Convekta Logo
Chess Software Sourcebook
Join Chess Reviews<>
Convekta Logo
Multivariation analysis in Chess Assistant 7 explored!
keywords: Chess Assistant, multivariation, infinite analysis, Ponomariov, Georgiev, Informant, Blokh, Flash
Maxim Blokh
Monday, August 25, 2003

This article discusses a couple of interesting features of the new infinite analysis mode in Chess Assistant 7.  We will look at using this mode to examine a game published in Informant #80. The game in question was an interesting, complicated, sharp struggle, and was fought between the young 15-year old GM Ruslan Ponomariov (UKR) and GM Kiril Georgiev (BUL) at the Istanbul (Turkey) Olympiad in 2000. You can download a copy of the annotated game from this link. Incidentally, since this article was written originally, Convekta created a flash version, which you can view here.

In this article, we will use the Tiger engine in infinite, multivariation analysis mode. To start the analysis,  press the Ctrl-Space hotkey or use the Infinite analysis icon on the toolbar.

The screen below appears. Select Multivariation (1 engine) in the Scheme drop-down list and set all options identical to the picture below:


Press Ok. You will see different move candidates offered by the engine. You can compare these with moves from the Hugebase and CAP trees, which are placed under the chess board and game variations.

If you want, play through the game quickly, using the arrow keys or game animation features of CA 7. We can see that the end result was a typical drawn endgame with opposite color bishops. After the move 24 Bf6, Tiger confirms this, demonstrating that the position is drawn. If we go back to the position before move 24, we see that Tigers evaluation is that black has the advantage. Indeed, Ponomariov considers 24... Bf6? as a blunder, the move 24... Rad3?! as dubious and recommends 24... Re8! instead. To check this, press the Ctrl-Space hotkey and switch on the Consider moves only in tree panel option.

Tiger confirms the analysis by Ponomariev. Let's try to search for another possibilities. Press Ctrl-Space and switch on Ignore moves in tree panel option. Tiger immediately offers the very strong variation 24... Rc8 25. Be2 Bc3 26. Rf1 Kf8-/+.

Now the situation is clear. Black must avoid exchanging the rooks, and instead utilize his development advantage, and the poor position of the white rooks. Rc8 prevents white from placing the bishop on c4, and the black bishop at c3 prevents the white rooks from becoming active. Now the pawn e6 is unprotected and will be consumed by the black king on its way to the center of the board. Black has good winning chances.

Now switch on the Consider all moves option again and continue the analysis. Go to the position before 21. Ba3. Ponomariev marks the move with ! sign. However Tiger offers the  simple 21. dxe7!?Qxe7 22. Qa4 exf3 23. Bb2+/=  . Make these moves on the board. Tiger confirms its opinion. It seems that in the heat of the fight, Georgiev overestimated the threats to his king and underestimated the strength of promoting the pawn on e6. It is not clear why this variation is not considered by Ponomariev.  So the move 21. Ba3  should not be marked with a ! but with a ?.

So after 20... Rc3, the evaluation is +/=. Can black play stronger here? Tiger offers the simple 20... Bxa1? 21. dxe7 Qxe7 22. Ba3 Qf6 and  =/+ . Make these moves on the board.

At this point Tiger, after some thinking,  realizes that after 23. e7+! Rf7 24. Be2 Kh7 25. b6! the evaluation is +/= (and possibly even +/-  ) in view of the dual threats of supporting the  promotion of the e7 pawn on a4-e8 diagonal,  in conjunction with the deflection of bishop on a1.

So Tiger's early evaluation of the position after move 20. Bxa1 was wrong. It was unable to calculate this long variation due to its sacrificial nature. But note that Ponomariev was able to do this at the board!

So let's give Tiger another chance. We already know that the counterattacking move 20... Rc3?! made in the game can lead to a +/= evaluation. So let's try one more move offered by Tiger among the 3 best moves  (but not 20... Rc3 and 20... Bxa1). It is 20... Qxd6!? . Make the moves on the board 20... Qxd6 21. Ba3 . Tiger offers 21... Qb6!?., as well as 21... Qd5 22. Qxd5 Nxd5 23. Rad1 (evaluated with by Ponomariev) Bc3!?. After careful analysis we can see that in both cases black is OK. So 20... Qxd6!?= should be played,  instead of 20... Rc3?!.

Let's now go back to the beginning of the game and consider the critical position that occurred after the courageous and ingenious 17... d5! by Ponomariev. Even a GM can become confused after this move, but not Tiger!  It offers the natural moves 18. Nxd5 and 18. exd5  as well as the very nice 18. Rd1!?= . After 18... d4 19. Ba3! Rxc3 20. Qxc3 dxc3 21. Rxd8 Rxd8 22. Bxe7 Re8 23. Bxf6 Bxf6 24. Rc1 Rc8 25. exf5 gxf5 26. Bxb7 Rc5 27. Bc6=, the endgame is draw.

So the novelty 16...Rc8!? played in the game in the variation 11...Nf6 12. Bf3 c6 13. b5 cxd5 14. cxd5 h6 15. Ne6 Bxe6 16. dxe6 promises equal chances for black. Of course, not every chess player likes to play the positions that resulted in this game. So let's search for other opening possibilities. The best tool for this is the  Hugebase and CAP trees already present underneath the board.

Go to the position after white's 11. Ng5. Tiger prefers the Kasparov maneuver 11... Nf4 with advantage +0.4 pawn as compared with second and third moves! Look at the tree. The move 11... Nf4!?  occurred about 100 times (though it's more rare than11... Nf6).

Lets make the move. The main theoretical answer is 12. Bxf4. Tiger at first offers other alternatives, but at a depth of 16 chooses 12. Bxf4. 

So continue 12. Bxf4 exf4 13. Rc1 . Here Tiger, after some thinking, offers the reliable 13...Bf6 and after 14. Ne6 Bxe6 15. dxe6 black can choose between the sharp 15... Bxc3 16. Rxc3 fxe4 17. Bg4, that is choice of many players (as well as Tiger) and the reliable 15...c6!? (with one game in the tree) 16. Bd3!? (16. Bf3 Be5! 17. b5 Rc8=) Be5!=. Analysis shows that the 15...c6!? continuation looks quite interesting.

So here we conclude our analysis of this complex game. I would like to thank both Grandmasters for this interesting game, as well as Convekta and Christophe Theron, for their great programs.


ChessAssistant is a trademark of ChessOK
Syndication available through rss.xml
Click on my name to send me e-mail (must have javascript on)