By Permanent Brain (*****) Date 2012-04-02 14:09
It is a test if knowledgeable chess player's common sense can withstand. :lol: Sometimes, computer results (faked or true) need to be "waved aside"... After 3.Be2, White has to deal with 3...Qh4+. Also, after 3...d5 4.exd5 as proposed in the article, again 4...Qh4+.

Therefore, 3.Nf3 just can't be wrong. I may be wrong but that is how I think about this, for now.
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-02 14:35
On further reflection I now realize it HAS to be a joke.  But I am bitterly disappointed!
Parent - - By Fritz202 (**) [us] Date 2012-04-02 15:03
What does your book say about Be2? No way its the best move or even the best draw move.
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-02 16:25
I'm at work and have no book access right now, but my impression was surprise that Be2 was the only possible drawing move.  But then, what my book says means nearly nothing if someone has done massive analysis of the KG.  Nonetheless, the writer of this article (whom I have since ferreted out) clearly knows a lot about what is plausible in the opening and what isn't.  A number of conceptual statements in there are true or skirt on the edge of truth, which is what makes it so credible.
Parent - By MasterGreg (**) [us] Date 2012-04-02 16:29 Edited 2012-04-02 16:31
I thought that Bobby Fischer "busted" the King's Gambit after
E4     E5
F4     ExF
N-F3  D6 !
Fischer states that this is a "High class, waiting Move."
Greg
http://www.academicchess.org/images/pdf/chessgames/fischerbust.pdf
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-02 16:59
You know, the real question is...if the conditions described were real--3000 high-end cores managed by the cluster, four months of highly focused work following intelligent algorithms--what could actually be accomplished?  Seems to me one hell of a lot of progress could be made, just as is described. 

But of course, you'd need a federal grant.  Something like that would cost $500,000 to buy the basic components, the power bill would be huge, and of course you'd need a place to house and cool all this equipment, another massive expense.  Call it over a million dollars altogether, conservatively speaking.  All that to solve the King's Gambit?  Won't happen, unless you're an eccentric oil sheikh or Russian oligarch!

But here's the thing: if someone WERE to solve one opening that would be equivalent to a manned moon landing in computer chess terms.  It would be a giant step toward solving chess.  Proof of concept would be a tremendous achievement.
Parent - - By Uly (Gold) [mx] Date 2012-04-02 20:16

> But of course, you'd need a federal grant.  Something like that would cost $500,000 to buy the basic components


Would it be possible to solve 0.1% of the King's Gambit with $500? Or the escalation doesn't work like that?

Also,this thread is missing an actual link to the article:

http://www.chessbase.com/newsdetail.asp?newsid=8047

What I found fun is that when Googling this, this Rybka Forum thread appeared first, and the actual article appeared second :smile:

Also, while this is almost certainly a joke, it could also be the reason Vas was behind schedule for Rybka 5...
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-02 20:53
I think you either solve it or you don't solve it, so 0.1% is worthless.  But, if your question is "could this be solved through a distributed computing solution" then the answer is "maybe" though you'd need a lot of software to direct and develop the workload in addition to huge amounts of dedicated hardware.
Parent - - By Uly (Gold) [mx] Date 2012-04-02 21:03 Edited 2012-04-02 21:06

> But, if your question is "could this be solved through a distributed computing solution" then the answer is "maybe"


Yeah, the idea is gathering just 1000 people interested in solving the King Gambit, and then have each of them solving their 0.1% of the positions. If we're going on the basis of "4 months per solution", then we can halve the needed people, and have 500 people solving it in 8 months, or 250 people solving it in 16 months (taking into account the technological progress that would happen in that time, and assuming all the people upgrade their hardware accordingly). I was asking if it works like this, assuming each people has already a $500 value hardware (and taking into account overhead 250 people solving it in two years still sounds great).

Of course the most important part of this is having the algorithm that tags drawn positions as drawn, and won positions as won, accurately, so that most of them could be pruned and only a small fraction needs to be checked. If that algorithm is just fictional then we're back to square one.
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-02 21:56
Vas could create something like that I am sure.  The question is whether he'd participate in any way, and the answer is obviously not, for a lot of good reasons.

The basic concept is the same as Aquarium, only taken to a monstrous level.
Parent - - By keoki010 (Bronze) [us] Date 2012-04-02 22:42
FinalGen?????
Parent - By keoki010 (Bronze) [us] Date 2012-04-02 22:42
:lol:
Parent - By Uly (Gold) [mx] Date 2012-04-03 06:51
No, the concepts here of solving an opening and the FinalGen's approach are opposites. While one relies on having a solution with as few positions as possible, FinalGen focuses on building a solution that includes all positions as soon as possible, with side effects like not being able to solve positions where one side has more than two minor pieces (while the side effects of the other approach is that you can't be 100% certain that it doesn't have holes.)
Parent - By Permanent Brain (*****) Date 2012-04-02 18:02
I think 3.Be2 idea comes from Tartakower, from the famous tournament New York 1924. Tartakower played the KG. with 3.Be2 aganist Bogoljubow, Yates, Alekhine and Capablanca, and scored 2,5-1,5 from it. That's not something to easily argue with. But noone of his great opponents simply played 3...Qh4+ to put him to the test.

(Please keep in mind that my comments are from a patzer perspective. :lol: )
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-05 11:03
Forgot to reply to this.  I have the position after Be2 at 50.8% and 21.9% draws in my book, which doesn't tell you much.
Parent - By Geomusic (*****) Date 2012-04-06 07:14
I saw naka play Be2 the other day I forget if it was a 3 0 or 1 0 game. But after the move he said, "Best by test" lol! He won of course.
Parent - By koroljow (*) Date 2012-04-03 22:26
My reaction is exactly the same as yours.  I am so sorry....
Parent - By Master Om (*****) [in] Date 2012-04-03 01:31
AFAIK The Kings Bishops Gambit leads to Draw. Qhr+ is good for white!!.
Parent - - By Mark Eldridge (****) [gb] Date 2012-04-02 18:09
I agree fully. I assume we will no very soon.
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-02 18:45
Oh, we know now.  I was a little dense at first, but apparently this has been in the works for quite some time and I know who was behind it.
Parent - - By Mark Eldridge (****) [gb] Date 2012-04-02 19:10
Go on then tell us.
Parent - - By Nelson Hernandez (Silver) [us] Date 2012-04-02 20:47
I'm sure the guilty party will own up to it soon enough!  :grin:  It wasn't me, that's for sure!
Parent - By NATIONAL12 (Gold) [gb] Date 2012-04-02 23:00
Nick knows his time zones quite well.:smile:
Parent - By ernest (****) [fr] Date 2012-04-03 16:31

> I was a little dense at first


You must be joking... :evil:
Parent - - By Gambit-man (**) Date 2012-04-03 22:25
April Fool!!
Parent - - By googyi [hu] Date 2012-04-04 09:32
I also think this is April Fool...
The problem is with the draw positions. For them the +- 5.21 margin does not help at all they must be calculated until repetation/50 move role/etc. So the complexity is too much anyway..
I would be happy if this is true, but I can't believe it.
Parent - By turbojuice1122 (Gold) [us] Date 2012-04-04 10:23
The main problem is that 3000 CPU running for months is nowhere near enough computer power to take on something like this.
Parent - - By Dragon Knight (*) Date 2012-04-04 00:58
Indeed interesting article! I do not believe as an April Fool Joke.
Guys start-up the strongest chess engine you have (coz I have none!), please help me figure-out where following opening lines leads to...favor whom, white or black?

Game - !
1. e4 e5 2. f4 exf4 3. Nf3 g5 4. h4 g4 5. Ne5 Nf6 6. Bc4 d5 7. exd5 Nxd5 8. Qe2 Be7 9. Nc3 c6 10. d4 O-O 11. O-O {Opening Development Complete}
*

Game - 2
1. e4 e5 2. f4 exf4 3. Bc4 Nf6 4. Nc3 c6 5. Bb3 d5 6. exd5 Nxd5 7. Nf3 Be6 8. d4 g5 9. h4 Bb4 10. Bxd5 Qxd5 11. hxg5 Nd7 12. Qd3 O-O-O 13. Bxf4 Bf5 14. O-O-O Qa5 {Opening Development Complete}
*
It is my personal belief that above both lines hold equal-equal chances for white and black, both?
What do you say?
:)
Parent - - By Ed Collins (*) [us] Date 2012-04-05 02:19
Chessbase has already confirmed what most already knew... it is indeed an April Fools joke.

Houdini prefers White in the first game:

1. e4 e5 2. f4 exf4 3. Nf3 g5 4. h4 g4 5. Ne5 Nf6 6. Bc4 d5 7. exd5 Nxd5 8. Qe2 Be7 9. Nc3 c6 10. d4 O-O 11. O-O

+1.04  f3 gxf3 g3 Qe4 Nd7 Bxd5 cxd5 Nxd5 f5 Nxe7+ Qxe7 Qf4 Nxe5 dxe5 Re8 Qxg3+ Kf7 b3 b6 Kh1 Qxe5 Bf4 Qg7 Qxg7+ Kxg7 Rg1+ Kf7 Rad1 Be6 Kg2 Rad8 c4 Bc8 Kf2 Rxd1 Rxd1 Kf6 Rd6+ Kg7

Both Kings are slightly exposed but White is better developed and can regain the  pawn easily.

In the second game the chances look equal:

1. e4 e5 2. f4 exf4 3. Bc4 Nf6 4. Nc3 c6 5. Bb3 d5 6. exd5 Nxd5 7. Nf3 Be6 8. d4 g5 9. h4 Bb4 10. Bxd5 Qxd5 11. hxg5 Nd7 12. Qd3 O-O-O 13. Bxf4 Bf5 14. O-O-O Qa5

+0.20  Qe3 Rhe8 Ne5 Nxe5 Bxe5 Bxc3 Qxc3 Qxa2 Rhe1 Re6 Re3 Rd5 b3 c5 b4 Ra6 Rd2 c4 Rf2 Be6 Kd2 Ra4 Ref3 Rd7 Kc1 Rd5 Rf6 Ra3 Qb2 c3 Qb1 Qxb1+ Kxb1
Parent - - By Dragon Knight (*) Date 2012-04-05 06:50 Edited 2012-04-05 08:12
hi Ed Collins,

Thanks for your inputs!
The first one I played as black and second as a white, both of us managed to draw both the games by three fold repeatition. :lol:
here is how it went (to the best re-collect of my memories, or may be one or two move can differ in sequence)
1. e4 e5 2. f4 exf4 3. Nf3 g5 4. h4 g4 5. Ne5 Nf6 6. Bc4 d5 7. exd5 Nxd5 8. Qe2 Be7 9. Nc3 c6 10. d4 O-O 11. O-O f3 12. gxf3 g3 13. Qe4 Nd7 14. Bh6 N7f6 15. Qd3 Nh5 16. f4 Bd6 17. Bg5 f6 18. Nxd5 cxd5 19. Bxd5+ Kg7 20. Qf3 Qe8 21. Bf7 Rxf7 22. Qxh5 fxe5 23. Qh6+ Kg8 24. Qxd6 exd4 25. Rf3 Qe6 26. Qxe6 Bxe6 27. Rxg3 Rg7 28. Kf2 Rc8 29. c3 dxc3 30. Rxc3 Rxc3 31. bxc3 Rd7 32. Bf6 Rd2+ 33. Ke3 Rg2 34. a4 Kf7 35. Bd4 a6 36. Ke4 Bd7 37. Ke5 Bc6 38. f5 Rg4 39. Kd6 h5 40. Re1 Re4 41. Rxe4 Bxe4 42. Ke5 Bc6 43. a5 Bb5 44. f6 ...{Both of us were stuborn, just kept moving bishop here-n-there endless, finally draw by three fold repeatition out of frustration} 1/2-1/2

Most of the time we analyze GMs game with two-or-three good human players, and find-out something really good and interesting.
With my limited knowledge of computer, Can anyone suggest easy and inexpensive or free tools for analyzing games played by GMs on computer?
Can some one show me how to analyze on computer?
Tks!
Parent - - By googyi [hu] Date 2012-04-05 09:14
Hi,

The most easy is to start with this free stuff:
http://www.chess.com/download/view/portable-winboard-45b--houdini-15a

Just extract it to any folder, no installation needed.
Start winboard exe, select "Play against a chess engine". First engine should be houdini, you can select it easily, as it is preconfigured.

Select "File" -> "Load Game" -> select any pgn file you wish with GM games in it. (You can find pgn files on chessbase.com)
After you opened a pgn file, you can select any games from it. There will be a list of them in another window.
After you selected the game you wish to analyze select "Mode" -> "Analyses mode"
Then the engine will start to calculate the best move in the given position.
0.00 means equal postion, +1.00 means nearly a pawn advantage for the side to move, etc.

Googyi
Parent - - By Dragon Knight (*) Date 2012-04-05 16:45
Hi googyi

many thanks for the link and suggestions.
tks,
Parent - - By Uly (Gold) [mx] Date 2012-04-05 22:14
The problem with Free Houdini-15a is that it forgets analysis after hash entries are overwritten or you unload the engine. To avoid this I suggest using Critter 1.4 with Session File On.
Parent - - By Dragon Knight (*) Date 2012-04-06 05:33
Uly

Thanks.
I am barely interested in computer chess, plus my computer knowledge is limited to internet surfing.
If I start using houdini, may be I will keep shouting houdini what next?...houdini what next?  at every move  rather than check-mate,  I mean I will forget what I know.
I am happy the way I am and the way I play chess. I may drop idea using computer for analysis util someone teaches me in details on my computer. I am looking for person who know computer chess in details in my friend circle, even a kid.
Thanking you everyone for help and response.
:lol:
Parent - - By Homayoun_Sohrabi_M.D. (***) [us] Date 2012-04-06 07:03
Hello Dragon Knight,

Of course you do not want Houdini to tell you what move to make next but analysis of your game with a chess engine AFTER the game is a good idea.    That way the engine can point out to you where you or your opponent missed some tactical shots.   Also the engine can quickly show you the result of other alternate lines of play that you may have been thinking about during your game.

You don't need any computer knowledge ( or a kid:lol:) to do this.   Go to www.tripplehappy.com, download and you are ready to go in 5 minutes.   Everything you need is packaged for you in one free product.   You can also go to www.shredderchess.com, then to the download section and download the "demo" shredder chess classic and use it for one month, no obligations.   Very easy to use and again everything you need is right there.    It of course has "help" files which are clearly written and will answer any questions you may have.  

www.geniusprophecy.com has some tutorials on all of this stuff in the beginner and intermediate sections, but you don't really need any of that, it's all very simple.    What gets confusing is when you try to figure out which product is the best because there are many to choose from but they all basically do the same thing.   My advice is just to stick with one that you like, learn to use it well, and that's all you'll need.  

Finally, there is a book "how to use computers to improve your chess" by Kongsted.   It's 10 years old, I know, but it's very well written and I think you'll find it an interesting read.
Parent - By Dragon Knight (*) Date 2012-04-07 02:35 Edited 2012-04-07 02:37
Homayoun_Sohrabi_M.D,

Thanks.
I got No install Tarrasch Chess GUI for Windows with Rybka 2.3.2a and komodo3 from www.triplehappy.com, and I am happy too!
It's easy -n- fun. Enough exercise for my little brain. I am getting ito it, I guess.
Thanks!
:lol:
- By dragon49 (****) [us] Date 2012-04-06 16:43
I am not often deceived, but this one got me.  Good job Chessbase!
Up Topic The Rybka Lounge / Computer Chess / Vas has busted the Kings Gambit- Chessbase

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill