More about preconditions
Let's start with a typical Aurora attack in the Colle Zukertort. This is how it might work:
 |
| Diagram 1. White to move |
r4rk1/pbqn1ppp/1p1bpn2/2ppN3/3P1P2/1P1BP3/PBPN2PP/R2Q1RK1 w - - 2 4
Remember what we try to accomplish: finding the preconditions. The preconditions are related to the LoA landscape. Let me first map the obvious LoAs (lines of attack).
 |
| Diagram 2. White to move. The LoAs |
- blacks Nf6 messes with two LoAs
- The long diagonal is clogged up
The first thing to look at is can we clear the long diagonal with tempo?
1.Nxd7
- 1. ... Qxd7 (Aurora Australis)
- 1. ... Nxd7 (Aurora Borealis)
Let's assume the Aurora Australis
2.dxc5
- 2. ... bxc5
- 2. ... Bxc5
- 2. ... Be7 (the lesser evil, giving up a pawn)
Let's assume that 2. ... Bxc5 is played. Then we get the following position after 1.Nxd7 Qxd7 2.dxc5 Bxc5
 |
| Diagram 3. White to move |
r4rk1/pb1q1ppp/1p2pn2/2bp4/5P2/1P1BP3/PBPN2PP/R2Q1RK1 w - - 0 6
3.Bxf6 getting rid of the defender of the LoAs
3. ... Bxe3+ in between move that doesn't change the result
4.Kh1 gxf6
5.Bxh7+ Kxh7
6.Qh5+ Kg7
7.Qg4+ Kh7
 |
| Diagram 4 |
r4r2/pb1q1p1k/1p2pp2/3p4/5PQ1/1P2b3/P1PN2PP/R4R1K w - - 4 11
And mate in 5 follows.
This provides a baseline. It shows the LoAs. Every move of the tree of analysis is related to the LoAs. So the LoA landscape acts as a pair of scissors to prune the tree of analysis.
It is a matter of fiddling around to get all the details of the preconditions.Stockfish immediately sounds the alarm when a precondition is not met. By changing the begin position you will find out what every piece does in relation to the LoA landscape. So far I unearthed the following areas of attention:
- Black can try b5-c4 in order to chase away Bd3
- Qc7 and Nc6 can be dangerous after cxd4 exd4 Nb4 forcing white to give up Bd3
- Prevent that black gets the upperhand on e4, where he can clog up the diagonal d3-h7
- Prevent the pawnbreak e5 by maintaining the upperhand on e5
- Whenever black gets the chance to play cxd4, he prevents that you can open the long diagonal at will
- Be7 instead of Bd6 makes that dxc5 cannot be played with tempo
- Bc8 in stead of Bb7 can have an influence on the pivot square h3, in some variations
- In some lines you need to play Bf6 instead of Bxg7 in order to prevent that black can open an escape route for his king by playing f6
UPDATE
Not all LoAs are aiming at the same targets.. Some of them are aiming at a sitting duck (king, weak pawn, invasion square), some of them are aiming at the killbox or a defender of a LoA. This gives different roles to the attackers.
Take for instance the potential roles of Bb2:
- in an old school rule way it is looking at its own pawn
- it is aiming to sacrifice itself against g7, which is part of the killbox around the black king
- it is aiming to sacrifice itself on f6, thus preventing black from playing f7-f6, thus creating an escape route for his king
- it is aiming at trading itself for Nf6, which is the defender of h7
- it is aiming at its own pawn at e5, which is a powerful attacking asset (it chases away Nf6)
Furthermore, Bb2 has two adjacent LoAs:
 |
| Diagram 5. White to move |
These adjacent lines of attack play no role at the moment. But when in certain variations where the log diagonal is blocked, these adjacent LoAs can play a role. The diagonal a3-f8 might help to prevent the black king from escaping. While the diagonal c1-h6 might play a role when black plays g6, thus weakening the dark squares around his king.
What is the role of Bd3?
Usually it will pry open the killbox by saccing itself against h7, luring the king further away from an escape route. A lot of the skirmishes around this LoA are:
- trying to trade the bishop by Nc6-Nb4-Nxd3
- trying to chase the bishop away by b5-c4
- trying to block the LoA by Ne4
- trying to block the LoA by f5
- trying to block the LoA by g6
What does all this mean? The preparation of an attack AKA the Vukovic gap is no longer Terra Incognita. It is about LoAs and the effect of pieces on the LoAs. This means that we have an objective method to judge the effectiveness of a move. All this must now be done in the study room by means of system 2.
ReplyDeleteThe question is, can we absorb the logic? That is the only way to make the calculation of long variations possible. We talk about pruning the tree of analysis in an objective manner and automate the logic.
UPDATE in blue
ReplyDeleteAfter reading your post, I realized I know virtually nothing about the Colle-Zukertort. All of my previous experience was in the Colle-Classical (also known in the USA as the Colle-Koltanowski). I had no knowledge of the Aurora Attack variations. So, as usual, I went spelunking on YouTube and found this informative and very helpful (for ME) video.
ReplyDeleteLINK: https://www.youtube.com/watch?v=vAC1C7pBBUg
Colle Zukertort System: Crush Strong Players with This Setup
Synopsis:
Chess Opening Full: Colle Zukertort.
The Aurora Attack 00:00:01
The Noah Ark Attack 00:14:44
The Knight c6 Variation 00:29:36
The Fight for e5 00:38:00
Fighting the Caro-Kann Structure 00:41:01
d6 & King's Indian Setup 00:47:23
Other Approaches 00:50:40
The Aurora Polaris Attack 00:58:17
The Colle Zukertort System is a flexible and powerful opening for White based on solid development and a strong central structure. Instead of entering heavy theory, White builds a simple setup with pawns on d4 and e3, bishops on d3 and b2, and knights supporting the center. Once the pieces are in place, White can launch dangerous kingside attacks, often leading to tactical opportunities against the enemy king. Because of its clear plans and attacking potential, the Colle Zukertort is a favorite weapon for players who want a reliable system that works against many different defenses.
AI-generated video summary:
Master this aggressive attacking weapon tailored for the Colle-Zukertort System. Learn to apply pressure and disrupt opponents by controlling key central squares. Explore tactical variations designed to put immediate strain on the opposing king.
END of synopsis
There are many variations, but I did perceive several “common” tactical shots among them. I was somewhat surprised to see how many different ways there are to prosecute the attack on the enemy king. On the surface (prior to unloading the Greek Gift sacrifice), it appears that the WBb2 LoA is hopelessly blocked, but with BPc5 juxtaposed against the WPd4, that is not the case.
On a broader topic: it demonstrates how thoroughly learning all of the vicissitudes of a specific variation/opening can lead to identification of all possible tactical patterns from a certain point forward to a definitive evaluation (if not mate).
That is the video I'm referring to indeed. It is going to be my main preparation for the tournament in July.
DeleteI'm trying to find a broader perspective, though. The essence is the LoA landscape as mental model for an attack of any kind. All patterns are related to getting the upperhand at certain LoAs. Not all LoAs have an equal function. Some just assist to annihilate a defender or to close an escape route.
In this post, the final mate is usually delivered by Queen and Rook, and sometimes Knight.
Have another look at diagram 2 in the post of March 8
ReplyDeletePART I:
DeleteI did take another look at Diagram 2:
FEN: r1b2rk1/ppqn1ppp/1n2p3/2ppP3/5P2/1P1BPR2/PBPN2PP/R2Q2K1 w - - 1 13
One of the “obvious” things is that Black misplayed the opening. He wasted two moves getting the BNb8 to b6, where it MIGHT support advancing the c5-pawn to c4 at some point (if more urgent matters do not arise against his king). Capturing WNe5 with BBd6 resulted in displacing the BNf6 to d7, wasting more tempi and giving White control of f6. The BRa8, BBc8 and BQc7 are blocked from entering the game in defense of the kingside. Black only has the BRf8 as a quasi-defense against the kingside attack. Lots of tempi loss.
Comparing White’s trumps: WBb2 backing up the WPe5 controlling f6. White’s sacrifice via Bxh7+ cracking open the Black kingside. The White Rook already on f3, ready to swing to h3 with check, opening the LoA for the White queen to get to h5. That’s the two White heavy pieces that are requisite for a decisive kingside attack, positioned to gain tempi. Black cannot escape forward and must retreat to g8 (otherwise he gets mated quickly). Farther down the line, White can bring the WNd2 into play by either advancing the e3-pawn or by routing it through f3. After clearing the back rank, White can reposition the WRa1 as needed to support the attack on the Black king if it tries to escape to the queenside. All those potential LoA bolstered by virtually all the White pieces auger well for the success of the Greek Gift sacrifice.
NONE OF THOSE “GENERAL CONSIDERATIONS” MEAN ANYTHING WITHOUT CONCRETE VISION OF HOW TO EXECUTE THE ATTACK. This type of position must be explored during training until it becomes second nature so that we can “SEE” it all the way through when playing a game.
It’s easy to “SEE” the move(s) from the given position to the first steppingstone position. 13. Bxh7+ Kxh7 (13… Kh8 Rh3 appears suicidal) 14. Rh3+ Kg8 (forced: 14… Kg6 15. Qh5#) 15. Qh5, reaching that first steppingstone position.
THIS IS WHERE KNOWLEDGE OF SPECIFIC TACTICAL MANEUVERS IN THIS TYPE OF POSITION HELPS DRIVE THE PRINCIPAL VARIATION FORWARD!
It is easy to get “lost” here in the multitude of unforced variations if we use trial and error as the means to looking ahead. Instead, look at the general situation FIRST.
Black has no resources to bring to the defense of his king position. Consequently, he must try to skedaddle over to the queenside where he has defenders—that is his only viable plan. Since he is faced with mate-in-1 (Qh7# or Qh8#) and has no defenders available, he must create an escape path via f7. He only has two alternatives: f6 or f5. As GM Tisdall noted, PICK THE ONE THAT SEEMS MOST LIKELY. We may pick the “wrong” alternative, but that’s okay: “stuff” happens.
IMHO, 15… f5 seems less likely to succeed given that it opens up more lines around the king, so I pick 15… f6 as the most likely defensive attempt. (I intend to ignore the 15...f5 alternative for reasons of brevity; this alternative should be explored as well.)
Run through the principal variation again to refresh memory: 13. Bxh7+ Kxh7 14. Rh3+ Kg8 15. Qh5 f6. Get that position firmly in the mind’s eye, visualizing the new positions of the pieces that have moved. (Look away from the board if necessary to avoid “contamination” of a retained actual image.)
What attacking resources does White have in this new position?
First, get the Black king moving by getting the White queen “up close and personal”: 16. Qh7+ Kf7 (blocking the action of the BRf8 on f6 and creating a pin on the g7 pawn, which limits the defensive options Black has on f6).
THIS IS THE SECOND STEPPINGSTONE POSITION. (Keep in mind that steppingstone positions are idiosyncratic to each person.)
White wants open LoAs, so it is “natural” to consider capturing on f6, forcing open more lines. So, 17. exf6 Nxf6 18. Bxf6 Kxf6. Eliminating defenders and opening lines must be good for White.
PART II:
DeleteTHIS IS THE THIRD STEPPINGSTONE POSITION.
The g7-square becomes the next PoP; bring the WRh3 to bear on it (and also on g6) with 19. Rg3. The threat (it’s always good to proceed using forcing threats!) is to capture with check, skewering the Black king and “attacking” that undefended BQc7.
What can Black do? If he tries to defend by bringing the vulnerable Black queen to f7 with 19… Qf7, he gets mated in two moves by 20. Qh4+ g5 (or 20… Kf5 21. Qg5#) 21. Qxg5#. Perhaps the best thing to do is to either skedaddle to e7 (not attractive after 20. Rxg7+ K moves 21. Rxc7) or at least protect his Queen by moving it to where it is defended, perhaps d7, so 19… Qd7. That leaves open a path to the queenside for the Black king with the e6 pawn as a defensive shield.
THIS IS THE FOURTH STEPPINGSTONE POSITION.
Following GM Aagaard’s advice, bring the last piece into the attack: either the e3 pawn or the WNd2 or the WRa1. The most “obvious” is 20. e4, threatening mate with 21. e5#. It’s time for Black to skedaddle with 20… Ke7, but after 21. Rxg7+ Kd6 22. Rxd7+ Bxd7, it’s time to evaluate the material (since no mate is in sight).
White has Queen+two pawns for Rook+Bishop. Additionally, BNb6 and BRa8 are somewhat out of play. At this point, it’s time to stop trying to calculate ahead: White has sufficient advantage to win the game from here—the rest is a matter of technique. (HA! HA! HA!)
Summarizing the principal variation (in case I screwed up the numbering above):
13.Bxh7+ Kxh7
14.Rh3+ Kg8
15.Qh5 f6
16.Qh7+ Kf7
17.exf6 Nxf6
18.Bxf6 Kxf6
19.Rg3 Qd7
20.e4 Ke7
21.Rxg7+ Kd6
22.Rxd7+ Bxd7
GM Stockfish can (and should) be used to explore the alternatives at each steppingstone position until all tactical maneuvers are committed to LTM.
PART III:
DeleteGoing back to an earlier branching point:
IMHO, 15… f5 seems less likely to succeed given that it opens up more lines around the king, so I pick 15… f6 as the most likely defensive attempt. (I intend to ignore the 15...f5 alternative for reasons of brevity; this alternative should be explored as well.)
At first, it seems that there is no real difference between 15… f6 and 15… f5. I overlooked that with WPe5, White can capture en passant after 15… f5 16. exf6 e.p. Unfortunately, that temporarily brings the Black knight into the defense on f6 after 16… Nxf6, but White can chop it off with 17. Bxf6 Rxf6. Now the Black rook is in a more active defensive position, changing the principal variation.
THIS IS A STEPPINGSTONE POSITION.
White can now set up the attack on g7 without Black being able to prevent it. 18. Qh8+ Kf7 19. Rg3.
If 19… Rg6 (making g6 B.A.D. [1:1]) 20. Qh5 pinning the BRg6 and Black has no other piece to add to the defense of BRg6. White is ahead in material and still has attacking chances.
If 19… Ke7 (trying to run), he loses the Black queen: 20. Rxg7+ Kd6 (20… Rf7 21. Rxf7+ Kxf7 22. Qh7+ Kf8 23. Qxc7 winning) 21. Rxc7 winning the Black queen.
If 19… g6 or 19… g5, 20. Qh7+ Ke8 21. Qxc7 winning the Black queen.
We do not have to work out the technical details following these considerations. I ASSUME sufficient technique to win with a significant material advantage and the Black pieces uncoordinated defensively to boot.
ASIDE: GM Stockfish opines that White can mate in 8 by NOT capturing en passant on f6 with 16. exf6 e.p.:
16.Nf3 Qd8 17.Ng5 Qxg5 18.fxg5 Rf6 19.exf6 Nxf6 20.gxf6 gxf6 21.Qh8+ Kf7 22.Rh7+ Kg6 23.Qg7#
It seems a little strange to toss the Black queen, Black rook and Black knight into the wood chipper trying to ward off mate. I guess delaying the inevitable makes sense to GM Stockfish.
I just heard GM Neiksans say that the difference between a casual grandmaster and super grandmaster is mainly determined by a difference in the speed of calculation.
ReplyDeleteWhat I can only translate into a difference in amount of absorbed relevant patterns. I suspected that from the beginning. Say, the year 2000, when I started my quest. The prove was already overwhelming, lately. But it is good to hear him say that.
The video of Aox showed that the statistical tipping point is the age of 12. But it added that new methods with the aid of a computer, might change that.
At present the training assistance that can be provided by current software is focused on transfer of declarative knowledge (“KNOW WHAT”) rather than development of skill as procedural knowledge (“KNOW HOW”). The underlying assumption is that if you acquire sufficient knowledge, then you will automagically have the corresponding skill to take advantage of that knowledge.
ReplyDeleteCurrent training software does not (cannot) analyze problems with missing patterns. “Missing” in the sense of absence of a specific pattern in the personal repertoire and “missing” in terms of connecting two or more patterns into a cogent whole as part of skill.
For example, the emphasis on tactical training is acquisition of knowledge of the common tactical themes/devices and checkmate sequences (including a name for each pattern). System 1 could care less about names but it does care about putting isolated pieces of knowledge together into larger, more abstract patterns, “SEEING” connections between apparently disconnected fragments of knowledge and triggering the “Aha!” emotion. Unfortunately, if you never experience the “putting together” of those disparate pieces of knowledge in novel circumstances through repetitious practice, then those knowledge fragments do not become integrated into usable higher level or broader patterns. Even exposure through practice does not guarantee that the requisite patterns will be triggered when needed in any specific circumstance.
Analysis at the move level (as per GM Stockfish) does not provide explicit pointers to the abstract patterns that must be “SEEN” as the basis of SKILL. If the trainee cannot “read between the lines” he will never acquire the requisite patterns.
The literature of cognitive science estimates that a Class A player utilizes approximately 500 "patterns", an Expert utilizes approximately 10,000 "patterns", and a Master utilizes approximately 300,000 "patterns" - with NO real definition of what constitutes a "pattern" except the FUNCTIONAL relationships between the pieces and squares in a given position. BTW, the range of patterns supposedly required for Master-level play is estimated at 10,000-100,0000 by Chase and Simon, a rather broad range. I personally do NOT believe there is such a wide disparity in SKILL based solely on the number of recognizable patterns.
The problem with those estimates is that no one has any idea of what those (supposedly independent; otherwise, how would you be able to count them?) patterns actually are. Are those patterns on the same level, such as pins, forks, skewers, etc. or are there higher level patterns (perhaps composed of those common named low-level tactical patterns PLUS other unnamed patterns)? For instance, GM Averbakh grounded his seminal work on advanced tactics on patterns of elementary contacts between pieces. That is comparable to considering programming to be nothing more than establishing contacts between bits in the computer hardware. That sort of knowledge is UNUSABLE for developing programming skills and also for developing pattern recognition SKILLS in human chess players.
Consider the Greek Gift pattern. Is it primarily recognition of the possibility of sacrificing a Bishop on h7, followed by an “attack” (by major pieces or is there more (MUCH more!) to it? Given the present discussion, I assert without proof that there is much, much more to it—but that is the identifying characteristic promulgated in the training literature!
Is Lasker’s double Bishop sacrifice a variant of the Greek Gift? If so, do we ASSUME we “know” how to set it up in myriad circumstances after we become “familiar” with the original Greek Gift sacrifice?
I focus on the logical patterns that constitute the battle of the LoAs. In this post, there are different LoAs. Each LoA has a different goal.
ReplyDeleteDepending on the goal, the "technique" differs. It are very much standard scenarios. The skill is geared around the standard scenarios at LoA level (system 1). Combining and coördinating the activities that are going on on two or more LoAs is more a task for system 2.
In order to be able to calculate without thinking (by seeing), the technique must be automated.
In diagram 2 of March 8th, my opponent had no clue about what was going on a LoA level. He didn't see them clear and coherent.
In order to win a chess game, you must lure your opponent into territory where you see more than your opponent. That is true at club level, but it is equally true at master and grandmaster level.
Most of the time, we are talking about mini skills. If you are beaten by a 7 year old girl, you are beaten by mini skills.
On the one hand, you need to be very precise is finding and mastering the mini skills you need, while on the other hand, you must explore new areas. That is why I opted for the Vukovic gap as main focus for my tournament preparation, and to put the endgame on the back burner.
ReplyDeleteSomething I found accidentally while looking for something else about the Colle-Zukertort scenarios:
ReplyDeleteThe Feynman Technique — Stop Memorizing, Start Understanding
LINK: https://www.youtube.com/watch?v=NYSe2Ln0Tf4
I was totally unfamiliar with Feynman’s approach to learning. It is universally applicable to any subject, including chess. It explains a lot about why we don't really acquire SKILL from all the effort we put into adult chess improvement. It differentiates KNOW WHAT (by recalling the applicable label to “regurgitate on command”) from KNOW HOW (understanding, correlation and application of principles) using the K.I.S.S. principle.
Try it for yourself.
Or maybe I'm just "full of $#it". It wouldn't be the first time someone observed that might be the reason my eyes are brown!
Lol. That is the essence of this blog. Spiralling and coming back to the same issues over and over again until I understand it. This is my blank sheet of paper. And you help me to make clear what I don't understand yet.
DeleteThere is no such thing as acquiring mini skills. They are a result of understanding. System 1 looks over the shoulder of system 2 and does its trick automagically.
Repetition is only needed when I didn't pay attention the first time.
Creating a logical narrative is telling the story to a beginner, who doesn't know anything about chess.
Plateauing is the arrogance to think that I know stuff while I not really understand it. Think about it. Speed comes from understanding. Because understanding prevents me from doing things that are irrelevant. Trying to solve irrelevant problems is what makes me slow.