Saturday, May 10, 2008

At the club

Look at the right spots.

Ideas tested at the club.
Friday I tried my scansystem against an opponent against whom I have never won a single game over the years. He has about 200 ratingpoints more and my score is +0 =12 -3, which is actually quite good.
For some reason our play doesn't match. I cannot get grip on his pieces and he cannot get grip on my pieces. Our latest encounters were all draws.

Although the scans are still time- and effortconsuming I managed a few times to do the full scans during the game. That showed me that the amount of elementary tactical motifs was way below the average in a masterlevel problem. There were a few knightforks now and then, but then I had to put my knight on a square which it couldn't reach because it was solidly protected.

No tactical shots.
The scans told me with certainty that there were no viable tactical combinations around. (The game became a draw again.) That brings me to the next point. I'm very happy with the scans and I'm quite sure that it will get me lots of points once the habit is fully settled. Yet there are situations that there simply is no combination around. No matter how good your scans are. For this situation a positional scan must be invented.

Positional scans.
I'm sure that a positional scan will sound as trivial as the tactical scan after developing it, yet I must take the pains necessary to go through the process of formulating a positional mnemonic and testing it. The positive side of triviality is that it is the best way to keep things secret. The items, however trivial, are carefully selected from a bunch of irrelevant trivial items. The fact that the application of a mnemonic is subject to subtility and it needs considerable efforts to establish the habit is sufficient guarantee that my ideas will remain "secret" for the mass.

With a tactical approach the elementary tactical motifs seem to come out of the blue. A scanmethod is only invented to detect them once they arise accidentally during the game.
This raises the positional question: how can I put my pieces so that the tactical scans start to deliver positive results? In other words, how can I inflict my opponent with double attacks, batteries, pins, overloaded pieces and convergency squares?

First I need a strategy to do so. There is no doubt that a lot of it can be found in positional books. But before you can decide which positional path to go you must scan the features of the position at hand. That is where the yet to invent positional scans come in. In the coming time I will elaborate on this. Next to my daily tactical scan-drills.

Just a coïncidence?
I have done a vast of amount of experiments the past 3.5 years. It always surprised me that none of the exercises yielded any interest. You never get something extra for free. A positive side-effect which you had never thought of. There was always a direct relation between your efforts and what it brought you. While I'm solving masterlevel problems with the aid of the mnemonic, all of a sudden my intuition begun to work. The killermoves started to rise up in my brain and I didn't know where it came from. Just an urgent feeling to make the right move came up. I wonder if that is a coïncidence.

A few questions.
Likesforests wrote the following comment on my previous post:

Dabattpinolc, dabattpinolc, dabattpinolc. It seems like it can't be THAT easy but you have extreme experience studying tactics so I will also try this mnemonic out as I study tactics and report back my success or lack thereof. I have a feeling this will speed up my process, but won't be able to solve some positions, that have say a queen that can be trapped or somesuch that's not a DA BATT PIN OL C. What about the traditional checks, captures, and threats? Do you try that first? That sequence seems to crack simpler problems very quickly.

I understand the doubt that it can't be THAT easy. The difference is made by the following:
You can't educate your autopilot while you are on autopilot. The conditions are consciousness and an active attitude. You don't focus on the solution of a problem but on the scans that trigger the pattern recognition.

Some tactics are not included, like a queen trap. I have focussed on the bulk. Initially I had included traps in the mnemonic. Since I usually don't miss a queen trap, I dropped it. Feel free to include it yourself. I'm rather pragmatic and try to avoid any ballast. It is difficult enough as it is.

The scan for checks, captures and threats is designed to narrow down on the candiate moves. It is in fact a kind of an heuristic. The scans I suggest have another goal. In stead of narrowing down your view, it is widened. The used method is guided pattern recognition. Just make sure you look at all important spots and your pattern recognition will take over. I have tried CCT myself but it yields to much irrelevant moves.

That sequence seems to crack simpler problems very quickly.
I need no aid to crack simple problems.

Blue Devil asked:
Do you think what holds you back in chess is not finding the right move in complex tactical positions? E.g., is missing such things the reason you aren't 2000 rated?

Partly. The other part is the positional side of the story, as I explained above. The solution for both problems is the same though: guide your pattern recognition. Both tactical and positional.


  1. Cool beans. Let us know what you like on the positional front. Piece activity should be god there, no?

  2. Yes, but how to translate piece activity into scans is another matter.

  3. Yes, true. I had trouble finding a sustained, useful treatment of piece activity (PA). Heisman talks about it extensively in 'Elements of positional evaluation' but in a way I didn't find all that practical.

    To brainstorm a bit, I look at PA as three things:
    1. Mobility (number of squares to which a piece can legally move).
    2. Freedom (number of squares to which it can move while still carrying out essential defensive roles (e.g., if a knight is pinned to the queen it has effectively zero freedom).
    3. Coordination. Are the free pieces working together toward some goal? (E.g., invasion).

    I find this useful in practice. When a piece is not mobile, mobilize it. If many pieces are free and mobile, it is time to coordinate them for an attack. This often involves pawn moves.

    It is fairly easy to scan for mobility (and certain concepts are tied to mobility of specific pieces, like a weak color complex which is often useful when your opponent has it, doesn't have a bishop of that color, and you do). Because there are specific 'rules' like that for specific pieces, it isn't all that simple, either, as maybe your dark-squared bishop presently has low-mobility, but your opponent has a weak dark-square complex around his King. I would try to hold onto that Bishop, because of its potential mobility.

    Heisman discusses potential and actual mobility quite a bit in his book, which was pretty helpful for me. It helped me to value putting rooks on half-open files, even when (and sometimes especially when) those files are still clogged with my own pieces--a good example of a strategic factor planting a tactical seed (discovered attack).

    I find it less easy to scan for freedom (e.g., I might forget a piece is pinned). Coordination is much harder, but it seems you have worked on that pretty well, implicitly, by working on attacks, convergence squares, and the like.

    This leaves out pawn structure. It took me a while, but I am just getting better at seeing weak squares. I find pawn structure to the be trickiest thing to evaluate quickly (of course a scan could include the standard list like backward, passed, doubled pawns and the like, but whether such things are good or bad is so dependent on the position (partly because whether a pawn form like that is good or bad is determined by how they affect piece activity!)).

    At any rate, the above is meant to spur thought/discussion about how to translate the abstract notion of 'piece activity' into more concrete, scan-friendly, terms. Perhaps others have other criteria they use.

    I'm not sure what else there is besides piece activity and pawn structure in the strategy department. Things like 'space' tend to be special cases of piece activity (where space is defined in terms of how advanced your pawns are, giving your pieces mobility, and threatening promotion (a tactical factor)).

    I wonder what the best book out there is on this topic? I know doing the PCT puzzles helped me. So did playing gambits, where I really learned the value of activity.

  4. I'm not sure how original you need to get, given how much has been written on strategy by GMs. The wheel may already be out there.

    Many people advocate doing a 'piece scan' (e.g., Buckley, and that new book 'understanding chess tactics' too), evaluating in some detail each piece, where it most wants to be on the board.

    Not original, but doesn't mean not good. (The same goes for mobility/freedome/coordination--those are all already out there, and probably the key points). For pawn structure, I have no idea as I understand that worse than anything.

  5. OTOH, turning this stuff into a usable scan is different. I feel I can do it well with activity: MoFCo (mobility, freedom, coordination)

    I can't with pawn structure, so if you have thoughts on that please tell us! :)

    Nimzovich might add a 'Ce' for center, but most would say that is a special case of mobility and coordination.

    But flexibility is another thing that I didn't even consider,--a kind of meta-activity. In how many different ways does your piece have mobility, freedom, can it coordinate in multiple attackes. Does it have mobility on one side, or both sides of the board? Is it part of two potential pools of attackers or only one?

    Shit, a new thought I'm gonna have to chew on. I think it may be important enough to include in Chessplanner. Version 4.0 here we come. :)

    Unfortunately it is tactics I suck at more than anything. My coach was like "Wow, you are good in the opening and are making good positional moves, but you just suck at tactics. You are missing basic 2 move tactics left and right." Unfortunately, he also said that was the thing it was hardest to coach, that I should play and do puzzles for a few years until my tactics catches up with my other stuff.

    Ho hum.

  6. As far as originality (I was thinking about this during lunch): if you look at your tactical scan, it isn't originality that was important. Discovered attacks etc are ideas that are out there. The key was finding the personal list of things that you have trouble finding without guidance.

    With strategy, it may well be the same thing, so the key question is, "What positional criteria do you tend to have trouble with?" and then make the acronym. I like Mxyzptlk.


  7. That the end result will sound trivial is something I'm convinced of. That was why a named a post triviality is the spice of life. Yet to get there there is some out of the box thinking needed. As you said. To personalize your own wheel as it were.

  8. Yes, but personalize it using the mundane categories everyone uses (my FSTDD scan is basically a direct transcription of a list from every book, but it doesn't include mate/en prise capture which I automatically look for first anyway). As you suggested, the originality is pretty minimal.

    With strategy, I've been trying to get you to discuss more specifically the pool of possible scan items from which to narrow down to the ones you need to use, but you keep giving Tempo one-liner cryptic non-answers. ;O

  9. In Tempo's defense, he did say in reference to the strategy scans:
    "In the coming time I will elaborate on this."

    So the comments here are for when he does.