## Posts

Showing posts from January, 2024

### Slowing down

So how can the word salad monster (wsm) and the vulture work together in harmony? Apparently perception is key for the vulture. The wsm wants to go on to the next problem since he wants to do 100 problems before dinner. Otherwise his scheme will not work. I noticed with clay modeling  that when you are patient enough, at a certain moment the mind's eye takes over. The clay remembers system 1 how the feminine characteristics look like, it is projected before the mind's eye, and your hands start to materialize them. There are two ways to improve. To invent new strategies (system 2) and to boost perception (system 1). In the first category I am making huge progress lately, but observation showed that I don't make progress with system 1. It is probably best to let system 2 blow off steam when you are busy with a problem, in order to let him invent every strategy he can think of. There might be something useful in it. Only when system 2 is ready, there is time for the vulture t

### Interference

Quite a lot of combinations seems to consist of short sequences. Somewhere, these sequences interfere with each other. It is necessary to see both the sequences and the interferences. Black to move 4bnk1/6p1/6qp/PN1pPp2/5P2/5N2/3Q3K/8 b - - 1 1  [ solution ] First sequence : 1. . . . Bxb5 LPDO: the knight drops off Second sequence: 2.Qxd5+ Ne6 3.Qxb5   Double attack: on both black's King and Bishop on b5 Both sequences 1 and 2 interfere at b5 Third sequence: 3. . . . Qh5+ 4.Kg2 Qxf3 5.Kxf3 Double attack: on King and Knight. Fourth sequence: 5. . . . Nd4 6.Ke3 Nxb5 Fourth sequence interferes with sequences 2 and 3 at b5 and f3 In an Xmind scheme: This gives an idea what you must achieve in the realm of visual perception. In order to find the move 2. . . . Ne6 you must already see the knight fork in the fourth sequence.

### First impression

I read the first three chapters of Betty Edwards book Drawing on the Right Side of the Brain. These are my first thoughts. Which may well change into the opposite overtime after digesting the material. I get a bit the feeling as if the left side of the brain (system 2) is treated as the enemy of the right side (system 1). I don't think that that is the case. They each have their own task and complement each other well. The reason why one system is dominant over the other at a certain moment is probably guided by efficiency. Brain resources are scarce, and generally it is cheaper to deploy system 2 when a task is not very often repeated. In order to ingrain a task in system 1, more energy is needed to make long term changes to the brain. Think of the priest who needed a cheat sheet to be reminded of the exact text of his daily prayer. The plasticity of the brain requires a certain amount of energy, and nature has build in a certain resistance against lighthearted changes of the bra

### Perception attempt 1

When logic and perception were intermingled, I couldn't decide on the best choice for the dataset with puzzles to use for training. Since both logic and perception have different needs. The constraints for visual perception puzzles are: The problems must be as simple as possible I should be able to see the solution immediately, but don't Subject is that tempo's are important I decided on the puzzles of ChessTempo: Only one move deep (should guarantee simplicity) Rating between 1700 and 1800 (should guarantee not easy to see) For instance this one: White to move 1r5k/5b1p/p2p1r2/5P1p/1q1N1QnP/2N3R1/1PP5/2KR4 w - - 1 1  [ solution ] This is my first attempt to focus on visual perception. Where visual can refer to the mind's eye. You will have to bear with me as long as nothing is crystallized out yet. Let's invent a method. I'm using Xmind. With that I'm able to take the chessboard away so I can focus on visualizing the concepts, without being distracted by t

### Rehabilitating the eye of the vulture

It took me 23 year to discriminate between the sense and nonsense in all the good advice which is so generously disseminated over the poor chess player who wants to improve. There is no such thing as bad advice, but there are certainly issues with a lack of context which renders the advice as worse than useless.  Now the cloak of confusion was finally lifted last year, immediately logic emerged as the nec plus ultra method to educate our hungry system 1. And I started to apply logic to every phase of the game, especially the phase between the opening and the killing of the hostile king who rests in his mean castle. It convinced me that logic was indeed profusely lacking in my approach to chess. However this will undoubtedly improve my chess, there remained some nagging questions in my head. Consistently fed by Robert. I'm always grateful when people are willing to disagree with me and don't refrain because of my lack of response. Was this the trick that child prodigies use to

### Clay logic?

When I found that system 2 has to act as a guide for system 1 by means of logic, I wondered how that would translate to other areas than chess. Since I have a firm belief that a method of educating system 1 should work in a whole lot of areas other than chess, where adults plateau for years. Yesterday I tried again to make a clay model with female characteristics. It wasn't clear at first where logic would fit in the story. But Robert provided a few links with knowledge about the differences between male and female faces. In my previous attempts, I just fiddled around until after 6 hours accidently a few feminine characteristics emerged. If system 2 doesn't guide you, everything happens by chance. After reading the links, system 2 had learned about 8 characteristics that are different in males and females. And so it guided system 1 by steering the attention along these characteristics. This is the result after 1.5 hours: The logic of the differences wasn't invented by me.

### Seeing the invisible

Robert said: " It's interesting that we can " SEE " the LoA on the surface, so to speak, while the tempi remain invisible. " I don't know whether it is possible to SEE  the tempi before the mind's eye. If it is possible, and if a child prodigy is able to do that, it would explain a lot. There are a lot (or maybe just a few) of sources from which extra tempi can arise. Two of them we have already found: B.A.D. (Barely Adequate Defended) pieces Multifunctional moves These sources can already be seen before the mind's eye with some effort. What we must try is to get rid of the effort. In the past I talked a lot about duplo moves, which are moves that attack two pieces with one move. I.e. fork, double attack, discovered attack, simultaneous attack, pin, skewer and X-ray attack. This generalisation was based on the pieces that are involved. It was an attempt to simplify matters. Now I'm trying to base matters on tempi, it might be useful to change ma

### Pruning the tree of analysis

System 2 must guide system 1. How do I know that so for sure? My good friend GM Bent Larsen learned me that the hard way . We must be very careful with the deployment of system 2 during a game. For the very reason that its activities are very slow, cost much mental energy and are error prone. We must prune the tree of analysis as much as possible. Furthermore we must delegate as much tasks as possible to system 1. The latter is what we want to accomplish in the study room. System 2 is not good in the parallel processing of tasks. Not to mention that it is totally unsuited for that. We must adjust our approach in a game situation accordingly. Find a beginning, work out a sequential line from there. Only when we bump into a refutation we switch to another line of thought. This way, there are no branches on our side of the board. There can only be branches created by the opponents moves. But even so, if you follow the logic, there is only just one line at the time that needs investigatio

### Monitoring the time

Forcing tactics are based on space (trap, mate) or time . It is important to know how "time" looks like, in a position. Time can originate from two sources: a B.A.D. (Barely Adequate Defended) piece or a multi function move. Let us see whether we can clarify these vague descriptions. The goal is to get some sort of definitions that simplify matters and reduces the chance of confusion. BAD The term B.A.D. (Barely Adequate Defended) piece is originally coined by CM Dan Heisman, if I remember well. What we want is a term for targets that are currently under attack, but where the balance of attack and defense is equal. Why are BAD pieces interesting? If we can disturb the delicate balance between attackers and defenders, the BAD piece all of a sudden becomes a ripe fruit, ready to pluck. We need to: change the balance of attackers:defenders pluck the fruit Where does the time come from? With a BAD piece, we already have the attacker in place. When we harass the defender, the opp

My approach at the moment might look a bit haphazard. But I don't want to lose momentum by trying to over organize things too much now the juices are flowing. The tempo battle differs from the LoA battle. It has a different focus. In practice both battles will intermingle. But at the moment I want just to  inventorize the pillars under the tempo battle. The LoA battle is more focused on space . A clear pillar is the multipurpose move. And a multi purpose is based on time , not space. You do two attacks while the opponent can only defend against one. Another pillar is the BAD piece. Where BAD should be read as B.A.D. (Barely Adequate Defended). Robert provided this position: Black to move 4r1k1/1b2qppp/1pr2n2/pN1p4/3P4/P4P1B/1P4PP/2RQR1K1 b - - 0 1  Here Re1 is BAD. It is attacked twice and defended twice. Rc1 is not BAD, since it is attacked once and defended twice. So the logic focuses in on Re1. By taking on c1, you take a defender away from e1. The main task of logic here is to

### Modelling clay

The result of my investigations during the latest tournament is twofold. I devised a method to educate system 1 by letting system 2 create a PoPLoAFun based logical framework for the middlegame. I haven't created that framework yet, because it takes a lot of time, But I expect that within a few months or so, I will have a logical framework ready for the middlegame that is usable for the Colle-Zukertort. Since I have adopted five new openings, it will take five times a few months to cover the spectrum of all openings that I play. I'm quite happy with the PoPLoAFun system now, since it is perfectly suited for the gap between the opening and the moment that Vukovic kicks in. I'm sure it will earn me a lot of points. The second discovery that I did, was that there is another gap in my chess education besides the battle of the LoA's, and that is the tempo battle. I wrote about that here . The PoPLoAFun system is not very suited for describing the tempo battle, so I'm go

### LoA skirmishes

In accordance with my discoveries the last tournament, I started to inventorize the lines of attack (LoA's) in the Colle which form the landscape where the skirmishes take place. I analyzed one of my own games, and the 16 tactical themes that are common in the Colle. This is what I found: 6 attacking LoA's 5 counter attack LoA's 5 defenses against LoA's If you want to murder your opponent's King in bed, it is very likely that you will make use of quite a few of these LoAs. It helps if you know them well, and the tactical maneuvers that can be applied. If the tactics don't crack your opponents castle, you can often use them to make favourable exchanges or to get a favourable endgame. Inventorying the LoA landscape that is common for an opening is only the first step. The next step is to formulate the chess logic that helps me to work with the lines of attack. Luckily the amount of LoA's per opening is fairly limited.

### Tempo logic

Again we are making a giant leap forwards. Invasion and King assault are best described by the battle of the LoA's. They are geared around the lines of attack (LoA's). And are perfectly described by the PoPLoAFun system PoP = point of pressure LoA = line of attack Fun = function of a piece But there is a vast area of different tactics, which are geared around tempo's. The battle of the tempo's, so to speak. From the previous post: White to move r2q1rk1/pb2bppp/1pn1p3/2ppN3/3PnP2/1P1BP3/PBPN2PP/R2Q1RK1 w - - 1 11 The first step is to write out the tree of analysis: Purple = white, green = black This is typically what system 2 must work out in order to find out that 1.Nxe4 wins a pawn in all circumstances. When I wrote the tree of analysis down, I was actually rather discouraged. How on earth am I going to abstrahere a simplified logic from this? But after a few hours, I started to see matters from a different standpoint. How can I describe this from the viewpoint of te

### Crystal clear logic

There is a chapter in the book about the Colle by GM Simon Williams with 20 thematic tactics that are typical for the Colle. I'm very happy with that. Since it are tactics that I will encounter in my games. I'm sure that a thorough study of those tactics will deliver me a lot of points. 16 of them are about an assault on the king, and I'm sure that I can describe them with the PoPLoAFun system until they become crystal clear. Furthermore, there are 3 traps, which aren't rocket science either. But there is one tactic I have difficulty with. Not because I can't calculate it, but because I have difficulty in to describe the logic in the right way. In order for logic to work as educational stuff for system 1, it must be crystal clear. I know it when I have accomplished that. White to move The last move of black was 1. ... Ne4 White can win a pawn here. As said, I have no problem to calculate it. But calculation is an activity of system 2, and I'm sure that it must

### Middlegame framework

As a programmer, I have done some debugging in the past 39 years. The most important thing in debugging the way I do it, is that I change only one parameter at the time before I retest. So that when the outcome changes, I know which parameter is the culprit. Of course there are other ways to debug. If you change 10 parameters at the same time and the outcome changes, you know that the culprit must be somewhere between those 10 parameters. Usually guys who debug this way lose the thread somewhere in complex environments. On the other hand, they may have much faster results in simple situations. Chess has multiple parameters. Getting mated can be seen as the ultimate bug. Given my step by step approach, systems with multiple parameters take an awful lot of time to penetrate into. That is why I have trouble with systems like HTRYC. I cannot keep an eye on 10 parameters at the same time. I need more hierarchy in my parameters first. After 23 years of chess debugging, a big picture starts