I have the feeling that we are really on to something. That focusing on the salient points slowly builds some feeling for salient points. But not every salient point is immediately visible for the mind's eye.
You need some logic to make them visible. The problem is, that each node initiates diversity. Like a tree fork from where new branches are sprouting.
I advocate to ignore the complexity that sprouts from the nodes in favor for the most simple and logical continuation until the big picture emerges. Only when the big picture is clear, you go back for working out the details of the nodes.
 |
| Black to move |
2r3k1/2r2ppp/1p1b1n2/pP1Bnq2/8/PP2B1P1/3R1PKP/1N1Q1R2 b - - 0 1
[solution]
There are three salient points here that aren't immediately visible before the mind's eye.
- Bd5 doesn't look B.A.D. (Barely Adequate Defended).
- Rd2 doesn't pop out as a critical defender
- The diagonal d5/g2 looks taboo for the black Queen
Logic must reveal these salient points and make them visible for the mind's eye. You can only know that this are salient points with hindsight.
The point of pressure that pops into my mind is c1. If we invade a rook there we attack a Queen AND we deprive the white knight from a defender. It is a double attack.
The simplest reaction white can have is to trade a rook that invades on c1. That is the most straightforward idea.
Of course we can come back later, and investigate other branches that sprout from this node. But for now we prune those branches in order to keep it simple.
The trade of two rooks against a Queen changes the position. The white queen had a double function: to guard c1 AND Bd5.
So the bishop on d5 has become B.A.D. (Barely Adequate Defended). And our attention is guided to that BAD bishop. A new salient point has become visible.
And from the BAD target our attention slips to the defender Rd2. Because of simple logic. And another new salient point has become visible. We cannot harass the defender. But we can trade the bishop and transform the BAD defended target into a NOT defended target.
As we know, LPDO. And only now a new salient point becomes visible: the double attack against the white King and rook.
Sofar, three new salient points have become visible:
- Target Bd5
- Defender Rd2
- Vulnerable diagonal d5/g2
Only now the big picture emerges. By zooming out. We got here by following the most simple variation. And since the most simple line works in its most simple form it is now time to look back at the nodes and see whether we can see more complicated branches which refute the whole idea.
Summary
- Start with the salient points that are already visible (by zooming out)
- Follow the most simple and straightforward logic while ignoring the branches that sprout from the nodes (by zooming in)
- See the new salient points that become visible along the road (by zooming out)
- Follow the logic until you win a piece (by zooming in)
- See the big picture (by zooming out)
- Go back to the first node and look for alternative branches that might refutate the most simple and straightforward line (zoom in)
- Lather, rinse, repeat
I consider the tree of scenarios to be something completely different than the tree of analysis. A standard scenario is something like:
ReplyDeleteSalient point reveals a target. Is there a defender? If yes, harass the defender.
Or, trade the target so the defender becomes the new target. Changing the target from B.A.D. (Barely Adequate Defended) to NOT defended.
The amount of standard scenarios is fairly limited.
In an ideal world, the salient points lead us to the standard scenario which leads to the next salient point, while both SEEing salient points and SEEing the standard scenarios are automated.
ReplyDeleteVERLICHTING!!
ReplyDeleteIn my comments since around 2015, I have been referencing (pushing) GM Jonathan Tisdall's analytical approach of variation processing — without being able to comprehend and articulate the guidance mechanism used to propagate the principal variation forward. This specific blog post is an enlightening example of what must be "SEEn" in order to consistently DO variation processing!
GM Tisdall in 1977 figured out exactly how many hours of study would be required to advance by one rating point. He then estimated how long it would take him to reach grandmaster level from his current rating level. It took him 18 years to become a grandmaster in 1995. Your odyssey of 23 years is about to reach the rocket stage: "BLASTOFF!"
Time to (finally!) consign GM Kotov's "thinking process" to the ashbin of history — based on logic.
I'll toss another 'word salad' regarding this post and my "Eureka!" moment, when I have more time. (I don't have the time at present because my special needs granddaughter is coming for a couple of days visit.)
You even start to talk dutch, now and then.
DeleteIt never hurts to learn a new language - especially when playing and studying chess!!
ReplyDeleteMy granddaughter is arriving later than planned, so I have a few minutes to write about this latest discovery.
The salient points are not necessarily points (squares) on the chessboard. They are also pertinent ideas or conclusions.
What I previously did NOT “SEE” was the connection between local areas of tension (such as B.A.D. squares, invasion squares, etc.) and variation processing. If we look for those “areas of interest” and the interconnections between them, we can more clearly SEE the path of the principal variation.
GM Tisdall gives the following advice [Improve YOUR Chess NOW, pp 20-21]:
Looking back at this version of the solution, it seems to have taken a vast amount of time and space. But the vast majority of the work was verbal, and in real-time, this happens very quickly. THE MOST IMPORTANT ASPECT OF THE SOLUTION WAS CONSTANTLY INTERPRETING THE RESULTS OF EACH SEARCH. One notes the dynamics of the position, recurring themes, key tactical ideas, the ideal posts for the pieces—factors not usually associated with the calculation of variations. Also, you will note that there were more variations covered in the second pass of the position, and they were calculated using shorter variations (since we did not, as Kotov did, try to solve it from the start position).
One question that naturally arises from the examination of this position is the ordering of variations. Would it not be even more efficient to tackle the variations that appear to be most difficult first? Then, if these turn out in our favour, we could go back and tidy up the inferior variations later? There may be occasions where this is true. I have a feeling that this might be an effective method when one is defending a tactical position. In the above example, when one is attacking, and more importantly, sacrificing, I think the method given is best.
One argument is that by examining the potentially weaker defences first, one can actually get a clearer picture of the hurdles to be surmounted. If great difficulties arise breaking down apparently inferior lines, one has to examine the entire sacrificial enterprise in a colder light. In the mean-time, one is GATHERING VITAL HINTS ABOUT THE MOTIFS IN THE POSITION that will usually be of great use in wending through the most difficult lines later. The assessment of lines for elimination, or to place them in some sort of hierarchy, is being done through an intuitive or abstract process. This is a far less exact science than calculation. It has value as a sorting tool, but not for passing absolute judgment.
Referring to the example position, I “SAW” the c1-square as a “B.A.D.” potential invasion square [2:2]. At a minimum, Black can force the exchange of two Rooks for the White Queen. I also “SAW” the B.A.D. d5-square [2:2; WRd2+WQd1 vs BNf6+BQf5. “Looking through obstacles, perhaps all the way to the edge of the board” is vital when looking for squares under tension.
It is important to NOT get “hung up” on deciding which local area to address first. Since the process is a series of “ZOOM IN, ZOOM OUT” explorations, it really does not matter which order we “SEE” the individual parts. As noted, the point of all this is to get an accurate “big picture” of the overall position.
"The salient points are not necessarily points (squares) on the chessboard. They are also pertinent ideas or conclusions."
ReplyDeleteThat is a very important point. I tried to describe that in the penultimate post. Somehow the salient cue has changed into the salient point. On the one hand we have the roles of the pieces like target, defender or attacker, and on the other hand we have the altering of the roles by a standard scenario. Like eliminating the defender or trading the target into a piece with different properties. Both the amount of roles and the amount of scenarios are fairly limited.
There are several ways to eliminate a defender, but again, just a limited amount. It is the combination of salient cues that make the numbers.
In the past I described 23 scenarios I'm going to have a fresh look at that old post, but I'm sure the amount is fairly limited.
ReplyDeleteTempo - Have you ever thought about creating the index of all the topic and group these into the categories to make possible to surf through your articles without the need to browse all few hundreds of these you have written so far?
ReplyDeleteI am wondering you must have some catalogue/index of the topics and articles you are writing about and publish here on the blog. It would be great to have something like table of content on the separate page, to be able to browse all the fantastic content you are sharing with us!
I sometimes think about cataloguing the 1293 posts and 9492 comments. And than I sigh and use the search function which works pretty well.
DeleteBut no worries, I will find the scenarios for you.
DeleteFWIW, in the current context [emphasis added], without comment:
ReplyDeleteGM Jacob Aagaard in Thinking Inside the Box
Chapter 7. What is Calculation?
Calculation = The process of finding what you do not "SEE" automatically. How is this done? Slow down the flow of variations running through your head and look for alternatives."
Chapter 8. The Calculation Process
The purpose of calculation is to aid decision making, NOT to SEE everything. Do not over-calculate; you risk playing too forcing chess, not allowing the opponent to make his own mistakes.
It is important to get control over your impulses and to respect your opponent. Do not be naive. Even if he has played poorly so far, he might still have an idea behind his latest move.
Use falsification to check if your sharp lines work and be wary of going too far.
End your lines with a look at the position, not with an emotional evaluation.
Resistance should trigger slowing down, just like so many other tactical obstacles.
Great calculation skills require a stubbornness and a level of unbroken concentration which is difficult to develop. Difficult is OK; you have greatness within.