The bottleneck
I don't know where this post will be heading, so you are forewarned!
The method I concocted is totally empyric. It came about by a mix of elimination, trial and error, and observation. Hence it is not optimized in any way, nor scientific. This means that I'm always on the lookout for improvements and more understanding.
The core of my method is generalization of concepts. This means that I have to conceptualize what is happening in a position, and once the concepts are clear, the concepts must be generalized, so that they become useful for other positions as well. Both the creation of concepts and the generalization are done by system 2. The next thing is to ensure that system 1 has absorbed the matter.
Attention is the spotlight, and both system 2 and system 1 are following the attention. Because that is their very nature.
The bottleneck
Because system 2 plays such big role AND is very slow, it is the bottleneck of the process. Any optimization of the process will focus on the efficient use of system 2.
There are of course a few tricks we can use. These tricks are mainly revolving around what we can learn from better players. Most of us have a small (or big) library of chess books. After my house had been on fire this library was lost, but it has been replaced by a digital library at an astonishing rate.
But at the end of the day, you will have to think for yourself. So I want to focus on the process again, and see whether there are any chances of optimizing it myself.
Robert showed us an interesting position some while ago, and I noticed a potential hidden gem in it. Besides the gem he already showed us.
Although I'm avoiding endgames at the moment, I will make an exception here because it might be a good way to get a precise observation of how the process conceptualisation works.
 |
| Black to move |
This position might arise from the position that Robert provided. Stockfish estimates it as a draw. But why?
Fiddling around a bit with the position learns me that the black king must prevent the invasion of bishop Bf6. Hence the black king must stay in contact with f6. With Ke6 or Kf7. Everything else loses.
What is the problem with Bf6? It forces g4 for black. And that creates holes where the white king can invade.
So already two concepts become clear:
- prevent invasion of the bishop with your king
- prevent invasion of the white king by keeping your pawns abreast.
- white must protect his pawn
- white must prevent the promotion of the d pawn
- black can use its f pawn to threat promotion
- the white pawn can't stop the promotion of the f pawn because it can't be traded
The conceptualization is the most time consuming part. Once done, you need to generalize it and to absorb the patterns. But that is relatively easy.
ReplyDeleteThis means that in order to make fast progress, you can only do that by generating the concepts fast. How can I improve on that?
My rating went down from 1785 to 1680 steadily from 2008 to about a year ago (despite continuous training!). Since a year it is on the rise again, currently being around 1750. I'm invited to play in the second team the coming season. I started in the fifth team two years ago. I look forward to it! I had quite a few neat scalps in the internal competition lately.
ReplyDeleteEven better, I start to feel confident, every now and then!
There are different areas that need improvement. I even know already how my approach must look like. I have not decided on the priorities, though.
ReplyDeleteIf I see this post for instance, I get the feeling that a little effort on endgames would be justified and would yield benefits easy.
Yet I must not divide my attention too much. In Januari I will retire, then matters will become easier, I suppose.
The openings seem to be under control, lately. I don't lose games due to a bad opening anymore.
In July I have a nine day tournament. I will use that to get a clearer picture of my priorities.
I watch a lot of videos about AI and the problems it will face to get better. LLM's cannot conceptualize problems. Hence it starts to hallucinate whenever they are pushed beyond their limits.
ReplyDeleteScientist have a lot of ideas about conceptualizing problems. I'm looking whether I can borrow a few ideas.
While reading your post and examining the position as you described it, I became aware of another concept in it that is often referenced: a weak color complex.
ReplyDeleteBecause White has a Bishop capable of attacking anything sitting on the black squares, it would seem logical for Black to place his pawns on the white squares so that the Bishop can never attack them. The problem is that White can then use his King to attack them from the unprotected black squares. Following this logic, of course Black can protect them – from white squares, but not necessarily from black squares. After all, he has two extra pawns and they are currently in the phalanx position (side-by-side). If one of them is attacked, Black can just advance it and then defend the rearmost pawn. It would seem illogical to advance one of the pawns in the phalanx so that both pawns are on black squares; that would just make it easier for White to attack them with the King and Bishop.
The missing concept is Zugzwang. Given good coordination between the White King and Bishop, White can force a situation in which the Black King must abandon the defense of the rearmost pawn IFF both pawns are on white squares. If in the interim White has secured his single pawn on a black square (defending it with the Bishop only if necessary), then Black could lose both pawns. Because White can always gain the opposition by “passing” with a Bishop move, Black will eventually lose all of his pawns and White will promote his own pawn. This scenario does not have to be calculated concretely to the end; it is sufficient to grasp the concept as a guide.
There is an alternative based on Capablanca’s basic pawn strategy: move in such a way as to hold up two pawns with one pawn. This suggests that the white pawn should be on g3 with the two black pawns on g4/f5, and then the Bishop can prevent the promotion of the d5 pawn while White re-positions his King to attack the rearmost of the black pawn duo. Black cannot abandon the two pawns unless the white pawn is eliminated first.
Black must look for an alternate strategy while trying to avoid creating a weak square complex that allows the White King to invade.
The only viable alternative is to (somehow) get rid of White’s single pawn, even if it costs all three black pawns. White would then be unable to win with just the Bishop.
This generalized concept is applicable in a wide variety of positions and can play a planning/evaluative role at an earlier stage in the game. The consequences of trading down to a Bishop for two pawns must carefully be considered, and not solely on the basis of the relative material imbalance.
The observed concepts are remarkably universal.
ReplyDeleteA bishop wants:
-to invade
-prevent invasion
-clear the path for its own pawns
-stop the movement of alien pawns
-conquer alien pawns
A king wants
-the very same
A pawn wants:
-to promote
-to keep invaders out
-not become a target
-to assist its brother
-cooperate with its pawn and king
What can be achieved in an actual position depends. But the universal goals remain the basis of the evaluation.