Any thing new from Rybka 4?

[Sunny Singh]

When Rybka 3 came it changed the way the computer chess was played and it played Caro-kann very well and it was a revolution in the opening book making. Caro dominated opening book theory for very long. But wondring anything stored in Rybka 4 something like this or is it a just another product, a product with so called ELO game 50-500 gain? Rybka experts, Vas Felix, Uly and Shahar in particular. Please don't sya TBD or NO or find it yourself. You are the creater and testers of the products.

More than two years something should be different isn't it?

[Albert Silver]

When Rybka 3 came it changed the way the computer chess was played and it played Caro-kann very well and it was a revolution in the opening book making. Caro dominated opening book theory for very long. But wondring anything stored in Rybka 4 something like this or is it a just another product, a product with so called ELO game 50-500 gain? Rybka experts, Vas Felix, Uly and Shahar in particular. Please don't sya TBD or NO or find it yourself. You are the creater and testers of the products.

More than two years something should be different isn't it?

There is no way to answer this question yet. Even if one saw different evaluations for certain openings, how would you know whether they were better or worse? GM practice and evaluation is pretty much the only way so time will tell. As to Rybka 3 being the second coming, I think you are overlooking a couple of important facts. Though Rybka 3 was indeed a big improvement over Rybka 232a, by about 100 Elo, the biggest reason it became so popular wasn't its particular ability to analyze the Caro-Kan or any other opening, it is because beyond being the absolute strongest program around, it was also now distributed by Vas himself, ChessOK, AND Chessbase. This really helped get it in the hands of so many master and grandmaster players. As to GM commentary on Rybka 4, I can only give you a couple of snippets by Larry Kaufman when asked about it in the Rybka forums:

Question: Larry, could you comment briefly on the playing style of R4 relative to R3 from the human GM point of view?

Kaufman: In my opinion, the style is not so different. The evals are pretty close in most positions, with king safety and certain endgames being the most notable exceptions. Basically, Rybka 4 is faster and evaluates more accurately, occasionally avoiding serious eval errors. Perhaps R4 is a bit more aggressive, but I don't think the difference is as much as the difference between R3 and R3 Dynamic for example, in most situations.

Question: In which type of endgames does R4 perform better in your opinion?

Kaufman: One that caught my eye was two rooks, minor piece and pawn vs. queen, minor piece, and pawn. R3 had the queen and evaluated it as equal, whereas R4 had the rooks and scored it as a full pawn better. Sure enough, R4 won the game. I haven't done a systematic endgame comparison, my primary interest is opening play.