After looking at Kaufman's coverage of the Benko, I decided to look at another sideline, the Zurich Variation of the Nimzo-Indian (4.Qc2 Nc6) to see how Kaufman covers this. Unfortunately this isn't one of the strong points of his work - he fails to consider Black's best moves in the key lines and doesn't in any way explain how to play the resulting middlegame positions as White, often just saying that White is better because of the bishop pair or similar at the end of a variation. If you own Palliser's Tango! book and play an opponent relying on Kaufman's repertoire book, you will equalise out of the opening and probably outplay the opponent in the resulting middlegame. 

I've started a thread on this line in the relevant section (http://www.chesspub.com/cgi-bin/yabb2/YaBB.pl?num=1345425246/3#3) and would be interested to see you back up your explication with some concrete variations.

First of all, he basically recommends to accept the pawn and then to push it. That is the beginner approach to the Benko. If you don't know theory, this is what you would do. Regardless if it is good or bad, this is exactly what the Benko-player is hoping for and this is the reason why this gambit became popular in the first place.

Second, just check his line. At one point of his analysis he plays c5-c4. This move is just as ridiculous as Bologan's f5. Yes, both c4 and f5 are thematic, but not at this very moment. The move f5 comes in the pawn endgame, where basically all heavy pieces got exchanged and the move c4 comes when a black knight can jump to c5 and to d3. You don't play these moves just whenever you get the chance.

I love Kaufman's comment on 12. Qb6. He gives a few moves including the pointless Qb3 for black and says "White is starting to consolidate his extra pawn". This shows how clueless he is. Consolidation doesn't get white anywhere. It is like consolidating the pawn e5 in the Berlin Defence. Black is perfectly fine with white having an extra pawn, as long as nothing else happens.

The fact that Carlsen and Nakamura have played 12. a4 doesn't mean much. Carlsen has played many lines, many suboptimal lines and even many obscure lines. Since modern players only need a good computer evaluation to give it a go, I am not surprised he chose it. Still, I am convinced that playing like this will get the readers of this book in trouble.

Back to the main topic. In the meantime I have checked a few other lines and I am shocked. If you compare the lines in the Kings Gambit with the new repertoire book from Chess Stars, it becomes clear that Kaufman has a problem. His line in the Kings Gambit is the same, except he recommends weaker moves. Also his line in the Evans Gambit is the result of the game Short-Nielsen, but the guys from Chess Stars improve on this game, so there is no need to deviate. Basically Kaufman with his Elo 2382 is too weak of a player himself to identify the spots in the computer candidate list where it makes sense to investigate deeper. He just lets the computer run for 15 min and then recommends the top line on the list. That is an amateur approach to chess openings and while it may work, it definitely can be improved upon.

If what Neric says is true, then it seems arrogant to make such scathing judgements against a player clearly stronger than oneself. ...f5 is a common theme in the Benko to undermine the centre and d5-pawn, as the study of any Benko book will demonstrate, and it's very rough to judge a player on the basis of a few games.

Where did Yermo say that?? LOL

Neric wrote on 08/15/12 at 07:46:06:


Both Carlsen and Nikamura tried this in the latest Biel tournament. Carlsen jumped right on one of the main idea to put a knight on b5 while Nikamura ( attemted to improve over Carlsen's game -- ChessVibes ) followed Kaufman a litle longer. Both games ended 1-0.

I haven't read the new book. In the old book Kaufman was fairly generous why he considered one side better or the position equal, not just claim a computer evaluation.

I too have played CC. I much prefer to use one engine and test lines back and forth until I get a decent feeling of the ideas of the computer. It usually refutes most of my own ideas convincingly. Only occasionally do I test position with a second engine.

I too play backgammon. It's a very different game due to the influence of the dice. Letting computer programs play eachother in critical chess positions is an idea, but not a very appealing one. I much prefer humans working with the computer and not let it run without guidance.

Neric wrote on 08/15/12 at 07:46:06:


How is this funny? He is frank about the assesment of two programs and a fairly strong player himself (with much chess program experience) who gives his own view. Shouldn't it be that way?