|
|
|
Novice NookDan Heisman
Translate this page
|
Attack it with Something Worth Less Quote of the Month: What if he just attacks it with a …? In chess the most important skill is analysis. The single most important issue within analysis is safety (tactics). There are more issues to safety than just losing material and checkmate; one is the effect of a threat to lose material, addressed in both It's Not Really Winning a Tempo and The Three Types of Reasonable Threats. A related issue that often arises in students' games is that they too often allow a piece to be undesirably (and often unnecessarily) attacked by pieces of lesser value. Suppose the opponent makes a move and you ask "Is it Safe?" You could first use Counting (see The Two Types of Counting Problems) to determine if any opponent's piece is en prise or whether it allows any sequence of captures on any squares to win material. If so, it's likely not safe, but if Counting finds no danger, that still does not necessarily mean that the move is safe. The next question that should be considered in determining the safety of the move is "What happens if the piece that moved is now just attacked by another piece?" If there are possible attacks, then those attacks can be categorized into three cases:
Each case has its own issues. In the first, since the attacker is worth more, then in order for the attacked piece to be unsafe, it must be at least inadequately defended (see The Inadequately Guarded Piece): White to Play
Assume White plays 1.Qe2. Since the knight on e6 is barely adequately defended (one attacker vs one defender), the queen attack from e2-e6 on the knight makes it unsafe and 2.Bxe6 becomes a threat. In Case Two the pieces are the same value, so similar safety issue as in Case One apply – an attack on a barely adequately defended piece would now create more attackers than defenders, but there are additional issues, such as "activating" previous attackers that were worth more than the attacked piece, or whether a fair trade for equally valued pieces would favor the attacker or the defender. But these are not the issues I would like to address this month. I would like to concentrate on the third case: AWL. AWL is the most forcing case, and thus the one you should consider first, assuming such an attack is possible. In The Five Ways to Make Piece Safe, I discussed the various ways to make an unsafe piece safe. One of the most popular, but not always the best, was to guard the piece. But when AWL is involved, guarding is almost completely eliminated as a defense. This makes attacking something with something worth less not only the primary way to make a piece move, but also is the primary starting point for trapping a piece. Summarizing, if there are no Counting issues on the square to which the piece moved, nor any other affected square, then you should next ask: "What if I just attack the piece that moved with something worth less?" Resolving this simple issue before more complex, time-consuming multi-square motifs such as double attacks, pins, and removal of the guard, is efficient analysis. Those multi-square possibilities must later be checked too, but to skip the "AWL?" step often results in missing important safety information. Of course, don't forget to also ask "What are all the things that move does?" Because even if the moved piece is not safe, it could have made a discovered threat, so capturing or trapping it may not be sufficient. However, if the moved piece creates a threat, then one defense to the threat may be capturing the attacking piece. I have seen many weaker players see a threat and make a passive defense when they could have just captured the attacking piece and not only eliminated the threat that way, but also win material. As noted above, "AWL" can result in not only winning a tempo, but also trapped pieces and removal of the guard. Let's consider an example of each: Black to Play after 1.e4 d6 2.Bb5+
If you give this position to an engine, it will tell you that Black's choice is very close between all four reasonable replies (2…Nc6, 2…Nd7, 2…Bd7, and 2…c6). Most masters would probably play the simple 2…c6, getting in a "free" helpful move (see It's Not Really Winning A Tempo). At 25 ply, the Houdini engine slightly prefers 2…c6. Similar ideas occur in most openings. When a beginner asks why they can't find 1.e4 e6 2.Bc4 in their opening book it is because of the AWL 2…d5, "winning a tempo." Although 2.Bc4 fights for the key square d5, and it is thus an initial candidate move (see Initial and Final Candidate Moves), it must be rejected because of the AWL idea. Look for rejection of your candidate move because of AWL in many positions. Black to Play after 1.Qa4+
In this position the "normal" way of saving the double-attacked knight would be 1…c6, discovering the defense on the a5-knight by the black queen. Note, if you are not familiar with this pattern, put this position into your personalized study book (see the latter part of A Case for Time Management), because this queen defense of a knight – usually after a check – appears quite often. However, in the diagram 1…c6 does not work because of the removal of the guard 2.b4 and afterward, with c6 falling hard, Black's game is in tatters. Instead, Black should play the pseudo-sacrifice 1…Bd7! using AWL. Even strong players might reject 1…Bd7 superficially, because it loses the knight to 2.Qxa5, but that rejection would be a Quiescence Error. The AWL idea does work as 1…Bd7 2.Qxa5?? allows Black to trap the queen with 2…b6. So instead White must retreat the queen and Black holds on. I have used the following position hundreds of times to illustrate AWL: Black to play after 1.Bf4
On the previous move White had several alternatives to save the pinned knight on e5, but chose 1.Bf4. When I ask intermediate students what they would play here for Black, most consider either piling up on the pinned knight with moves like 1…Nbd7 or attacking the guarding bishop with something like 1…Nh5 (often failing to "see" when visualizing ahead that this fails to 2.Qxh5). However, almost every strong player instinctively knows that it is far more effective to look for AWL before giving thought to a move like 1…Nh5 (even if it were safe). Additionally, these strong players are much more likely to not fall victim to quiescence errors and thus are able to quickly spot the "unsafe" AWL idea 1…g5! White should now muddy things with 2.0-0 gxf4 3.Re1, but after 3…Be6 Houdini evaluates Black is still a little better. However, if Black had tried other first moves, he would just be down a pawn for no compensation. An important principle to support AWL is Don't place your pieces on squares where they can easily be attacked by pieces of lesser value. This attack is often – but not always! – favorable to the attacker (see Strong Principles vs. Important Principles). I have frequently stated that in games between two players rated less than 1500 FIDE/USCF, there is about a ninety-nine percent chance of a basic tactic opportunity for one side (frequently missed!). Not understanding AWL is also common. I would estimate that a great majority of games between players of that level feature an AWL error: one or both players place pieces where they can easily and effectively be attacked by a lesser piece. For example, suppose there is only one open file on the board; e.g., 1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d4 Here Black may have been aware of the famous trap 3…Nxe4? 4.Qe2 Nf6?? 5.Nc6+ winning the queen and thus feels it is worthwhile to try the similar (but in this case counterproductive) trap. 5…Qe7?! hoping for a blunder like 6.Bd3?? Nc3+ winning the queen (no, I don't call this Hope Chess although I could; see Question Three in Novice Nook Quiz) When I tell students that this idea is not worthwhile, they often fail to understand why. In order to leave the beginner ranks and become an intermediate, you can't make "bad" traps and hope your opponent will fall into them, and the reason this one is bad is primarily AWL: White to Play after 5…Qe7?!
White will simply play 6.Be2 to stop the easily-thwarted threat (see The Three Types of Reasonable Threats), followed by 0-0 and Re1 (AWL!). Black's queen is not only blocking his king's bishop, but will itself be in danger of pin tactics along the e-file. The fact that White can so easily occupy the open e-file with a rook is a main reason why the queen should not be placed there. If Black was more familiar with AWL, he would realize why e7 is an awkward spot for the queen, trap or not. In his book Grandmaster Secrets: Openings, GM Soltis states that the opening principle with the fewest exceptions is Develop your rook to the same file where you opponent develops his queen – even if the file is currently not open, I might add. It does not take much insight to realize that developing a rook in this manner is simply an extension of AWL! One reason to play …a6 in the Sicilian Defense is to keep a white knight out of b5, which not only protects d6, but provides the black queen less vulnerability on c7. In his classic book My 60 Memorable Games, Bobby Fischer shows how Mikhail Tal's omission in their game at Bled 1961 allowed a "simple" AWL-based attack for his white forces. This was a clash between future and past World Champions, as Tal had lost his defense of the title to Mikhail Botvinnik earlier that year: Fischer – Tal 1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 e6 Also stopping a later Nd5. 5.Nc3 Qc7 6.g3 With the latent AWL threat of Bf4. Now Fischer writes, "A perfectly legitimate treatment which Botvinnik labeled a ‘very cunning and well-masked idea.' Actually no trap is intended. It becomes one only by virtue of Tal's reply." 6…Nc6? Fischer continues, "Probably the losing move! Tal looked worried immediately after having made it, but I'm not sure he was convinced he had really been careless. Correct is 6…a6 [stopping Nb5-DH] 7.Bg2 Nf6 8.0-0, etc." White to Play
7.N4b5 AWL One. 7…Qb8 Fischer gives 7…Qa5 Bd2 as worse for Black and, forty years later, Houdini strongly agrees. 8.Bf4 AWL Two. 8…Ne5 Fischer states, "Tal took a long time on this risky reply. The alternative 8…e5 9.Bg5 a6 10.Bxf6 (not 10.Na3 b5 11.Bxf6 b4!) 10…axb5 (not 10…gxf6 11.Na3 b5 12.Nd5) 11.Bg5 gives a clear advantage." However, Houdini thinks that 8…Ne5? is the losing move – not Fischer's suggested culprit 6…Nc6 – and the super-strong engine calculates that 8…e5 is forced. It agrees with Fischer's analysis of 8…e5, but differs with 10…gxf6 as best, and after 11.Na3 Bxa3 12.bxa3 Ne7 White is only slightly better. 9.Be2! Dominating the knights – a good positional idea to know. After this move both Fischer and Houdini think White is winning. Fischer won in forty-seven moves. So if Tal can fall into an AWL problem, any of us can. But when he does it, it is news; so you don't want to allow your pieces to be pushed around too often. An instructive example from a recent student game: Black to Play
In this position Black played 1…Bg4?!, which looks like an AWL, because it develops the bishop and attacks the queen. However, this attack is ineffective, since White can reply with his own AWL, 2.f3, which not only saves the queen, but also attacks the bishop on g4, reinforces e4, and protects the e3-bishop from harassment via a later …Ng4 (similar to the function of f3 in the Sicilian Dragon's Yugoslav Variation). So 1…Bg4?! is counterproductive for Black and 2.f3 productive for White, and instead Black should have played 1…c5, when later threats of …c4 and …a6 and …b5 to trap the unfortunate bishop on b3 would give Black the initiative. White in turn also erred by replying with the natural (but incorrect) 2.Qd2. This type of play is very common in the opening, so understanding this position should help with many similar positions. Are there exceptions to AWL? Absolutely! Sometimes the attack is not favorable to the attacker. For example, in the Levitsky Attack 1.d4 d5 2.Bg5 White's idea is that neither pawn attack 2…f6 or 2…h6 is all that helpful for Black – the good they do is at least balanced by the harm, such as the weakening of the g6-square and the king's safety. Nevertheless, at 25 ply Houdini rates the break move 2…c5 as best (the evaluation is -0.04 pawns), but both 2…f6 and 2…h6, along with 2…Nc6, tie for 2nd-4th (0.00). In the Ruy Lopez, grandmasters and computers believe White is slightly better after enticing …a6 and …b5 than he is in the Italian Game without those moves. Specifically 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 b5 5.Bb3 is better for White than just 1.e4 e5 2.Nf3 Nc6 3.Bc4 even though Black is "ahead" the pawn tempi in the former continuation. There are several reasons for this – the bishop is more vulnerable on d5 and the black pawns are vulnerable to attacks with a4. However, the bottom line is that allowing the bishop to be attacked and driven back by 3…a6 is not at all detrimental. Therefore, you are much less likely to see the Italian at the grandmaster level, and most strong players pass up the opportunity to play 4…b5 in the Ruy Lopez and more often choose 4…Nf6. Defending AWL From the defender's point of view, if the opponent does use AWL, then the right response is usually to just move the attacked piece to a safe square. Counterattacking by attacking another piece is very dangerous; it can be a great resource, but often it just creates unfavorable complications. For example, if the opponent's Piece A attacks your Piece X, then counterattacking with your Piece Y against opponent Piece B often fails, because Piece B can be saved, while creating a threat, say, against Piece Z. At that point A is still attacking X and now B (or something) is also attacking Z and there may be no defense to these two simultaneous threats. A simple example should suffice: Black to Play
Here the pawn on e5 (A) is attacking the knight on f6 (X) with an AWL. Instead of properly and simply moving the knight, Black decides to erroneously counterattack with 1…Bb4+?? where the bishop (Y) attacks the king (B). Here White can get out of check with 2.c3 where the c3-pawn (C) attacks the bishop (Y), still also retaining his attack against the knight on f6. Black has no defense against both threats A against X and C against Y and loses a piece. I see this type of error quite frequently from weaker players who don't want to submit by simply moving their attacked piece, but want to show how "clever" they are by finding a counterattack. Sometimes this is a good idea, but often it is either a waste of time or, worse, a terrible mistake. As noted in The Five Ways to Make a Piece Safe, counterattacking is a great weapon, but it usually creates complications and thus is very dangerous. This implies that counterattack can be very effective when you are losing and looking to complicate, but used with great caution when you are winning and want the position to be simple. Question I've read that ninety percent of results come from five percent of effort. What do you think that five percent is for improving chess? Answer Assuming this even applies to chess, I would answer, "Analyze as much as possible with very strong players." If you analyze your games with these strong players, so much the better. This is my top tip for chess improvement. Number two is to consistently play long time control games – trying to use almost all your time each game – against somewhat stronger players and analyze the games with them afterward. You should also play some slightly weaker players to learn "technique" and maintain your confidence. Have the perseverance to absorb losses and learn from your mistakes. Number three is to consistently check to see if a move (your opponent's or your candidates) is safe by asking "Can that move be defeated by a check, capture, or threat that cannot be met?" If you do this for all analytical positions, I call that "Real Chess." If you don't, I call that "Hope Chess." Question I am rated 1925; what percentage of study time should be devoted to openings? Answer About five percent; unless you are preparing for an event where you know your opponent's openings, then your opening study time should be higher. That is enough to both look up your games (including speed games) to see what you would do differently next time if someone played the same moves and also occasionally learn new tabiyas. © 2012 ChessCafe.com. All Rights Reserved. A PDF file of this month's column, along with all previous columns, is available in the ChessCafe.com Archives. Dan welcomes readers' questions; he is a full-time instructor on the ICC as Phillytutor. Yes, I have a question for Dan! Comment on this month's column via our Contact Page! Pertinent responses will be posted below daily. |
Purchases from our
|
|
|
|
[ChessCafe
Home Page] [Book
Review] [Columnists] © 2012 BrainGamz, Inc. All Rights Reserved. |
|
