Do any strong players actually think white has a clear advantage after 8. Qf3? I have never seen this advocated anywhere else... most of my 2N books remark is unclear or plus for black thus stating why white should prefer different lines like Be2 Bd3 stuff.

I wouldn't recommend placing too much trust in Fritz's evaluations in such a gambit line.  Quite often in unbalanced positions Fritz assesses the position as being strongly in favour of one side, only for the assessment to drop towards equal when you push it into the recommended lines.

Against 8.Qf3 you also have 8...Rb8 and 8...Be7.

Fromper wrote on 03/02/11 at 18:44:00:


On page one I gave a game where I got an advantage against Fritz12; whether Black can improve is unclear. Well into the middlegame, White still has his pawn and Fritz 12 couldn't demonstrate any initiative for the pawn. White appears to have a slight plus against 8.Qf3 h6.

TN wrote on 02/27/11 at 13:21:36:


8.Qf3 or 8.Be2 are possible ways. As far as 8.Qf3 cxb5, look at earlier posts where I was able to document advantage White. It takes time and exhaustive analysis to prove each branch is favorable to White. The 6.Bb5ch Bd7 line, for instance, is very difficult to analyze to a clear plus just because Black has no pawn weaknesses after c6 & can restrain any ability of White to create a passed pawn for a long time. To "prove" a win would take dozens of additional precise moves.

Other lines of the TKD will be covered on other posts.

TN wrote on 02/27/11 at 13:21:36:


There is 'theoretical' refutation and there is practical refutation. A theoretical refutation is one where any advantage, no matter how small, with best play can be converted into a win even if it takes a 1000 moves of best play by both sides. A refutation by a Class D player may be finding checkmate in one move. All other practical wins lie in the middle.

A good illustration of a refutation of a Queen and pawn versus Queen ending is when I think it was IM Jerermy Silman (?) was a pawn up in a Queen and pawn versus Queen ending. The computer claimed that he had a checkmate in 83 moves. Silman made a move and the computer promptly claimed checkmate in 87 moves. Thus a theoretical win for a computer in a Queen and pawn ending is always favorable to a computer over a human and what constitutes a refutation depends entirely on whether you are referring to a computer refutation or a human refutation.

When you hear the phrase that White is on the better side of the draw, this has no meaning in a theoretical sense, but does so in a practical sense.

I think in the future that it will be generally regarded that White has a slight theoretical plus in the TKD (The first move advantage) but that the wins will take over 100 moves to prove in many examples. For example to prove a win in the line 5...Na5 6.Bb5ch Bd7, the line that it will take an extremely long time to win, White will win the heavy piece ending (the minors will oppose one another and they will eventually all be chopped off), but then White has to try to demonstrate in a heavy piece ending that the extra pawn matters.

The winnning procedure will consist of pushing the Queenside pawns slowly, rearranging the majors, eventually creating a passed pawn and then create a target on the Kingside. Then using threats on both sides of the board White will be able maybe 60-80 moves ahead to Queen a pawn.

The extra pawn is a theoretical win in this opening but an incredibly difficult win for a human player in the heavy piece ending but a simple practical win for a World Class computer.

sloughter wrote on 02/27/11 at 10:00:03:


I've always played 8...Nd4!? there; e.g.

[Event "corr ICCF thematic"]
[Date "1996.??.??"]
[White "Haas, B."]
[Black "Tait, J."]
[Result "1/2-1/2"]
[ECO "C57"]

1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. Ng5 Bc5 5. Bxf7+ Ke7 6. Bb3 Qf8 7. O-O d6 8. Nc3 Nd4 9. h3 c6 10. d3 Nd7 11. Ne2 Nxb3 12. axb3 Qf6 13. Nf3 Kf7 14. c3 Rf8 15. d4 Bb6 16. dxe5 dxe5 17. Ng3 Kg8 18. Nf5 Nc5 19. Bg5 Qe6 20. Be7 Nxe4 21. Bxf8 Qxf5 22. Bb4 Be6 23. Qc2 Qf4 24. Ba5 Bxh3 25. Ne1 Nd2 26. Bxb6 Bxg2 27. Ra4 e4 28. Bc7 Qg4 29. Qxd2 Bf3+ 1/2-1/2

A strange but interesting game.

Jan Gustafsson has played 4.Ng5 d5 5.exd5 Na5 6.Bb5+ c6 with Black today in the German Bundesliga and won the game. 

Perhaps in the training session with Svidler last week (see his blog) he has been convinced about Blacks chances, since on his DVD he had expressed some doubts if Black gets enough compensation.

Or he simply wanted to avoid the opponents preparation.

walkingterrapin wrote on 02/25/11 at 18:58:17:

Conversely I was under the impression that 4.Ng5! was doing pretty well at high levels- see the serious problems Black faces in the line 4...d5 5.exd5 Nd4.