IM Rajlich Claims the King's Gambit is Solved by Rybka Cluster

Assuming it's not a late april's fool joke by Chessbase, this is a very surprising claim by IM Rajlich, the author of Rybka.

In short, he says 1.e4 e5 2.f4 exf4 is solved by the Rybka cluster powered by IBM's super computer, and what's even more surprising is that he says there's only a single move that draws for white which is the quite "unorthodox" 3. Be2, all other moves winning for black by force.

He also says Fischer was actually very accurate in most of his famous "refutation" of the KG.

I wonder what a lot of KG fans out there, including our IM Pruess is going to do now .

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

OMG!

Has it been played in a real game? I'd like to see how it played out

I am a huge fan of the KG, but I am really not a strong player, although I gained many nice wins against many opponents (here on chess.com, OTB in rated games, OTB in friendly games against better players than I am, OTB in rapid [one Croatian Candidate Master is my achievement]). I just write my opinion: every gambit is refuted as long as your opponent knows the refutation.

If you are not playing against someone rated 2100+, I think you can try almost whatever opening you want to. KG is in my opinion one of the best openings against an unprepared opposition, perhaps the best against "lesser" opponent.

I don't think it is, as it's published on April 2nd, and it would be a little unreasonable to joke with the latest developments of one of their products and a world famous tech company.

Calm down everyone. This seems fake if you look carefully at the pictures.
First of all this is how IBM Watson look like. http://www.almaden.ibm.com/images/featured/watson/WatsonPower7.jpg
They don't show a picture using this machine. Instead of this they show a picture of a guy who is using like 10 pcs and 5 monitors. Not too impressing...
Lastly the guy is doing the whole stuff in his room and using ordinary fans to cool the pcs. I assume if it was real it would have been done in a professional enviroment with proper cooling. The whole thing seems quite dubious.

That is the picture of a very older version of the Rybka Cluster, which was already in circulation in the Rybka forums and chess sites for a few years. I don't think Rajlich actually sat in front of the IBM computer and did the analysis manually at any point, they probably have developed a software optimized for this specific analysis, which runs in their cluster, and then it distributes its operations to the IBM super computer. That is what I got from the text. But it's clear that it is not the current version of the cluster, and definitely not the IBM super computer itself.

We know for a fact that a 300-core Rybka cluster does exist and it was rented by Topalov for his preparations against Anand, together with the latest Rybka engine which wasn't public. However, I think this news item was the first one to mention it was connected with IBM Power 7 for the purpose of "extreme" opening analysis.

I say proof or didn't happen and those pictures are definately not the ones they are talking about

OK, after a little confusion, the discussions in the rybka forum indicate this was in fact a late april's fool joke (they have caught that the interview "took place" on April 1st according to the article), but it's definitely cheating to make an April's fool joke on April 2nd! This might be the second most serious cheating controversy involving Rybka and Mr. Rajlich.

guesso 1 - me 0 .

philidor_position

Great comment!  I'm still laughing.

The interview was April 1. ChessBase always tries something clever. If this is not their April Fool's joke, identify the story that is.

Yeah OK they got me eventhough they cheated a little. But what's the fun in April's fool if you're not going to fall for anything?

Exactly.  Before following the link to read I scrolled down until I found the people saying "oh, so it was a joke"

Firtz 5 is just as well suited as Rybka to "solve" an opening.  The speed of calculation and storage capacity for a proof are the issues I believe, not superior middlegame evaluations heh.

Although they do try and make it convincing in the article.

Now I am ticked, I am a KGer, and well a Bishops gambiter, mad at that. Its probably April Fools

Little slow, Jetfighter.  Read the previous comments :P It was a joke.

chessbase had quality april fools jokes in some previous years, this year rather disappointing.

Solving King's Gambit in the sense that you mean isn't that far removed from solving chess as a whole.

Really? I thought this was a quite good one.  Just rather bad form to post it on April 2.

Gonna play 3.Be2 from now on then lol