Friday, March 30, 2007

Chess VIII -- The CCT-rule and pattern recognition

I imagine that the following objection to the CCT-rule can be raised: "following that systematic way of thinking is a waste of time. Are you familiar with the concept of 'pattern recognition'? It's basically the idea that you train your mind to recognise pattern so that you see things immediately and don't need a systematic way of thinking. That's what the aspiring tactician should strive to archive, to solve enough tactical problems so that they start to get a 'feel' for a given position and see what needs to be done or thought about. You're assuming that the human mind works like a computer, but I got news for you: it doesn't! We can never be as efficient as humans in doing brute force, and need to work with the strength we got. So take your CCT-rule and shove it up your a**!"

What is my answer to the above? Let me begin by saying that I think its garbage. Well, not really, I actually agree with a lot of it, but it's stupid of the objecter to think that I wouldn't. Most of it is completely irrelevant to the CCT-rule. And I know, I'm was rude (for saying 'garbage' and 'stupid'). But so what, he began. You did notice that part about shoving the CCT-rule? Did you, did you? Punk. Oops.

Firstly, and this is point that I haven't stressed in the past but is somewhat self-evident: If you do see something interesting immediately, a way to win a piece or such, then by all means go for that (or examine variations thereof, trying to win more). That's perfectly all right. The CCT-rule is more a rule to fall back on when you *don't* see something immediate and have no leads. As soon as you see something, even if it's a capture or threat while you are examining checks, switch to that which you see, even if it means deviating from the rule. The rule is just a tool to help you find good moves, and often your mind will make 'unexpected' jumps. For example, you see a check that would be great to do but you can't do because the square is guarded, but then you see that you can capture the piece that guard the square. In this scenario, check mentally what happens if you capture that piece, and do this despite not having examined all checks yet. You've found something, go for it! If it leads to nowhere, go back to the rule and the (semi-)systematic approach.

Secondly, the CCT-rule does leave a lot of room for intuition; everything isn't structured. The rule just gives you the start point of certain variations; you still have to decide by other means which variations are interesting enough to think more about, and evaluate the resulting position. I have no rule for that (yet. :) )

Thirdly, far from being anti-intuition, the CCT-rule is actually an efficient way of training your intuition. By constantly going through these series of questions (what are all possible checks, and what happens on the board when I make them? What are all possible captures? etc) you train your mind to start to automatically notice these things. You get faster and faster at answering these questions. That's how intuition is built. (Okay I know, I made the exact same point in my earlier post. Repetition is the mother of all knowledge etc.)


Wednesday, March 28, 2007

Chess -- CCT-rule revisted

That simple CCT-rule has really grown on me. Having spent time with it, using it and thinking about it, I can now see that it's even more beneficial than I initially thought. Specifically, I can see that it covers more than I thought, and I see more and more useful connections between it and other things. Using it is like having a tree of knowledge growing inside you, or something. Maybe that sounds cryptic, let me explain.

The CCT-rule, as I've explained, is the check/capture/threat rule. That these things should all be checked (all moves that fall under these, it's usually not that many and you don't need to go that deep within each), and in that order. This is what I've discovered using it:

You don't miss much. Since you look at everything within these areas, you will find moves that you otherwise wouldn't. The rule doesn't cover everything in every situation, but it's great when attacking or being attacked. Great for finding tactics or just keep the attack and initiative going. (I'll write about the limits of the rule, and what you need in addition, later.)

You get much faster at finding certain things on the board. (This point will get its own post later.) Since you always systematically look for these specific things, your mind eventually starts to notice them without you even asking the question. This is the stuff chess intuition is built of.

The "capture"-part has a tight connection to "pressure" (that I've written about) and "removal of the guard" (a tactical manoeuvre). Every possible capture is a "pressure point", and the sum of these is the total amount of pressure. And for every pressure point you might ask yourself: why can't I take this piece? (Assuming you can't). That will often lead you to noticing the guards, and then you might ask yourself "can I remove this/these guards?", and that may lead to ideas you wouldn't otherwise have. (And btw, there are exactly four different ways to remove the guard, and maybe I should go through them all! I haven't decided. Again, I don't want to make it too complicated, but I don't want to miss anything either.)

The "threat"-part has a strong connection to the tactical inventory and the enemy's weaknesses. That's how you find the threats. You don't just mentally randomly try out moves and see if it threatens something; you begin by looking at the enemy and see what can be threatened. You find the loose pieces and weak spots and other weaknesses (see the post on the tactical inventory), and you move a piece to attack this. If you can't find a direct threat, just increasing the pressure is fine.

So you see, by applying the simple CCT-rule you will automatically get into all these other things. You don't have to remember to check out "pressure" or doing a tactical inventory or even examining "removing the guard" as specific points, rather its integrated into the CCT-rule. You create strong mental associations between the elements in this rule and these other things, and your mind will automatically remind you to do these things. That's how it should be done: you shouldn't have a long list of disorganized things to check, but a simple rule with logical connections to these other things. That's mental integration.

Of course, the final proof of the value of the rule (and whatever other rules I can come up with) is my own coming success. : )


Monday, March 26, 2007

Chess VI -- Rapid Chess Improvement part I

There is (or was) a debate on the net regarding a certain book on chess skill development. His ideas on how to develop skill is different from many other books (which he criticises). The book is "Rapid Chess Development" and the author is Michael de la Maza. I haven’t read the book, but I’ve read about it, and I’ve read the two articles on which it is based (sort of), and I have a few comments.

I think he has some good points and some bad. Where do I begin? Actually, I think I have to split it to several posts.

He place a lot of emphasis on tactics. I think that’s good, for non-expert players tactics is THE most important thing to train, vastly more important than opening theory or such. (Not that you should ignore other areas completely.)

He mentions a few insights he reached. One is "chess knowledge isn’t the same as chess skill". That is certainly true, and is exactly why my chess training nowadays (if this relaxed thing I’m doing can be called training at all) consists not of learning chess (except training tactics, but that’s very skill-related training as against acquiring chess knowledge) but how to bring the relevant knowledge to my mind at the right time through a systematic way of thinking, and asking the right questions. A systematic way of thinking is HUGELY important. There is no point in having a lot of knowledge in chess unless you’ve learned how to use it. That’s precisely what "the psycho-epistemology of chess" is, even if Michael de la Maza wouldn’t put it that way. However, I don’t really agree with what he is suggesting on how to think (I think I can come up with something better, at least for me), although it’s clearly better than some other advice I’ve seen.

I’ve seen a slight improvement the last week or so: I’ve found tactical solutions through my system that I otherwise wouldn’t have seen, or would have taken longer. On the other hand, it feels somewhat unnatural thinking in this new way, but that’s because it is new. And I’ve barely begun.


Wednesday, March 21, 2007

Chess V -- pressure

I think that the concept of ’pressure’ might be of help for me to integrate into my thinking when playing chess.

So firstly, what is it? It’s what you put on your opponent when any of your pieces (pawns included) can capture an enemy piece, whether it’s a good or a bad capture. So there is a simple way of measure pressure: just count the number of possible captures. If two pieces can capture the same piece, count it twice. Now, some possible captures may be completely unimportant, and some positions may have fewer but more important captures possible, so the number of possible captures isn’t all that is important, of course.

The more pressure you exert on the enemy, the better. It means more options, more possibilities of there being some good a winning path (more possibilities of some nice tactics). Also, exerting pressure forces the enemy to defend, it ties his pieces up and fixates them. Pressure is good even if there is no immediate tactical gain from it, but eventual tactical gain is often the result of pressure. It’s a positional advantage with a clear connection to tactics. Also, there is, very generally speaking, a somewhat clear plan if you’re the one pressuring while the enemy is defending: add even more pressure, throw in a currently under used piece; place an already active piece even more active. Or move back and forth between targets, because the attacker in these situations is often also the one with spatial advantage, which sometimes mean your more able to switch between targets than the enemy is able to switch his defence.

There is clear connection between pressure and space, but they are not the same. One way of measuring space is to count the number of fields (whether they are occupied by pieces or not!) on the other half of the board that you can either move to or capture (in case of pieces). However, space is more about maneuverability and is far more a positional aspect than pressure, which is more tactical (though not purely so -- it’s in that area between tactics and position.)


Monday, March 19, 2007

Chess IV - The Tactical Inventory

So I’ve mentioned *the tactical inventory* a few times. Time to dig a little deeper.

Firstly, I want to place it in the wider context of chess theory. I place it between pure tactics and position. Pure tactics is calculation and maneuvers (these often have a name -- forks, skewers etc), and position is open lines and pawn formation and stuff. The tactical inventory positional aspects with a close connection to tactics. A major (*the* major?) example of this is undefended pieces. They are often suitable targets to attack. Another is a king on eigth/first row with no escape-route (which invites placing a rook or a queen at that row), and yet another is the enemy queen and king on the same diagonal. Notice the difference between these aspects and tactical maneuvers -- they are ’static’ qualities on the board while the maneuvers are... well, something you do. And often the maneuvers take advantage of precisely these aspects in the position that are identified in the tactical inventory.

I saw one page on the net using the names "theme" (for tactical maneuvers) and "motif" (for that positional aspect the maneuver is taking advantage of). However, this terminology seems both pretentious and arbitrary to me. But I do need a name for "that positional aspect the maneuver is taking advantage of". Well, maybe I can use the "motif" part? No, no. The word I’m looking for could be *constellation*. Which is more accurate and descriptive than "motif". So, you use tactical maneuvers to take advantage of a certain constellation. (It also occurred to me that "taking a tactical inventory" is to a large extent - or maybe completely - to identify weaknesses in the position.)

It is interesting to observe the tight connection between certain tactical constellations and what to *do* in the position, in contrast to what I’ve written about how it is not obvious how the identification of certain positional aspects leads toward finding a move. If the enemy has two undefended pieces, it is natural to look for ways to attack both at the same time. If the queen stands on the same line or row as the king, you wonder whether a rook do a skewer (or pin), and so on. Now, these are obvious examples of course, but bear in mind that these tactical constellations arise on and off through out the game, and the idea is to always be aware of them. Sometimes you can gain in more subtle ways than actually manage to pull of a double attack, for example moving somewhere where you *threaten* to make a double attack, forcing the opponent to defend.

Anyway, one of the things to automatize during a game is (during the opponents turn) to take a tactical inventory. To establish this habit one has to direct oneself explicitly again and again which the command "okay, let’s do a tactical inventory." And one needs to have automatized a (short) list of things to look for.

(It seems that "what’s the weaknesses of my enemy" is a better question than the imperative "let’s do a tactical inventory." Maybe that’s just a habit thing, but it’s also important to personalize the questions asked, so that they are psycho-epistemologically optimal)

The list:

1. Undefended pieces (and pawns?) (How about trapping a piece? A rook may not be undefended, but if you can take it with a knight or bishop that’s usually a gain even if it’s defended.)
2. Interesting lines, rows and diagonals (kings and queens on the same line/diagonal etc) (Or is this too close to purely positional analysis? It’s tricky to draw a line.)
3. Weak first/eight row (makes mating maneuvers possible). Or does this fall under "interesting rows"?
4. Pieces and pawns attacked once and defended once (they are basically undefended should they be attacked with one more piece).
5. Pieces performing more than one task (overworked pieces)
6. Pins.
7. Possible forks? (I don't know, I just feel that the above about lines and diagonals covers rooks and bishops, but I have nothing to cover knights.)

This list could (and should?) be longer. Or should it be shorter? Too long is bad. Am I supposed to go through every item repeatedly during a game? When? I don’t know, this is a damn tricky area. But I think I’m on the right track, and that what I’m doing here is the right way to improve.


Friday, March 16, 2007

Delphi and general design decisions in a language

Here are some pros and cons with Delphi, over and above what I’ve written (e.g. the great value of the Delphi-environment).

One thing I value in a language is ease of use when it comes to standard stuff. Those ordinary things you always do, like using basic I/O, file operations, conversions, arithmetic and so on. Most languages have their annoyances (actually, all I know have that, with the possible exception of Python though I’ve just barely begun using that. I don’t mind the forced indentation in Python.) Java requires too much work to do I/O (all that stuff with streams to do such as simple operation. But Java has that nice feature that auto converts non-strings to strings when using the "+"-operator. Delphi doesn’t, so you have to using function calls for that, although is an inbuilt always accessible easy-to-use function.

I like arrays in Delphi. Index doesn’t have to start at zero, and the indices in multidimensional arrays can be written within the same brackets, for example like this: [i, j] (instead of [i][j]). Simple and nice. Also, as in languages such as PHP and Python it’s possible to use ordinary arrays as map/key structures. That is, as indices you can use whatever values you want, including myArray[myString] where myString is a string. (Languages such as Java and C++ only let you use integers as index values. In ordinary arrays that is, there are of course available map-structures as classes, but that requires a little more work to use.)

The way you can receive several variables of a certain type with one declaration (in functions and procedures) is nice. You can type some_function(row, line: Integer), not having to type the type (no pun intended) twice, as in many languages).

And btw, I love the inbuilt powerful data types in Python. Having that powerful tool at your disposal with that ease of use is really nice, and worth writing about in its own blog-post some day.

Delphi deviates from the usual by having return-statements (called "Result" in Delphi) that works as variables. They do not end the function upon being called, but just changes its content and whatever the content is when the function has reached its end is what’s getting returned. That’s kind of weird and actually an semantic difference (compared to most other structured and object oriented languages) rather than a syntactic one. Haven’t decided for certain what I think about it, but I think I like it.


Wednesday, March 14, 2007

Is GTD a cult?

As an Objectivist I’ve often come across the accusation that Objectivism is a just a cult. Not that I’ve been accused of that myself, but I’ve seen it in debates and texts, there’s even a book called "The Ayn Rand cult" (which I’ve read, and I’ve read many other books on actual cults -- it was an interest of mine some years ago.) Now when I’ve started to follow what is going on in the GTD community, I see similar charges are being made against GTD.

Obviously I don’t agree at all with the charges against neither Objectivism nor GTD. They are typically unfair and based on nothing (or based on a twisted interpretation of something).

So what are the charges against GTD? Here are a few:

It’s rigid: no deviation from the system is allowed. (Not really true. Firstly, the system itself is pretty flexible, allowing for much personal tweaking. Secondly, if you want to skip a part or add something not part of GTD, just do it. Who will stop you? DA may recommend certain things, but so what? He’s just saying what he thinks works best and why, take it or leave it.)

If something in the system doesn’t work out, it’s the fault of the individual. (I haven’t actually seen GTDers say this. On the contrary, I’ve seen the opposite, that GTD may not be for everyone etc. The forums I’ve been in have been relaxed and friendly, not full of blame or such. On the other hand, some claims critics make do meet some opposition, but again, so what? People making false or dubious claims should be opposed. If someone misunderstands GTD, tries it and fails, they shouldn’t be surprised if someone is pointing out that they haven’t tried the actual GTD. Even this, however, is done in a friendly and non-condemning way.)

David Allen (the creator of GTD) is a part of MSIA (which is true), and he is communication secret messages through GTD to brainwash people into joining MSIA. (I’m not kidding, I’ve seen those exact words, also that he is conducting "thought reform", so subtly that you can’t find the messages if you look for them.) There’s also the claim that David Allen is trying to hide his membership in MSIA (not true). This one is just silly.

People are afraid to speak up. (Playing the victim-- "boo hoo, I get attacked every time I dare question the GTD dogma!") That’s their problem, people have all sorts of irrational fears.

Btw, I’m probably going to reconstruct my system into something that isn’t quite GTD anymore. Many GTD-practices are great, but some are not well suited for someone with my lifestyle and need to be changed.


Monday, March 12, 2007

Chess III

How’s it going with my thinking rules? Well, some problems have occurred.

I said I would think about positional stuff (including doing a tactical inventory) during my opponents turn, and specific calculations during my turn. One problem is: it is way to vague to have "think about positional stuff" as a mental guide. Out of all countless "positional" things to think about, which should I choose? I want rules that are as independent of the actual position as possible, to have the same rules as often as possible (some people, those who do not understand the way in which abstractions covers many different concretes, would immediate brand that as "inflexible" and tell me about the necessity of adapting the thinking to the actual position. Of course I would agree about adapting the thinking, in a certain sense different from the one I was talking about.)

Another related question is: how do I go from identifying positional aspects to finding a move? Okay, there’s an open diagonal, there’s a well placed piece, and there’s... okay, so what? The aim is to find a good move, and the connection between a positional analysis and a good move isn’t always obvious. Maybe I slant my analysis too much on what is rather than what could be? In any event, I need specific questions, and I want to integrate into the mix the following: Keres/Kotovs four positional areas, the question of purpose (what I want to achieve), plan (how to achieve it) and the question "what are my opponents weaknesses And of course the tactical inventory, which I think will help a lot when integrated fully.

Another problem: the checks/captures/threats-rule (CCT-rule?) It’s easy enough to check all checks and captures, and know you’ve looked at all. But the threats-part isn’t as clear. How do you know you’ve looked into all possible threats? Of course, I’m talking about immediate (or near immediate) threats here (otherwise it doesn’t fall under tactics), but it’s still much more difficult to handle than checks and captures. I need to break it down. Does it come down to "threaten to make a move which the next move will check and/or capture something"? Maybe, or should count such things as "moving a black knight e5, threatening to move it to a secured position at d3" (a black knight at d3 is kind of threatening regardless of captures and checks, but it is more a positional advantage than a tactical one unless some immediate tactical manoeuvre is possible).


Wednesday, March 07, 2007

Chess II

Okay, I’m trying out the following two rules:

1. This one I mentioned briefly in my last post on the subject: checking up all checks, captures and threat, in that order, and for both me and my opponent. And I’m doing this when it’s my turn to move, before I make a move, and for every interesting move I do the same for my opponent (i.e. checking which checks, captures and threats he can make in return.)

2. Calculate specific moves when it’s my turn, whether they apply to me or my opponent, and general principles when it’s my opponents turn, whether they apply to me or to the opponent. (One exception might be open positions, when calculating moves is of special importance. In these positions it might be a good idea to calculate moves even at the opponents turn.)

The general principles referred to above are basically of two categories: positional and tactical. Positional principles are things weak pawns, the placement of the pieces and so on, and tactical principles are tactical themes that are present in the current position (things like unprotected pieces, mate on the first row, the placement of piece in relation to forks, etc. These have a strong connection to positional principles, but aren’t the same. Tactical themes are basically positional aspects with a fairly direct connection to tactics, so it’s a mix of position and tactics, you might say). Checking for tactical themes is what sometimes is called doing a "tactical inventory". In some positions, in particular those with an open center, I think you almost only do tactical inventories rather than positional analysis.

It is really, really easy for me to overdo what I’m doing to here, to create a really elaborate system that takes into account everything but is impossible to follow. The idea is to really practice and make a habit out of the things I come up with. They should be few, and they should be the best possible out of many alternatives. An essential part of this whole thing is to try out everything I propose and see how it goes (and not keep building the system without a throughout testing of it at the same time.)

And btw, I haven’t here motivated the above rules, just stated them. But I might do that later.


Monday, March 05, 2007

Chess and psycho-epistemology

I played chess actively for a couple of years some years ago, and now I’m doing some playing again.

I’m think I’m going to make a systematic attempt at getting better, while at the same time not spend too much time on it (the last part is important -- it’s easy to start to obsess about chess and just play "one more game" etc.)

The question is how. How do you get better at chess? (Specifically faster games - blitz - played on the net.) Well, what I’m thinking at the moment, and which usually isn’t being taught, is a systematic way of thinking. I often found myself loosing after allowing something to happen which I *should* have seen, and *would* if I just had taken 1-2 seconds to specifically look for it. So why didn’t I? It didn’t occur to me, because I lack a systematic way of thinking (of course, a really good chess player would have seen it without even trying, but...). My thinking is rather random and chaotic. Not completely random of course, but directed by intuition and what happens to get my attention on the board. Those are not unimportant of course, and any systematic way of thinking must leave room for letting intuition be the guide.

That’s one important challenge: to not make the system to rigid or too extensive. For example, you can’t make a list of 25 things to go through in every position (e.g. "are there any open lines I can use?", "what’s the weakest point in the enemy camp?" etc), that would be very inefficient and very boring. And the fact that it is 25 things indicates that it is a system that is too specific (and too disorganized). By contrast, one simple three-step rule I saw the other day is a pretty good one: check all checks, captures and threats, in that order. That’s a simple (and just a small part of what I'm looking for) but efficient rule. In some positions you’re done in a few seconds, in others you may find interesting lines to think about for a long time. It sounds basic, but many, including some very good players, don’t follow that advice. They can miss a nice queen sacrifice because they didn’t actually look at every check. (Using the queen to take a pawn protected by another pawn is a kind of move that’s easy to discard mentally as a bad move, thusly it’s easy to not even consider that check... even though you would have seen the value of it immediately if you just had examined it.)


Friday, March 02, 2007

Non-explanations annoy me

There are number of explanations that doesn’t explain anything. What I specifically have in mind is explanations that just repeats what you are trying to explain. One example: it is easy to observe that people like different music and movies and so on. So, how could this difference be explained? Well, one way is to refer to the fact that people have different taste. But, having "different taste" just means "liking different things", which is what we are trying to explain. It’s just a restatement! That’s obvious if we substitute "different taste" for the equivalent "liking different things". Why do people like different things? Because they like different things! Some people actually think that they have explained something with that.

This is what happens when one is trying to think without knowing how.

Locations of visitors to this page