Saturday, September 30, 2006

Blank 117

Yeah. Did 117 problems at Chess Tactics Server without any error today. Of course I became super careful as longer as the winning streak lasted. So my rating dropped to 1406. It's enough for today, I'll continue the series tomorrow.

Wednesday, September 27, 2006

Slowmouse Is Back

Well it was more or less forced, this return of Slowmouse at Chess Tactics Server. I have been doing some 4000 CTS problems as Mousetrapper, but with the percentage focus of old Slowmouse, lifting my overall success rate from 78% to 82%. this brought me up from success rank 470 to about 270.

Slowmouse is unhappy with any session below 95%. Today I did 95.9% of 218 @ 1450 which brought me up to rank 21 in the success list. Half of the guys in the top twenty have about a hundred or less tries and most of them with very low ratings. Which means that they just wanted to get top ranked in percentage. If we take only the 1000+ tries, Slowmouse is already in the top ten.

Interestingly, Slowmouse has a rating of about 1440, 20 more than Mousetrapper.

But I have to be humble. Look at the success distribution. The curve has a big peak at 75%, a smaller peak at 80% and an even smaller one at 90%. If you imagine three superimposed bell curves and estimate the areas under the curves, then you must come to the conclusion that the guys with a 90% success strategy are about 1/8 of all tacticians. In other words: Would all share my high success (low blunder) strategy, I would score only in the top 200 or so.

Friday, September 22, 2006

About Consciousness In Chess

My position regarding the issue of unconscious vs conscious decisions in chess, raised by Temposchlucker. (Tempo, feel free to continue this discussion in your blog.)

Tempo gives a number of facts that support the view that GMs or even IMs can play and win games without thinking: Susan Polgar in a simul, a master while talking to his wife, and I can add a story some years ago in our club. I blitzed against our strongest player, and he talked to a friend, hardly ever glancing at the board, and he won, of course.

But every case has its counter case. The same IM has lost a blitz against one of my teammates recently, he was lost in position and also by time, which tells us that the IM had been forced to think. Hey, and I have won blitz games against this teammate.

Another counter case: The famous 2.6 seconds of Susan Polgar in a simul of hers (mentioned by Tempo time and again). I was not at that simul, but I had the opportunity to watch her sister Judit in her Zurich simul. That is: walk, move, walk, move, walk and on and on and on and ... stop and think (or look at the position, if you don't want to call it thinking). Sometimes ten seconds. Sometimes even half a minute. Even Kasparov did so.

Now let's have a look at the idea of unconscious move decisions. BTW I prefer the term subconscious, because any of these decisions may surge up to consciousness if necessary. And when is it necessary, even for grandmasters? Whenever the position becomes critical, this is my guess. Fast moving occurs in two kinds of positions: firstly in quiet positions where general strategic considerations prevail, and secondly in easy tactics.

Anyway, I mentioned grandmasters only because Tempo did so. More important for me is the question of how I should use the subconscious and the conscious decisionmaking myself. After tons of problems done at Chess Tactics Server I can put them into two categories: easy and hard. In easy problems the pattern pops out in a more or less subconscious way. In the hard problems I have to use a conscious search until I find it.

Okay, we are still in the phase of pattern spotting. Now comes the real problem, deciding a move. This involves not a positive pattern scan, but a negative one. The absence of bad patterns such as hanging pieces, counter patterns or simple defense moves.

Again, there are two categories of problems. In the easy problems, I have spotted all relevant elements in the six seconds before the problem makes its move. I just have to check if the move changes something. If not, I can play my own move grandmasterlike out of my guts. The zen of playing chess, yeah.

But there are enough other problems, the vast majority of them. They force me to check and check again for blunder. I would call this conscious thinking. And because I am a patzer and never will be a grandmaster, I must do much more of this sort than he has to do.

All is relative, by the way. Look at grandmasters playing grandmasters: they think and think and think again. Subconscious moving can only be seen in grandmasters against amateurs, and only if the latter play like patzers.

Thursday, September 21, 2006

Paying For Accuracy

Three weeks ago I used to play a level of about 1460-1470 at Chess Tactics Server. I was quite happy with a 90% accuracy then, compared to my former 80%.

But still I felt uneasy when I saw my rating drop. This caused me to play faster, and I paid this with a drop in accuracy, and after such a session I felt quite unhappy. I even got angry sometimes.

Then I realized that I still had the wrong view, despite my new focus on accuracy. I mean, I still regarded accuracy as a sort of money to buy as much rating as possible, trying to be more efficient, that is, paying less accuracy loss for more rating gain.

Yesterday I looked at it the other way round: Rating was the money now, and I was ready to spend it for more accuracy. When I gained two or three points in a good series, I said to myself, hello, here are some more points to SPEND. Earlier I always was anxious to LOSE them again.

And yeah, it worked: I had my best session ever, 96.5% in a session of 143 with a new personal record of 77 correct in a row, starting with 1442 and ending with 1444.

My next goal will be a blank session of more than a hundred.

Wednesday, September 20, 2006

All Skills Involved

Yesterday I had my best CTS session so far: 94% @ 1442, 199 tries, 57 correct in a row.

Today I looked back at yesterday's problem history and tried to figure out the lack of what skills caused the loss of points. Useless attempt! After going through three done problems I stopped it, because always all five skills are involved.

It takes too long until patterns pop out because my square vision is weak. Pattern vision depends strongly on square vision, but also on the discipline to scan the whole board for targets. This scan, in turn, is easier and faster with a strong square vision.

Maybe move vision is not so important at a CTS level of 1400-1500 which is the range of my problems now. But it will become more important with 1500-1600.

Evaluation is important for the quick decision of where to focus and what to neglect.

And of course discipline is always a main point: Oops, queen is hanging, and I have looked for an escape. Wrong! First look for a check while keeping the queen in mind. Yeah, the winning move was a check, and I have found it, but I wasted too much time considering a passive defense for the queen.

Conclusion: Most of the five skills have to be used all the time in all problems at Chess Tactics Server.

Tuesday, September 19, 2006

Five Basic Skills

I have the impression that some of us bloggers, me included, are in danger to get lost in theories, discussions and statistics about details of unknown importance, to say it politely. We, and me, are in danger of missing the real thing.

I mean, chess training is about improving basic tactical skills, and I dare to say that even grandmasters make use just of the same basic skills as we do. Only they are magnitudes more precise and faster than we are. I have tried to put these skills together, and I found five of them:
  1. Square Vision: Squares controlled by pieces (x-rays included) pop out and are perceived as distinct from other squares.
  2. Pattern Vision: Geometric patterns formed by pieces and controlled squares (such as basic checkmates, pins, forks, skewers and the like) pop out and are perceived as distinct from the rest of the position.
  3. Move Vision: Ability to visualize a position after one or more moves have been made (avoiding retained and anticipated images).
  4. Evaluation: Material counting with unequal trades in move sequences, counting of attack-defense balances, assessment of races (pawn promotion race, capture race), assessment of positional advantage, assessment of urgency, ranking of candidate targets, patterns and moves.
  5. Discipline: Always looking at the whole board (avoiding tunnel vision), always looking at opponent resources (as well as the own), coolness (not being euphoric in advantage nor frightened in danger), well-organized thinking (avoiding quiescence and overcalculation errors).

I think that it is possible to get the tactical abilities of a grandmaster just by optimizing these five basic skills. Of course, this will not be sufficient to become a GM, because a lot of positional knowledge will be necessary, plus tons of openings, plus more tons of endgames.

Of course I am aware that my square, pattern and move vision never will reach the GM level. Never. But I have already have made some progress. For instance, I have now done hundreds of problems at CTS without leaving my king in check one single time any more.

Thursday, September 14, 2006

Two Mothers Of All Patterns

After my farewell to never-ending lists I have been thinking a lot about how to find a new way to deal with the tens of thousands of pattern variations that are presented on the board. Storing them all in the memory cannot be the way. But what else?

Well, I just try the opposite. Instead of looking at all the different variations I try to see what they have in common. I have boiled them all down to just two mothers of all tactical patterns:
  1. Restriction: Making use of the restricted mobility of a pinned, overworked or trapped piece. This can also be called slow tactics because, typically, it is done in two stages: First setting the trap (mating net) or the pin, then attacking. Special cases: normal checkmate, stalemate. Defensive tactics: Escape, depinning.

  2. Acceleration: Doing two jobs with one move (double attack, attack plus defense, double defense) or overcharging the opponent with two jobs that he cannot do at once (overfeeding by two hanging pieces that cannot be taken at once). This can also be called fast tactics because tempo gain is involved. Special cases: Double check, double checkmate.

When I remember the most frequently asked question of mine when I see a new problem at Chess Tactics Server, it is this one: Can I take time for this job because the problem has no defense against it, or must I play fast (in an accelerated way) in order to prevent a defense?

This is my reason to believe that the «mothers of all patterns» are not just a useless theory but, in contrary, may help me to improve.

Saturday, September 09, 2006

Are We All Crazy?

When I read all these recents posts of mine and of other Knights, I must say, o God, o God, are we all crazy? I read such things as a systematic of a thousand pin patterns. I see lists of basic checkmates that go into dozens, and to these all traps must be added, queentraps, rooktraps, bishoptraps, knighttraps, oh, oh.

When I ask myself if such lists (hundreds of items, categorized, all this stuff) can help me to improve at Chess Tactics Server or at real games, I come to the one and only answer. No. No, no, no. Definitely no.

Why? Chess is basically a simple game. I mean, the pieces and what they can do are a very limited set of rules. Complexity comes into it by the combination of these simple elements to clusters. Now we have two ways of dealing with it: We can build huge lists of these clusters and make things complicated. Or we can improve our ability to filter the simple things out and to handle them quickly and with precision.

Today I have decided to quit the list club. Well, not completely. I keep the list of eight basic checkmates, because I find it very useful. And it is simple enough to be handled in practical situations.

I have come to the conclusion that chess training should meet two goals: improving chess vision and improving chess processing.

Chess vision must be optimized in two directions
  1. excentric: Look at a piece and spot all squares that are controlled by that piece.
  2. concentric: Look at a piece or square and spot all pieces that attack and defend that square.
Chess processing is all the stuff about values of pieces, priority (check before threat before other moves), tempo gain and tempo loss, checkmate, stalemate, zugzwang, eternal check, pawn promotion, discrimination of dangerous from harmless threats, and the like.

That's it, boys. And now I end this post an go to CTS, training my chess vision and processing abilities.

Friday, September 08, 2006

A General Strategy For Predators

Let's take the example of a hawk going after birdies. His first step is to spot them. Okay, he has spotted a pigeon far away and a sparrow much closer. As a second step he must sort these targets by size and by distance. Then he must focus on the prey he has ranked top, and then decide to go or not to go after this or after that or after another prey yet to be spotted (because the sparrow is too alert and about to escape, and the pigeon is too far away).

It seems that the four-step strategy «Spot. Sort. Focus. Decide.» is much older than chess. I remember a TV feature on marine biology that presented the school formation of herrings as an anti-focus strategy against sharks. Confused by the sheer mass of herrings the shark is unable to focus on a single herring, let alone to decide which one to swallow.

This reminds me of certain CTS problems where a near-mate, a piece win and a piece plus pawn win is presented. Or those tricky ones where a mate in three is false because there is a mate in two.

Thursday, September 07, 2006

Spot. Sort. Focus. Decide.

When, as a boy, I learnt to plane a piece of wood, I learnt that there is a limited set of conditions that make a good plane stroke. First the blade must grip the wood, but not too much. Then the stroke must go in the right direction, taking into account the direction of the wood fibers. Then the stroke must make sure that the shaving does not block the plane. All this must be controlled by subtle variations of speed and pressure, it is very complicated in the beginning but very simple after you have learnt it.

I have come to the conclusion that playing good chess makes use of the same principle. A limited set of steps that must be taken from looking at the board until making the move. It is up to every player to find his own set. For me, there are four basic steps that repeat all the time: 1. Spot 2. Sort 3. Focus 4. Decide

Just as one plane stroke is not enough to work a piece of wood, one Spot-Sort-Focus-Decide circle is (in most cases) not enough to find the right move and to make it. It may take two, three, four or even more circles.

First step: Spot

Make sure that you have seen everything (targets, attackers, defenders, patterns) before you take the next step. Sources of failure: tunnel vision, not looking at the whole board, missing hidden targets on crowded boards, missing patterns not adequately stored in memory for quick retrieval.

Second step: Sort

As soon as more than one element has been spotted, they must be sorted: Which elements match? Which is the most important one? Sources of failure: Underestimating and overestimating attacks or defenses, miscalculating values of pieces, missing tempo-gaining (intermediate) moves, and the like.

Third step: Focus

Focus on the top listed element. Assess candidate moves: checks, captures, threats, other. Sources of failure: Wrong ranking of candidate moves, for instance, looking at a threat before a check or at a distant check before a close check or at a bishop check before a queen check.

Fourth step: Decide

Is an element not important, discard it to save time. Is the top candidate the best move, then play it. Is it (probably) not the best move, do another circle, taking into account more spotted elements. This is the most important and the most difficult step. It is the step where failures in thinking become failures in moving. It is the step where you may lose too much time for the right decision. It is the step where you blunder by moving too fast.

My next project will be an analysis of my failures at CTS, sorted by Spot-Sort-Focus-Decide. Every problem where I lose rating points, even -0.01, will be eligible for this analysis. Of course failures will occur in every step, but I'll count only the main failure step in problems solved too slowly. In problems failed to solve there are two steps involved, one of them is the wrong decision.

Tuesday, September 05, 2006

Mapping Chess Mind

After more than twenty-five thousand problems done at Chess Tactics Server I am asking myself all the time what is the most important experience, and how might I take profit from it.

What strikes me most is how simple the elements of patterns and decisions are, and how huge the number of combinations of all these simple things. This brings me back to my idea of the brain as a tidy toolbox. A place for every thing and every thing in its place. A limited set of powerful tools and a quick grip to them.

Failures often have trivial causes. For instance, I miss the possibility to mask a threat, having only looked at take, escape and guard. Plus it could possibly be topped by a more forcing counter-threat. Obviously I have not always a clear picture of all possible decisions in a given situation. Or I miss targets or forget some of my own forces. Nothing complicated, just ordinary simple stuff.

Here are the basic outlines (seven main areas) as I see them at the moment:
  • 4 Basic Mind Steps: 1. Spot 2. Sort 3. Focus 4. Decide
  • Attack & Defense: Spotting targets and forces to attack or defend them
  • Urgency: What must be done first and what can wait?
  • Value: Accounting material win, loss and equal trades
  • Space Patterns (geometry)
  • Time Patterns (tempo)
  • Operations: from simple basics to tactical combinations

I'll come back to each of these map areas in subsequent posts.

CTS performance today: 88% @ 1455 (n=104)

Monday, September 04, 2006

What Is A Good Move?

A move that provokes a bad move of the opponent. And what is an excellent move? A move that forces a bad move of the opponent.

In my last game, when my opponent played Qa5, I was a bit shocked and asked myself, had I played hope chess? I thought this move was good, and even Deep Shredder ranks it as a top candidate. It forces queen swap, and on first sight this must be in favour of Black because I am down a pawn and must keep my queen for attack.

But then I discovered Rc1, and because Qa5 has provoked a good move, Qa5 is not really good. And why is Rc1 good? Because it leaves Black with two bad options: a) queen swap loses one more tempo for Black and wins a tempo for White (for rook doubling on c-file), and b) without queen swap the position of the queen is of no use which means also the loss of a tempo.

CTS performance: 92% @ 1450 (n=95)

Sunday, September 03, 2006

Morra Miniature

Yesterday I had my first test (2h/40, 1h rest) after an intensive phase of working on my thought process and training at Chess Tactics Server. Winning a miniature against a 100+ opponent is not bad so far, but I must admit that my thought process was far from being perfect. I need to work more on it.

This was the last match of our team this year. We won 3.5-2.5 and ended as winners of the group and can again play for promotion. (Our club has ten teams in various leagues, and if we succeed to promote, one of our youngster teams can go up in our place, so we serve as a sort of elevator team of our club.)

White to move and win. I had missed to exploit this tactic one move earlier (I would have won a knight), and my opponent had missed to prevent it, he even made it worse. Then he was so shocked losing his queen for the bishop pair that he preferred to lose a knight and a rook for nothing. He shook hands without a word, refused my offer to analyze the game and left, slamming the door.

[Event "SMM"]
[Site "Switzerland"]
[Date "2006.09.02"]
[Round ""]
[White "Mousetrapper"]
[Black "RG"]
[WhiteElo "1620"]
[BlackElo "1710"]
[ECO "B21"]
[Result "1-0"]

1.e4 c5 2.d4 cxd4 3.c3 dxc3 4.Nxc3 e6 5.Nf3 Bb4 6.Qd4 Bxc3+ 7.Qxc3 f6 8.Bf4 Nc6 9.Bc4 Qa5 10.Rc1 Nge7 11.O-O O-O 12.Bd6 Re8 13.b4 Qd8 14.Rfd1 (14.b5 $18) 14. ... b6 (14. ... a6 $16) 15.b5 Na5 16.Bc7 Nd5 17.Bxd5 Qe7 18.Bxa8 1-0

CTS performance today: 85% @ 1450 (n=164) :-(

Friday, September 01, 2006

Story of a Lost Point at Chess Tactics Server

The important things do not happen on the chessboard. They happen in the brain. Nobody can look into the brain. Therefore I have to use everyday pictures to describe what happens there over and over again. Or what I think that happens.

Oh, again such a chunky thing! Must make it smaller to fit in, then I win the point. Where did I put my plane? Aaargh! O, here it is. Let's plane that thing down. Oh shit, this is not wood, it is metal. Where is the file? Oh these files, where did I put them? Here it is. Oh shit, too fine, I need a coarse one. A coarse file, a scrub file. Where, where? O yes, here, finally. I got it. Now let's make this chunk smaller. It fits.

Chess Tactics Server: Yeah. You've solved the problem ... in a lot of seconds. ( Result 0.01 : 0.99 ) Your new rating is ... (-1.00)

Moral of the story: What I need is a well-organized, tidy toolbox in my brain. This is the first rule I learnt as a schoolboy in the manual skill lesson. The rule has been posted to the wall of the school work-shop. It said «Jedes Ding an seinem Ort erspart viel Müh und böse Wort» (Every thing in its place saves a lot of labour and dispute).

CTS performance today: 91% @ 1463 (n=226)