Monday, May 30, 2005

The Ultimate Day Zero

Yes, my friends. Yes. Yes. Yes. I did it. Finally and ultimately. Did it. Simply did it. Did what I first thought impossible. Was hard, really, but I am still alive and well. My brain is still working. Well, what else? Did puzzles the whole day long. As many as I never did before and at a speed I never did. Puzzle blitzing. And now it is over. Good bye circles. We had a good time. If I have profited, time will show.

Day 155 of the MDLM program, the Last Day!
Circle 7: all 1280 done in 9:11 hours
Stats: 88% (C6:87%) success rate, 26 sec. p.p.
All time high of all circles at maximum speed. Yeah.

Now it will take some time of break. No puzzles. No blogs. But I'll be back. Have a good time, Knights Errant, and good luck. It can be done. So do it!

The final countdown: 1 day

Yes, today is the day. Yesterday got up, out with dog, puzzles, brunch with wife, son and his girl friend, puzzles, visit dad, puzzles, cook, puzzles while chicken in stove, dinner with wife, puzzles, out with dog, puzzles puzzles, wife went to bed and said let it be, tomorrow is another day, I said no, impossible, this must be done now, she accepted, great wife, I love her, puzzles puzzles puzzles, uaaah, went to bed 2 o'clock in the morning. And now it is time to stop blogging and begin with the puzzles. Caissa, I am ready for the ultimate showdown ...

Day 154 of the MDLM program
Circle 6: all 1280 done
Stats: 87% (C5:85%) success rate, 31 sec. p.p.

Saturday, May 28, 2005

My List of Shame

After my last blunder in a better position I have decided to post this list here, as a pillory and must read before important games. It is supposed to be a warning for me and for all readers of my blog. The list begins with my start of the MDLM program. To be continued after my next rated game lost.

  • Missed piece-winning rook fork and pin, not considering all candidate moves (2005-06-11)
  • After opp queen (!) move not looked what it threats in better (!) position (2005-05-26)
  • Tried unsound tactics to free from unhappy but equal (!) position (2005-05-21)
  • Overlooked opponent's knight fork and pin in worse position (2005-04-12)

Friday, May 27, 2005

The final countdown: 2 days

Yes the pattern recognition is in action. They become familiar to me. But after the circles I shall have to arrange all this stuff into a framework in my brain. The whole thing is still quite chaotic to me, and I am sure that I still have many blind spots. Tomorrow will be family life again, but I hope to finish circle 6 sunday.

Day 153 of the MDLM program
Circle 6: 640/1280 done
Stats: 92% (C5:92%) success rate, 24 sec. p.p.

Thursday, May 26, 2005

Will I improve?

Is the MDLM program really useful? Will I improve? I decided to show everybody, be it a success or a failure. So I put my games and ratings record since the beginning of my MDLM program in the sidebar.

Just a couple of comments. Ratings are Swiss Chess Federation games. Unrated games are excluded. SCF publishes ratings 6 times a year. I use their Elo-calculator to get my estimated rating. This may differ a bit since the games are rated in the order of notification, not of date when played.

Wednesday, May 25, 2005

The final countdown: 3 days

Quite happy after this. Most of the simple patterns just pop out immediately. In many cases I need a minute or so. I still miss simple patterns, but this is no wonder while doing 320 in a row. My performance this afternoon was roughly 75%, but many of the missed ones have no clear solution after a 7 ply deep search of the computer. So this does not bother me, even grandmasters do not find all such moves in seconds.

Day 152 of the MDLM program
Circle 5: all 1280 done
Stats: 85% (C4:84%) success rate, 37 sec. p.p.

Tuesday, May 24, 2005

Bishop pair fetishism

For circles will take me half a day now, there was none today. Evening city championship game. Old friend, played sicilian, I attacked with Morra. He was not familiar with it and lost some time on his clock. I lost time with unnecessary pawn moves. At a certain point I was able to gain back the pawn but giving up the bishop pair. I did not want this and made a passive move instead. I once read in a blog that every passive move costs 0.2 to 0.3 pawn units. I think in critical positions this could be up to 0.5 pawns, and this is the value of the bishop pair advantage. So my bishop pair fetishism did not pay. My opp came into play, I got my pawn back, the bishop pair disappeared from the board and the game was drawn. We analyzed the active alternative and found that I would have got a position with winning chances.

Monday, May 23, 2005

The final countdown: 4 days

Been away from the circles for 2 days. Chess match. Family life. Today the 320 puzzles took me half a day! I am really tired now.

Day 151 of the MDLM program
Circle 5: 960/1280 done
Stats: 89% (C4:89%) success rate, 35 sec. p.p.

Saturday, May 21, 2005

Good and bad news

Good news: Welcome in our Hall of Fame, Sir Pale Morning Dun, Graduated Knight Errant de La Maza. You did a great job in your circles and blogs, and we hope you will continue these.

More good news: Our team won the match by 3.5-2.5, we are on second rank and if we keep this, we can play for promotion to a higher league!

Bad news: I lost. The story is quickly told: Bad pawn push, cramped position, try to free tactically, and I got what I deserved. But the lesson is very useful: I have to deal with positional patterns after the circles.

Game stats update
Rating at start MDLM program: 1627 (official)
8 games played, 5-1-2, 69%, 40 points gained
Actual rating: 1667 (self calculated)
MDLM progress rate: 0.28 points/day
(equivalent to 400 points in 1424 days)

Friday, May 20, 2005

The final countdown: 5 days

About the 92%, this seems to be my limit. My first 640 are a selection of the easiest from 1001 Reinfeld collection, so things get harder now. Tomorrow will be no circle work, but an important game in the Swiss Team Championship, a very slow G180. So I hope not to blunder there.

Day 150 of the MDLM program
Circle 5: 640/1280 done
Stats: 92% (C4:92%) success rate, 28 sec. p.p.

Thursday, May 19, 2005

The final countdown: 6 days

Recognition yes, memory no! I remembered several first moves quite fast, but did not see the pattern behind them. Of course it would have been a sin against Caissa to count these as 1 instead of 0. Speed is not the problem up to now. The problem is the sheer number of problems. The mass eats you. The only help is breaks, time and again.

Day 149 of the MDLM program
Circle 5: 320/1280 done
Stats: 91% (C4:91%) success rate, 27 sec. p.p.

Wednesday, May 18, 2005

The final countdown: 7 days

With this speed, circle 7 would take me 16 hours. I find it fast already, yet I must get more speed without losing accuracy. Tomorrow circle 5 begins.

Day 148 of the MDLM program
Circle 4: all 1280 done
Stats: 84% (C3:79%) success rate, 41 sec. p.p.

Tuesday, May 17, 2005

The final countdown: 8 days

Thanks for your support, Knights. I admire PMD, really, he is doing shift work besides the circles (or vice versa). I simply could not to that, but I am lucky being my own boss. And yes, I still have fun, and I welcome every pattern popping out ...

Day 147 of the MDLM program
Circle 4: 1120/1280 done
Stats: 86% (C3:81%) success rate, 41 sec. p.p.

Monday, May 16, 2005

The final countdown: 9 days

It has always been there, for weeks. The tiny little light. Growing slowly, becoming brighter. The end of the tunnel! Now I see it clearly in front of me, just a bit more of a week away. I have been doing the cycles for months, yet nearly half of all puzzles are still to be solved. Don't you call this insane? I do! I must have been crazy to begin such a torture! How can I get through? Wish me luck, Knights Errant!

Day 146 of the MDLM program
Circle 4: 960/1280 done
Stats: 89% (C3:84%) success rate, 40 sec. p.p.

Sunday, May 15, 2005

Improved

Things go quite well up to now in circle 4: More success in less time. But the hardest puzzles are yet to come.

Day 145 of the MDLM program, 10 days left.
Circle 4: 800/1280 done
Stats: 91% (C3:87%) success rate, 40 sec. p.p.

Saturday, May 14, 2005

Simple and complicated patterns

Yes, positional patterns are kind of simple, if not commonplace. I completely agree with Temposchlucker. And the number of positional patterns is quite limited. Dozens maybe. Compare this to the hundreds of tactical patterns, if we take all combinations, their number is infinite.

Yet this is also true: Tactics is essential to be a master, but to become a grandmaster you must be excellent in tactical AND positional play. Right? Then positional patterns should be a higher level? Wrong! Kind of a paradox, I must say. But I have an idea. The key may be not the patterns itself, but the decision when and how to apply them. In tactics this is simple: Either it works or it fails. In positional play there are never such yes/no decisions, but all shades of more/less. May we say that tactical thinking is digital and positional thinking is analog? And that analog thinking is more complicated (there is no analog computer yet). What do you think?

Day 144/155 of the MDLM program.
Circle 4: 640/1280 done, 4 days left
Stats: 92% (C3:88%) success rate, 40 sec. p.p.

Friday, May 13, 2005

Seven patterns for a win

That's how I won yesterday as White. I tried to figure out the patterns that were useful. Of the 7 patterns I identified, 5 were purely positional and 1 was mixed, but of course tactics was the key to win the decisive advantage.

  1. Play solidly in an unknown opening, using general principles of development. (My opp lost time, so ...)
  2. Trade off the strongest opp piece. (Remaining opp pieces were poorly developed ...)
  3. Open center when ahead in development. (Gave me the opportunity to ride a mating attack ...)
  4. When parried but still have pressure, retreat and switch target. (I doubled rooks in the center ...)
  5. Demolition of pawn structure by temporary knight sacrifice. (Yeah, here is tactics! Won a pawn ...)
  6. Swap pawns to create an isolated pawn as target. (Opp was forced to passive play ...)
  7. Trade off pieces when pawns up, in combination with removal of the guard. (After that I was 3 pawns up ...)

The ending was quite funny: My opp chased my king with queen checks (to prevent me getting a second queen) but I was quite happy because he chased my king from g2 to g6 where he helped my queen mate on g7.

Game stats update
Rating at start MDLM program: 1627 (official)
7 games played, 5-1-1, 79%, 53 points gained
Actual rating: 1680 (self calculated)
MDLM progress rate: 0.37 points/day (400 points in 1068 days)

Thursday, May 12, 2005

Value of a piece

Asked myself what that bishop of mine was worth, but found no price tag on it! Haha, said our IM, you better look at pawn structure. That's what makes the value of a piece. Well, I said, but there is much tension in these pawns, how can I tell what the structure will be? Wait, said the IM, be patient. Develop your pieces, then push or exchange pawns or be prepared to opp pushes or exchanges. As soon as the pawn structure is more or less stable, find all the strong and weak squares and where all the pieces can go and how easy. The better the perspectives of a piece, the more its value. And do this while your opponent is thinking. Use your own time to evaluate your candidate moves, to calculate forced lines and to do blundercheck. Great positional and endgame hero, our IM. Hope his advice will help me in my game this evening.

Day 142/155 of the MDLM program.
Circle 4: 320/1280 done, 6 days left
Stats: 91% (C3:87%) success rate, 42 sec. p.p.

Wednesday, May 11, 2005

Positional patterns

Yesterday at our chess club one of our IMs gave us a very instructive lesson I like to share with you Knights. Closed line of the Caro-Kann. Positions seemed to be equal. But then White saw that his light-squared bishop was bad and his opp's one was good. So he enforced bishop exchange. After that the opp was quite helpless to find a plan. He found a wrong one, got problems. White constantly increased pressure. The final was a nice queen sacrifice with a knight-rook mate. All of a sudden I began to realize that patterns play a key role not only in tactical but also in positional play! The key patterns are not more complicated than their tactical counterparts such as forks, skewers, pins. Just look at this one: Exchange your piece of less value against an opp piece of more value. After the IM told us this, the plan emerged clearly, despite of the closed position.

Day 141/155 of the MDLM program.
Circle 4: 160/1280 done, 7 days left
Stats: 93% success rate, 40 sec. p.p.

Tuesday, May 10, 2005

Circle 3 is history

I guess that about 10% of my puzzles are so difficult that even the computer does not find the solution in 10 secs. If I finish circle 4 near 90% success rate I'll be completely happy.

Day 140/155 of the MDLM program.
Circle 3 stats: 1280 done, 79% success rate, 1:15 min. p.p.
Circle 4 schedule: 160/day, min. 30 secs max. 1:00 min. p.p.
starting tomorrow ...

Monday, May 09, 2005

Circle 3 finish ahead

Uuuh, success rate dropping, and I did not manage to keep the pace of 1 min p.p. Did a lot of extra work this weekend on missed solutions. Hope this will help in circle 4.

Welcome aboard, Satish Talim

Day 139 update
Circle 3: 1200/1280 done, 80% (-3), 1:12 min. p.p., 1 day left

Saturday, May 07, 2005

Extra work

Need this before I start next circle. Look again at all the ones I missed. Figure out the best solution path: Ah, 2 loose pieces. Fork? No. Skewer? No. Discovered attack? Oh, Yes! 50 done this way.

Day 137 update
Circle 3: 960/1280 done, 83% (-1), 1:12 min. p.p., 3 days left

Friday, May 06, 2005

List of blind spots in chess

This list should be self-explanatory. If not so, or if I left out an important point, just drop me a line. Note that this is about pattern recognition. Blind spots in calculation may deserve a separate list. Exclamation signs mean: Pay special attention to ... !

  • Silent moves (zugzwang, other)
  • Long moves (Q/B on diagonals! Q/R on ranks! camouflage!)
  • Backward moves
  • Excentric moves (minor pieces! to border, corner)
  • Rook/Queen moves on ranks vs. files
  • Pins (of pawns! on diagonals!)
  • Latent geometry (yet to create)
  • Queen forks (camouflage!)
  • Pawn moves (pawn mates!)
  • Removal from (vs. of) the guard

Day 136 update
Circle 3: 960/1280 done, 84% (-1), 1:10 min. p.p., 4 days left

Thursday, May 05, 2005

Tunnel view

Spent some thoughts on backward moves (thanks for the remark, Tempo). There are three reasons why we miss them. One is camouflage that is inherent in the position itself (crowded board). The second one is a bad habit in our brain that I call tunnel view. We focus our attention on an area roughly the size of a Knight's reach, where we want to mate the King. So, if the mate requires a Queen switch to the corner h1 or so, we can spend minutes without finding the mate. The third one is the preoccupation that attacking means moving pieces forward. If we wipe this preoccupation out of our brains and if we adopt the habit of always scanning the whole board, I am sure we get at least half of the solutions we missed before.

Day 135 update
Circle 3: 880/1280 done, 85% (-2), 1:08 min. p.p., 5 days left

Wednesday, May 04, 2005

Camouflage

In my inner ear I hear this song, melody military style, lyrics about a marine called Camouflage. Camouflage saves lives of defenders, but, unfortunately, kills attackers. As solvers of chess problems we suffer from camouflage that hides the patterns and let us calculate nonsense.

Just again today: Wopp, new position on the screen, see exposed King, search for a mate, can King excape or not? Calculate, recalculate. Nerves. Hit the go button. Sigma Chess shows me the line. Now everything is clear. The music is playing elsewhere. And the final shot is taking a loose rook well hidden behind a crowd of pieces and pawns.

Specially prone to camouflage are: Long moves, diagonal moves, moves to and from the border or corner, rook/queen moves on ranks rather than files, file/diagonal/rank openings by pawn moves, trapped pieces. Need to anti-camouflage optimize my search strategy. Will deal with this in a later post.

Day 134 update
Circle 3: 800/1280 done, 87% (-1), 1:05 min. p.p., 6 days left

Tuesday, May 03, 2005

Zugzwang

Yesterday and again today I constantly missed patterns that I have missed time and again. This must stop or else I never finish the program! I came to the conclusion that I am a blind that must learn to see, so I will focus my next posts mainly on blind spots.

I think zugzwang ist one of the hardest to see, because it is not a pattern, but the absence of a pattern. You cannot see a zugzwang immediately, you have to calculate. In zugzwang the pattern recognition simply does not work. That's it. Today I had a zugzwang problem where I already knew that the Queen must check and then go to a square forcing zugzwang to opp King, Queen and Pawn. I knew perfectly what to do, but I did not find the square for the Queen. Hard, hard, hard! At least my rate did not drop further.

Day 133 update
Circle 3: 720/1280 solved, average 88%, 1:03 min. p.p., 7 days left

Monday, May 02, 2005

Blind Spots

Next to the 80 problems a day there is one top of my to-dos: A list of my blind spots. Today I encountered a blindness for long moves on ranks and diagonals. Earlier I remember chunked pieces attacking and defending each other: Which of them is the weak one?

Day 132 update
Circle 3 stats: 640 done, 88%, 1:02 min p.p., 8 days left

Welcome to our new Knight Errant: Harmless