Thursday, May 22, 2008

Up a pawn and drawn Shredder

As White against Deep Shredder I blundered my Queen's Knight for a pawn early in the game but then, always up a pawn, I managed to keep the game even throughout a queen endgame and queen swap to a rook endgame with my 3 against his 2 pawns. Only at the very end Deep Shredder snatched my last pawn, and all material was traded down to the naked kings. Playing more than forty moves against the silicon monster with the reserve of only one pawn and not losing, this may be called a success.

diagram
White to move.

In this position, I had forced the swap of the dark-squared bishops on a3. In odds games, trading pieces is a general strategy, but it should be done very carefully. In most cases it is better first to complete development and block counterplay.

The move h3 looks very natural in this position. Do you see what happens after the trade on f3? Comments are welcome.

Labels: , , ,

Wednesday, May 21, 2008

Brilliant endgame tactics of Deep Shredder

In my most recent training match as White against Deep Shredder 11 with the odds of knight and move, I have run into a wrecked pawn structure that allowed the software to put me under heavy pressure. I was forced to give the piece back and arrived in an endgame equal in material but with a serious pawn weakness.

diagram

Black to move.

I thought I could defend my position but then Deep Shredder surprised me with the brilliant strike 1.-Rg3! He gives a full rook for no checkmate, but after 2.fxg3 Qa2+ I must give my rook for nothing more than a tempo or I get checkmated. I am forced to trade Queens, but before I can do it, Deep Shredder's Queen snatches another pawn, and with his f- and g-pawns against my h-pawn, the endgame is lost.

Very instructive! It tells me that my training strategy of playing at odds against the full power of Deep Shredder is exactly what I need. A weaker engine probably would have agreed to draw by repetition earlier in the game, and I would have missed this tactical combination.

Another point in favour of my method: the E=mc2 of chess, that is, the balance between my material advantage and the "energetic" advantage of Deep Shredder.

Labels: , , , ,

Saturday, May 17, 2008

Play it home, Sam

I've discovered a very useful new training method for me, and I strongly recommend it to every chess player struggling for improvement. I probably never would have had the idea, but two days ago, I've been a full piece up and managed to blunder the game to a loss at last. I strongly felt that I have to cope with this loss, and this gave me the idea to play games against Deep Shredder 11 with the odds of the Queen's Knight, running it on my 2.4 GHz Intel Core Duo processor with 10 seconds calculation time per move by which Shredder reaches a full depth analysis of 12 to 17 ply.

I use a normal chess tournament board, a mechanical clock and I write down my moves plus time used just as in a slow over the board game. My time control is 60 minutes for 36 moves and 30 minutes for the rest of the game. After my move, I walk over to the computer, enter it, and wait for the reply.

In my first game, I played as White, with the odds of knight and move, and I chose the French defense (1. e4 e6 2. d4), Shredder responded 2.-d5, and I exchanged pawns in order to get an open file where I could trade the major pieces later. This plan worked quite well, I always was up between 3 and 4 pawn units until move 40. The critical moment emerged in this situation after the 42th move of Black.

diagram

White to move.

After 43. h4?, the Queen escaped to f5 and the light squares, where I have a weakness, paradoxically with my bishop. But it emerges that this bishop is just a better pawn, one that I have been forced to sacrifice earlier for activity. But now I gave my activity back for nothing. I squeezed my brain for seven minutes and found nothing better than 44. Qe5 which I should have played a move earlier. Now the Queen escaped with check on d3 and gained a pawn. I also gained a pawn with check, and ten moves later the game was drawn by perpetual check, with only two minutes left on my clock.

Instead, 43. Qe5 would have forced a queen exchange, along with the sacrifice of another pawn. But after that, the white King intrudes the position on the dark squares and wins easily. The move 44. - Qg3 which I feared was no option for Black because the queen loses too much time to get on the back rank.

Very satisfied with my performance

Deep Shredder has a rating of about 2800, and given the relatively short calculation time of ten seconds per move, it may be only 2500 or so which is still grandmaster strength. The odds of knight equals about 650 Elo, therefore a safe estimate of my performance in this game is around 1850.

Labels: , , , ,

Tuesday, May 13, 2008

My critical opening position in the Sicilian

My favourite plan against the Sicilian is 3. Bf1-b5, pinning a Nc6 or giving check. I stick to this plan since a strong (2200) player has admitted after the game (equal until I blundered in the endgame) that he hates Bb5 in the Sicilian. Another reason is Deep Shredder, considering it the best move against 2.-Nc6.

The critical line for my plan is thus 1. e4 c5 2. Nf3 e6 where I can neither pin a knight nor give a check.

The main purpose of e6 is a fast counterstrike d5 against White's center pawn, and if Black manages to reach this goal, he has an easy and equal game. White therefore has no time for slow moves such as c3, preparing d4 with equal pawn force in the center. If he wants to keep an advantage, he must prevent d5. The best move is 3.Nc3, keeping control over d5 and developing a second piece against none of Black.

Black now can prevent 4. Bb5 by 3.- a6, but the most played move is 3. - Nc6, allowing my thematic 4. Bb5. If Black attacks it by a6, I trade bishop against knight, and Black's double pawn and his lag in development give me a good compensation for the bishop pair.

diagram

White to move.

But Black can double his knights, playing 4. - Nge7. This is a critical position because after BxN NxB, Black has no double pawn and replaced his strong Nc6, and I have no sufficient compensation for the bishop pair. White therefore is forced to take action, and the best choice is 5. d4, of course.

That is, in this position I have to change my general plan, opening the position because keeping it closed will lose my advantage, allowing Black to equalize easily.

Labels: , , , ,

Thursday, May 08, 2008

Knights attacking the Bishop pair

In general, I like to keep my bishop pair. But, a paradox, I also like openings that give away the bishop pair very early, such as the Nimzo-Indian as Black or the Sicilian with Bb5 as White. Thus, I have to change my mind and learn how to attack the opponent's bishop pair with my strong knights.

In my next two games, I'll be White, and with my bad results against the Sicilian I am going to look at this position and its resources for both sides.

diagram

White to move.

Black has the "advantage" of the bishop pair which is only minimal at the moment with all pawns still on the board, and the disadvantage of a double pawn. He also has lost tempo by attacking the bishop with a6.

White therefore is ahead in development and has got rid of his light-squared bishop which would be a bad one in this position. In contrast, the knight pair has very good perspectives with two main plans: 1) pushing e4-e5 and moving Nb1-c3/d2-e4 and 2) pushing a2-a4-a5 and moving Nb1-a4/b2-c4. Deep Shredder prefers the second plan, rating it +0.51 pawn units at 18 ply depth, but the first plan (+0.36) is nearly as good.

When asked which side to take I would take White in this position. Black has not a very easy game at the moment. He must try to open the position, trade pawns and keep his bishop pair in the endgame. White, on the other hand, has two excellent plans for an active piece play in a pawn-clogged position where the black bishops play only a minor role.

Labels: , , , ,

Wednesday, May 07, 2008

Put off my Queen's Indian plan

Yesterday, as Black against a -60 rated opponent at our city championship, I was surprised by 1. d4 Nf6 2. Bf4. After this first-time encounter I was put off my Queen's Indian plan because I didn't want to spend too much time considering whether the Queen's Indian can be used against this type of position, and I decided to play a "normal" Queen's Pawn opening like this one:

diagram

Black to move.

Black has a bit more space than in the Queen's Gambit declined, but the pawn-clogged center is not what I like, and really I was not able to turn it into a favourable position. The game was drawn after 21 moves, and the Deep Shredder analysis shows that neither side made any mistake, and the advantage for White remained always within 0 and 0.7 pawn units during the whole game. I accepted his draw offer in a dead-looking position.

Instead, I could have stuck with my Queen's Indian plan against his buildup which could have resulted in such a position:

diagram

Black to move.

According to Deep Shredder, this is not better than what I have played, but it fits my general plan of controlling the light-squared center and diagonal with my pieces and holding back my center pawns, thus preventing the intrusion of his pieces.

In this position, Black can make use of the exposed Bishop and gain a tempo with 7.-Nd5 8. Bh2 f5 9. Bd3 Nf6, and Black has a very good kingside play, keeping control over the strong point e4. I guess that it would have been much easier to get an advantage against my opponent this way, because this type of position may have been less familiar to him and more familiar to me.

Labels: , ,

Monday, May 05, 2008

Considering the Queen's Indian big fianchetto

diagram

In my few first steps of the Queen's Indian defense, I always have used the classical fianchetto Bb7. GMs often play Ba6 with good success. I think I should adopt this plan, but what plan is it anyway? I should know before playing it. So let's have a try.

Obviously, this move gains a tempo because a pawn is hanging, delaying White's fianchetto Bg2. But its main purpose is the retention of the e2 pawn, x-raying both the pawn and the Rf1 after the castling of White.

By Qa4, White can force Black to go back to the normal fianchetto Bb7, but with other moves, Black will keep the Bishop there and block the light squares with c6 and d5.

This transposes the Queen's Indian to a Catalan position that gives Black a good and at least equal play. I think this is a good plan and I am going to play it next time when White will be avoiding to play Nc3 to which I have my Nimzo-Indian.

Labels: , ,