Sunday, February 03, 2019

Defensive mental mayhem

Now I know that I only need about 10 hours solving time per tactical theme, I decided to take matters up a notch.

The second theme, exposed king, was about reviving the good old mate patterns of papa Polgars first brick.

Yesterday, I started with the theme defensive move. I already figured I am very bad at that. As I use to say: "I must have a feeling for chess, because I always find the wrong move when skedaddling away when under attack. If I would have no feeling for chess at all, I would statistically find the right move every now and then."

Defensive move problems in the 1550 - 1620 range feel the same as 1900 - 2000 problems with the two other tactical themes. I often have no clue.

With a defensive move, you try to accomplish trivial things like skedaddling to a square where your king can't be checked again, skedaddle and attack, skedaddle and defend, take back with the right piece that adds an additional punch, take back with the right piece that adds an additional defense, that kind of things.

So lots of room for improvement here.

White to move
1rb4k/3n1r1p/p1R3p1/1P1Pb1B1/2B1P2q/P1N2Pn1/3Q3P/2R3NK w - - 1 1

This problem took me 3:34 to solve correctly. Go figure!


  1. Did take me 2:18
    1.Kg2 Nxe4 2.Bxh4 is saw instantly ( white is material up, so white just need to survive ) then it did take 2 minutes to believe that the only move left: 2.Nxe4 can be survived. I think many of the lower rated fast solvers did not see the danger or dont have the habit to calculate to the end. Sometimes better player are slower because they see something which has to be analysed.

    There are 2 types of defensive moves: you are already up material and now you have to survive a danger, or the weakness you want to play cant be played as usual because of a weakness in your own camp.

    If you want to do a defence training: use the "mixed" set

    1. I recognize the inclination to mourn about those idiots that move fast and accidentally correct who make me look like a moron. But I will not learn anything from denying that I indeed might behave like one.

      My mind was simply drifting away during solving this position, playing trial and error with the pieces.

      Your mind was at least busy in a disciplined way counting the material which lead to the next best winning move.

      What we should ask though, is what we can learn from this position. 1. ... Nxe4 is a sacrifice. And when a piece is sacked, all alarm bells should go off. It's all hands on deck: IS THIS SACRIFICE CORRECT?? should be the automated response. Can I win this piece? But my alarm bell produced a measly blip which drowned in the noise my mind was making.

    2. Result would be a list of alarm bells?
      you may reduce the list, hanging piece is enough ;)

      because of the superiority of the own material i did know: its about surviving
      a standard method for defence is to exchange material, especially the opponent queen close to the own king.. well or some other dangerous piece close to the own king

  2. I think it's interesting that you are in the process of confirming several things that have been speculative about previously regarding effective and efficient training.

    (1) Limiting study (initially) to two-move tactical processes. I am reminded of this blurb from GM Andy Soltis' The Inner Game of Chess: How to Calculate and Win, "The Myth of the Long Variations", pg. 6:

    A popular view among amateurs is that grandmasters are grandmasters because they routinely see 10 moves ahead. There are, of course, examples of this by GMs, but they are relatively rare.

    Much more common is the kind of calculation that calls for SEEING not more than two moves into the future. And most of the time these two-move variations lead only to minor improvements in the position. But these improvements add up.

    When Mikhail Botvinnik lost on first board during the 1955 Soviet-American match, the world champion explained the result simply, "It shows I need to perfect my play of TWO-MOVE VARIATIONS."

    (2) 3 sets of 70 tactical problems per theme, with two-move variations. 210 problems X 32 tactical themes (excluding named checkmates) = 6720 problems to cover all Chess Tempo tactical themes. The 28 named checkmates with 210 problems each = 5880 problems for the checkmates. Combined, that's 12,600 problems. Not trivial, but nowhere near the "guess-timated" 50,000, 100,000 or 300,000 touted in the (non-chess) literature. I'm sure there is considerable overlap of these problem sets, which means the number is an upper bound. That makes NM Heisman's estimate of approximately 2,000 problems seem more realistic as a lower bound.

    (3) Approximately 10 hours spread over one month for training one specific theme (or two specific themes, given the theme under study, and 1 separate theme NOT specifically being studied, limited to two-move sequences). This may represent an average number of hours, with an unknown variance around the mean. This appears to be considerably less effort than the MdlM Seven Circles of Hell (or the Woodpecker Method). 10 hours per theme X 60 themes (tactical plus checkmates) = 600 hours. Spread over one month per theme, that means it will take approximately 5 years to cover everything. This may require a schedule adjustment as you get used to the process.

    (4) The need to refresh the ideas occasionally (perhaps using spaced repetition). It remains to be established how much repetition is needed to keep the memories of all the tactical themes fresh in your mind.

    Apparently we are living in "interesting times"!

    1. I look a bit different at the figures. You gave me a list with frequencies of the themes. The first 20 themes cover 80% of the presented problems.

      Now I know what I'm after, and experience that the exercised themes play a role in my games at the club, I can do easily 1 theme per week. Which means that I will be ready to put myself to the test within 20 weeks. Given that I don't write too much at my blog ;)

      I'm sure that the theme "exposed king", which is about rogue mates, will effect my overall mating ability. I'm still talking about chess.

      210 problems x 20 themes is 4200 problems in 200 hours. Net solving time as reported by Chess Tempo.

  3. The theme defensive move is ideally suited for learning the ABC of trivial chess moves.

  4. I went back to the local chess club for the first (hopefully not my last) time in many years. There was a Quick Tournament scheduled: 5 rounds at 8 minutes with 8 second increment. For having no preparation, I didn't do too bad: 3 wins, 2 losses. In my loss against the highest rated player, I had no idea what to do with 1. e4 c5 2. c3 Sicilian. I tried 2. ... d5 3. e4xd5 Qxd5 and very quickly lost the thread. Against the second highest rated player, I had an excellent game all the way through, but lost on time (with less than 5 seconds on my opponent's clock). I've never liked fast time controls, but that's the only game available at the club.

    I do know that I "saw" more of the tactical potential than 4 of my opponents. In my last game, I made a stupid mistake that gave my opponent the opportunity to exchange his Queen for two Rooks. Fortunately (for me), he got fixated on the threats to his King and overlooked it, and resigned.

    I've got to study some of the openings played for next week. At the level I want to play, some detailed opening knowledge is essential. The club is starting a 24-minute (with unknown increment) Quick Tournament. I hope to do better over time.

    1. I learn my openings mainly by playing online blitzgames. After each game i analyse it with fritz and the IDEA of aquarium. i have an opening book in fritz made from mastergames so i can see the traditional theorie, and i ask super super super gm stockfish. At the end i store the new line in a file. while the computer is analysing i usually play an other game. of course.. sometimes i look into an opening book or look for a mastergame of "my" line in a database, listen to a openingvideo at youtube..
      1.e4 c5 2.c3 Nf6 is main line ( masters play it most of the time ) score 53% for white
      1.e4 c5 2.c3 d5 is no 2 in my computer opening book, score 53.2% for white.
      These anti sizilians need some extra attention from an sizilian player, this is not typical sizilian game anymore

  5. My play at the club indicates that I should do positional problem too, at one moment. Does anyone know a good problem set/ training site?

  6. I was nearing the 4 minute mark, so guessed wrong, as your comment said that your concern was with the time taken. Within another half a minute, I got the answer right without looking. The way I look at this problem is that it can take 4 minutes to understand the problem, and virtually no time to then know the solution.

    I believe there is too much focus on pattern-recognition discussed in this blog, and by it's members. This was more a straight-line calculation problem, as are so many on Chesstempo, versus the ones out of a tactics book, which require more creativity in the move-selection process.

    This position says that if someone, such as myself, were to walk up to a game - in the middle of the game - between two other players, figure out what was going on, and then make a move, that it would take me 4 minutes to do that in this particular position. Now, it might take me less time than that at a tournament, where my mind is usually in it's most prime shape.

    Anyway, any chess position can take a few minutes to figure out what is really going on, starting from scratch (without seeing and thinking about it from previous positions). It's not indicative that there is a particular chess problem going on in the position, other than to say that all positions have one.

    1. I believe there is too much focus on pattern-recognition discussed in this blog, and by it's members.

      Thinking is overrated. Maybe there is no such thing as thinking. Maybe there is just pattern recognition.

  7. One thing I should have added is that I believe that pattern-recognition is very important when it comes to double-attacks, as it's perhaps the easiest type of tactic to miss. Of course, Averbahk's book on "Tactics for Advanced Players" is great, in this regard.

    I would say that pin patterns are particularly important when first making Class A.

  8. I won my first game in the 24 minute tournament last night. I had White; the opening transposed into a variation of the Queen's Gambit Declined. My opponent played very obligingly to eliminate all of the minor pieces except for my Knight and his dark-squared Bishop. He allowed me to execute a classical minority attack on the Queenside. At one point, he overlooked a Knight fork of his Queen and Rook, enabling me to pick up the Exchange and a Pawn for the Knight. After that, I just ground him down by creating a passed Pawn on the Queenside. Near the end, he allowed me to establish one of my Rooks on the 7th rank absolute (Nimzovitch), and it was all over.

    I also played several informal 8-minute games against the highest rated player, all of them Smith-Morra Gambit (him) against the Sicilian (me). I lost every game except the last one, when I finally figured out what was needed to repulse his attack. I still need more study of the anti-Sicilian lines.

    I can tell that I am sharper tactically than back in the days (md-1970s) when I played tournament chess a lot.

    1. i did play the morra as white for some years by myself, so i accept the morra and play a e6(Bc4->f7),a6(Nc3-b5->c7),b5(Bc8-b7) against it. But you need to know what you are doing then. There are some antimorras buut, when you can stand the fire (morra) then dont play with the matches ( sizilian ) ;))

  9. In the one game I won against the Smith-Morra, I eventually arrived at e6, a6 and an eventual b5 as a viable alternative. The person I was playing said that e6 is the key to the Black defense against the Bc4 lines. I still need to study this more, with a lot of master games to cement the typical variations and ideas.

    I like playing with the Sicilian fire! My earliest book was How to Think Ahead in Chess: The Methods and Techniques of Planning Your Entire Game (Fireside Chess Library), by I. A. Horowitz and Fred Reinfeld. The authors advocated the Colle System for White and the Sicilian Dragon for Black. I forget what they recommended against 1. d4 or 1. c4; I think it was the Queen's Gambit Declined. However, that was nearly 50 years ago when I first started playing (and acquiring a bad case of Book Acquisition Syndrome), and I no longer have the book.

    I've always thought that was a strange combination of openings for a beginner. I can appreciate the simplicity of the Colle System approach, but the tactical hazards of the Sicilian Dragon (especially the Yugoslav Attack) are pretty difficult to navigate for a beginner. Maybe the idea was that if you persisted until you could actually survive, you would have to absorb sufficient tactical ideas. I guess it's somewhat similar to the advice of Master Ken Smith: Until you reach approximately 1800 USCF (about 1700 FIDE), your first name is tactics, your middle name is tactics, and your last name is tactics!

    1. and acquiring a bad case of Book Acquisition Syndrome

      I know what you mean. I have 42 meter of books of my own. Of which about 3 meters are chess books. If there is a relationship between rating and chess books read, I will need another 73 meters of chess books to become a grandmaster.

  10. I solved it pretty fast, dont know the time, but much faster than you for sure.
    How is this possible? Well, you title is about defensive moves, that was the major give away hint.
    First white is in check, so only move is a king move to start with.
    Second - with all the things going on, I was looking for a defensive good move. The best defenders are knights. If you have 2 knights close to your king, the attacker has a really hard time. So moving the other knight closer to the king came to my mind pretty fast - and bang - the knight in the center guards my queen --> solved it.

    Really, even in Bullet and Blitz games I often dont calculate much, but I KNOW I need to keep my knight(s) close, so I avoid to get under attack.

    If you solve puzzles, and you know it is a defensive move --> 90% of the time, the knights play a key role.

    1. with all the things going on, I was looking for a defensive good move. The best defenders are knights. If you have 2 knights close to your king, the attacker has a really hard time. So moving the other knight closer to the king came to my mind pretty fast

      I suspect it are these little gems of knowledge that make the difference.

  11. In many puzzles i have a question, which takes too long to answer. I changed my training more to this question and its propper answer. So i create own (sub-)puzzles. Just to see and understand the right answer is often not the real problem of such a puzzle.

    1. Seems to me like you are trying to lift weights with 120% of your maximum capacity.