Saturday, March 11, 2017

War on Trial and Error

While performing an investigation for the previous post, I realized how weak  trial and error actually is, in comparison to calculation in the direction of a goal. Just random trying in the hope for clues must be the last resort during a game. In the study room we need to eradicate trial and error.

Diagram 1 - white to move
2r2rk1/1p3pb1/p3bBpp/2p1P3/4NR1Q/8/1Pq3PP/4R1K1 w - - 0 1
[solution]

The black king looks like a duck, it sits like a duck and it quacks like a duck. So it is probably a sitting duck. So the main goal is clear: mate the king! If the opponent feels compelled to postpone his demise with a piece sac, I'm fine with that.

 I tried from here on to solve the matter with trial and error, but I failed. We need a clear goal here that gives direction to our calculations. What can that be?

What can we say in general about this position? White has 4 attackers (Bf6, Rf4, Qh4 and Ne4). Black has 3 defenders (Bg7, g6,h6). The most likely scenario is the trade of the bishops, saccing the rook or the knight for one or two pawns to get access to the king and give mate with the queen and the remaining piece. Beware that Qc2 and Rf8 don't enter the defense.

I'm not sure where this is going. I'm going to investigate it and will update this post later. Comments in the meantime are helpful!

UPDATE
Great comments! It took me quite a few hours before I even knew where to start. I think I have an idea now. The black king is protected by a fortress of 3 pawns: f7, g6, h6. The base of this fortress is f7, which is bolstered by Rf8 and Be6. We must ask the question: which defender is the weakest?

f7

At the moment f7 is well defended.

g6
g6 is protected by f7. Notice that f7 has a double function: protecting g6 AND Be6. So g5 can be weakened by attacking Be6.

h6
h6 is definitely the weakest point. If it is pushed to h5, g6 will get a double function: defending h5 AND keep the g-file closed.

1.Bxg7 Kxg7 is simply undermining h6

2.Qf6+ is designed for three things.
  • It prevents that f7 is pushed, which would open lines for black for counterplay OR for adding defenders from aside.
  • It clears h4 for the rook
  • It forces the black king in the path of the black rook, so it takes more time to let Rf8 defend h6
2. ... Kh7 must protect h6 (either by the king OR Rh8)
3. Rh4 Rh8 forgets its function to protect f7
4. Ng5+ double attack K and B the rest you can find yourself.

Now we have found already two universal questions to guide our calculation in the case of a mate attack:
1. Which piece is the sitting duck?
2. Which defender is the weakest?

39 comments:

  1. easy to see its about mate. the solution is the standard method: remove defender and get more and stronger pieces closer to the opponents king. The Re1 is to slow so its all about Bg7 Bf6 Ne4 Qh4 and Rf4

    I was not able to calculate it from the beginning to the end but i was quite confident that i chose the right moves

    ReplyDelete
  2. Too subtle for me. Thought in time a check could uncover a discovered attack on blacks loose queen at a move in the variation

    ReplyDelete
  3. 1936.1

    Black has two extra Pawns. There are no Pawn promotions in the immediate future.
    White has four pieces (with one in reserve) in the immediate vicinity of the Black
    King. Black has only two potential defenders (excluding the Black King). Black has
    no attack against the White King.

    Looks like White MAY have a winning attack against the Black King!

    At first glance, I "see" two obvious mate possibilities as patterns.

    (1) A Bishop and Rook mate with WBf6, WRh8. The White Queen could be substituted for the White Bishop.

    (2) An Arabian mate with WNf6, WRh7.

    Then I get stuck, trying to force one of these two patterns into existence. I spilled nearly 45 minutes
    trying to force one of these two patterns to work. The "obvious" solution is not going to work!

    At this point, I started to give up in frustration. Then I thought, why not try to work it out completely.
    Could it be that the mate (or maybe a gain of material?) is "hidden" somewhat deeper? Maybe try the most
    obvious forcing moves and take another look at the position after we reach a quasi-quiescent position?

    1. Bxg7 Kxg7 (forced; 1. ... g5 2. Qxh6 and the Black King will die very quickly) 2. Qf6+

    (a) 2. ... Kg8 3. Rh4

    Function! The g6 and h6 Black Pawns must "gimme shelter" to the Black King and
    consequently cannot move without further exposing it.

    (a1) 3. ... g5 4. Rxh6 any 5. Q(or R)h8#

    (a2) 3. ... h5 4. Rxh5! gxh5 5. Qg5+!

    (a21) 5. ... Kh7 6. Nf6+ Kh8 7. Qh6+ Qh7 8. Qxh7#

    (a22) 5. ... Kh8 6. Qh6+ Kg8 7. Nf6#

    (b) 2. ... Kh7 3. Rh4

    (b1) 3. ... h5 4. Rxh5+ gxh5 5. Qg5! Qxe4 6. Qxh5+ Kg7 7. Rxe4

    (b2) 3. ... g5 4. Qxh6+ Kg8 5. Qh8#

    (b3) 3. ... Rh8 4. Ng5+ Kg8 5. Nxe6 fxe6 6. Qxe6+ and Black is losing material.

    (b31) 6. ... Kg7 7. Qd7+ Kg8 Qxc8+ winning a Rook

    (b32) 6. ... Kf8 7. Qxc8+ winning a Rook

    (b33) 6. ... Kh7? 7. Qf7#

    And now, I check my solution against Chess Tempo, and find:

    I was right! The given "solution" is variation (b3).

    I can hardly believe I worked through all that, even though it took over an hour and a half!

    I think that shows that I have the capability to work at a higher level, even if I can't do it quickly in OTB situations.
    Repetition should help decrease the amount of time required to work through to a solution.

    That was very encouraging!


    ReplyDelete
  4. A good post to think. about on the weekend. Spent some time thinking about this as I had my coffee, went on the treadmill and did a couple of chores.

    Here is my two cents upon looking at the answer.

    Since chess computing analytic has a large amount of brute force trial and error. I wonder if we will ever get pass T & E. I think the best we can do is see what looks interesting, shuffle the pieces and recheck the best option works.

    A little intelligent guess work may bring the most promising thing to the top of the T & E list removing the need to evaluate lesser lines. That might be the most we can ask for.

    My evaluation- after looking at the solution ....

    It appears that White has the conditions for a mating attack. 22 points to the king more or less with Black has a weak h6 pawn , a single bishop in defense and a Queen that might be able to pop back to the action .

    The question is how to open this can, hold onto the initiative (as Aagaard says don't lose your breath) and if there is a final mating pattern based on the pieces and pawn at hand. I looked for an Anastasia mate and the rook or queen on H8 supported by a bishop. Both patterns which are often found buried within a position. Not seen here.

    I think this could be a classic case of how a finachettoed position is weaken with removal of the bishop. I wonder how this compares with other CT-Art problems in the crazy classifications like attack on King side fianacetto with pawns on g6 & h6. Perhaps some ideas could of been found there.
    Although I feel slightly uncomfortable using phrases like playing a dark square game,I'll say that's what this is as all the action occurs on the dark squares.

    Looking at the position and for a win to occur I think we can say a few things have to happen.

    A Line with have to open up (possibly the h file)
    The Queen and Knight will have to act like a sheep dog pushing the King around or making threats that must be dealt with (forcing moves)
    taking advantage of pins would need to play into the combination somehow.
    In looking at the position I felt that the rook would be better placed on H5 and in the course of play the queen and rook swapped places. Interesting that a new file didn't open up but a square was reloaded by the queen.

    That's what I got , hopefully we might find a seed of something useful.Cheers.



    ReplyDelete
  5. Robert here's a Sunday Insane Chess Puzzle of Chessgames.com from today . it is a mate. (to solve or just enjoy)

    Initial Postion , wtm.

    http://www.chessgames.com/perl/fen?fen=1q1r3r/pp1n1pk1/6p1/1Q2p3/3N4/2P1R3/PP4PP/5RK1


    Solution and commentary

    http://www.chessgames.com/perl/chessgame?gid=1795532&m=23

    ReplyDelete
    Replies
    1. @ Jim Takchess

      PART I:

      FEN=[1q1r3r/pp1n1pk1/6p1/1Q2p3/3N4/2P1R3/PP4PP/5RK1 w - - 0 1]

      At first glance, I have a "fantasy vision" (hallucination?) of a Rg7+Ne6 checkmate, with BKg8. No,
      that's just a hallucination; no "road kill" there. But that Black King seems like a "sitting duck."

      If I can (somehow) get the White Queen into the action on the Kingside, there is a good possibility
      that White can run the Black King out onto the road. Since I'm the "attacker" as White, I have to
      be aware of the "hidden" attack by Black on h2 by the Black Queen (presuming Black is ever given time
      to capture on d4).

      1. Rxf7+

      Strips away a little more of the Black King's cover in a very forcing way. White still has three pieces
      available for a potential mating attack, so we fulfill the "three piece rule" requirement for mating attacks.

      Running away doesn't work for Black:

      1. ... Kh6 2. Rh3+ Kg5 3. Ne6+ Kg4 4. Qe2#

      Not taking the White Rook results in the Black King becoming "road kill":

      1. ... Kg8 2. Rxd7 Rxd7 (any other reasonable Black move allows 3. Qd5+ followed by mate on f7)
      3. Qxd7 winning, with a piece up and threats of mate still in the air; 3. ... exd4? 4. Re8+ Qxe8
      5. Qxe8+ Kg7 6. Qd7+ K moves 7. Qxd4
      and the White Queen becomes the "Queen of the Hill" - the hill on
      which the Black King is going to die. It's also straightforward to just capture the BRh8 with
      6. Qxh8+ Kxh8 7. dxe5 and win the "easy" Pawn ending with two extra Pawns, one passed (e5) and another
      potential passed Pawn on the Kingside (2:1 majority with good Pawns).

      I'm reminded of a dialog from mister Lasker's Lasker's Manual of Chess, contrasting the
      combinative player with the positional player:

      Thus the following story is told of an onlooker at a game. He was a combinative player.
      Suddenly he interrupted the players: "I see a magnificent combination, a sacrifice of the Queen,"
      he excitedly called out to him who was to move. "If your opponent then takes the Pawn, he is Mated,
      and if he goes out of check, he is Mated in two." "Well," replied the player, "but the principal
      question is: what am I to do if he captures the Queen?" "That is the only variation," replied the
      combination player "which I have not yet looked into."

      Delete
    2. PART II:

      I supposed it's time to look into the implications of that capture on f7.

      1. Rxf7+ Kxf7 2. Qd5+

      This allows White to get his Queen a little "closer" to the Black King with tempo. As GM Tisdall said,
      "Big piece to the middle is always a good move [especially in mating attacks requiring the Queen!]."

      What about the alternative "forcing" move 2. Rf3+? Let's leave that until after we have figured out
      whether the Queen move achieves anything concrete. Getting the White Queen involved seems obviously more desirable
      than just opportunistically checking the Black King.

      There are now 5(!) possible Black King moves. GM Tisdall recommends that the attacker examine the easiest move(s)
      to refute FIRST when attacking, and the most critical move(s) first when defending. So let's take a look at some
      of the defenses from the attacker's viewpoint.

      1. Rxf7+ Kxf7 2. Qd5+ Kf6

      Walking into the lion's den seems suicidal, so we'll eliminate it first.

      1. Rxf7+ Kxf7 2. Qd5+ Kf6 3. Rf3+

      3. ... Kg7 4. Qf7+ Kh6 5. Rh3+ Kg5 6. Ne6+ Kg4 7. Qf3#
      3. ... Kg5 4. Ne6+ Kg4 (4. ... Kh5 5. Rh3+ Kg4 Qf3#) 5. Qe4+ Kh5 6. Rh3#

      That gives us some ideas that might be useful in the other remaining variations!

      1. Rxf7+ Kxf7 2. Qd5+ Kf8 3. Rf3+
      3. ... Kg7 (3. ... Ke7 or Ke8 allows 4. Qe6#) 4. Qf7+ Kh6 5. Rh3+ Kg5 (5. ... Kg4 6. Qf3+ Kg5 7. Ne6#)
      6. Ne6+ Kg4 7. Qf3#

      1. Rxf7+ Kxf7 2. Qd5+ Ke7 3. Qe6+
      3. ... Kf8 4. Rf3+ Kg7 5. Qf7+ Kh6 6. Rh3+ Kg5 7. Ne6+ Kg4 8. Qf3#

      1. Rxf7+ Kxf7 2. Qd5+ Ke8 3. Qe6+
      3. ... Kf8 4. Rf3+ Kg7 5. Qf7+ Kh6 6. Rh3+ Kg5 7. Ne6+ Kg4 8. Qf3#

      1. Rxf7+ Kxf7 2. Qd5+ Kg7 3. Ne6+

      This might be the critical variation because it seems to allow Black a little breathing room. The action of the
      White Queen is temporarily blocked by the Ne6.

      1. Rxf7+ Kxf7 2. Qd5+ Kg7 3. Ne6+ Kh6?? 4. Rh3#

      1. Rxf7+ Kxf7 2. Qd5+ Kg7 3. Ne6+ Kh7 4. Rh3+ Kg8 5.Ng5+ Kg7 (or Kf8) 6. Qf7#

      1. Rxf7+ Kxf7 2. Qd5+ Kg7 3. Ne6+ Kg8 4. Ng5+ Kg7 5. Qf7+ Kh6 6. Ne6! and mate is threatened with 7. Rh3#

      1. Rxf7+ Kxf7 2. Qd5+ Kg7 3. Ne6+ Kf7 4. Rf3+

      4. ... Kg8 5. Ng5+ Kg7 6. Qf7+ Kh6 7. Ne6! and mate will follow (eventually) with 8. (or later) Rh3#

      4. ... Ke8 5. Ng7+ Ke7 6. Qe6#

      4. ... Ke7 5. Ng5! and mate will follow with 6. Qe6# or Rf7#

      And so we have a successful attack with 2. Qd5+ and we can ignore the possibility of 2. Rf3+. It might be a quicker way to win, but having found one solid way of mating (or winning sufficient material) is more than satisfactory for ME! I don't automatically "buy" mister Lasker's maxim: "If you see a good move, look for a better one!") The "road kill" in the beak is infinitely more satisfying than the imaginary "sitting duck" that got away.

      It is possible that I did not copy down all of the moves accurately. However, I think the appropriate ideas are clear.

      BTW, I haven't checked the given solution on chessgames.com and haven't examined the position using GM Stockfish or any other engine. It did take quite a lot of time to get everything written down accurately (I can only hope!).

      Delete
    3. A correction to PART I:

      "It's also straightforward to just capture the BRh8 with 6. Qxh8+ Kxh8 7. dxe5 and win the "easy" Pawn ending. . ."

      should be

      "It's also straightforward to just capture the BRh8 with 6. Qxh8+ Kxh8 7. cxd4 and win the "easy" Pawn ending. . ."

      An error made in visualization from memory, without checking against an actual board. Sorry about that!

      Delete
    4. I love a nice king march. Nice use of pawns at the end in the game.

      Delete
  6. Replies
    1. That was quite a tour-de-force of analysis using the PLF system!

      It must be quite satisfying to find another player who "gets it!"

      Delete
    2. While I'm good at unearthing a system from almost nothing, I don't expect to be the first to become world champion with it. So yes, it is quite satisfying that other people pick it up.

      Delete
  7. Sometimes, your comments before seeing the solution are puzzling: "In the study room we need to eradicate trial and error." This comment seemed difficult to fathom. Now, once I saw the solution, that it's a mating attack, your comment made more sense in light of your whole post. When I look at the problem like it's a chesstempo problem, I also probably would have said "oh, it's a mating attack" and would have focused more on that, but I felt more psyched out by reading your comments first.

    So, ultimately, I did both. First, I thought, okay it might be a mating attack, but 1.BxB KxB, 2.Qf6+ Kg8, 3.Rh4 h5, 4.Rxh5 gxh5, 5.Qh6? (threatening Nf6 mate) f6!, so I said "Okay, maybe it's not a mating attack" and found after quite a while (still too long) 1.Be7 Rfe8, 2.Nf6+ winning an exchange and still holding a clearly winning position. After 1.Be7 Bxe5, 2.Nf6+ BxN, 3.BxBf6 with your standard winning, dark-square mating attack, and if 2...Kg7, then 3.RxBe5.

    Once I saw the solution (I had already seen that the given line was winning), I had to go back to this same mating line, and eventually figured out the win (after 2...Kg8) is 1.BxB KxB, 2.Qf6+ Kg8, 3.Rh4 h5, 4.Rxh5 gxh5, 5.Qg5+! (and not Qh6?) when Nf6+ will mate in two moves.

    Realistically, they should just tell you it's a mate since the other line I found appears winning as well. In terms of realism, my line would be just as devastating OTB to the opponent.

    I enjoy solving these major problems, just whoever is reading this remember that only your efficiency will carry over into your OTB rating. Most OTB games feature problems that are less than half as difficult as these ones, with the caveat that your stronger opponent will through more of these problems at you OTB and not require that you simply solve one of these problems and then you can go home and call it a day. No, they torture you with problems in 0.0 equal positions where you have to find the equalizer, so that the solution eats your clock and leads you back to square one.

    https://linuxguyonfics.wordpress.com/

    ReplyDelete
  8. A couple ways to think about this, One is that a piece is a sitting duck- easiest to take and another it is the most important piece in the opponents structure-the support beam if (re)-moved the house would come crashing down. (which piece is performing the most important FUNction.) Cheers.

    ReplyDelete
  9. @Robert -- Do you have an account on Goodreads ? www.goodreads.com. We need some good chess book reviewers . I would like to discuss chessbooks with you and that seems to be the appropriate place. My account name there is Takchess . thanks, Jim

    ReplyDelete
    Replies
    1. @ takchess:

      Yes, I have an account - I opened one to get a look at the Aagaard books you recommended. However, I haven't usually monitored it, since that was my only interest at the time. I just added you as a friend with a pending request. I'll keep a window up on Goodreads so we can communicate.

      Delete
    2. If connected to your email, you may get a notification once I send a message to you. Cheers, Jim

      Delete
  10. Replies
    1. PART I:

      Looks interesting but I only got a glimpse before the ad bot kicked in and demanded that I turn off my ad blocker or pay for the privilege. Sorry, but that ain't happening. I did find this article on Salon (not a site I frequent).

      Malcolm Gladwell got us wrong: Our research was key to the 10,000-hour rule, but here’s what got oversimplified.
      Yes, it takes effort to be an expert. But Gladwell based 10,000-hour rule in part on our work, and misunderstood.

      ANDERS ERICSSON AND ROBERT POOL


      Link: Malcolm Gladwell got our research wrong - Ericsson and Pool

      There is a reference to the book "PEAK: Secrets from the New Science of Expertise", Anders Ericsson and Robert Pool, Houghton Mifflin Harcourt, © 2016, ISBN 978-0-544-45623-5 (hardcover).

      There are quite a few good ideas to be gleaned from this book. Some excerpts:

      Delete
    2. PART II:

      "In any area, not just musical performance, the relationship between skill and mental representations is a virtuous circle: the more skilled you become, the better your mental representations are, and the better your mental representations are, the more effectively you can practice to hone your skill."

      "Note that these students weren't simply doing the same thing over and over again: THEY WERE PAYING ATTENTION TO WHAT THEY GOT WRONG EACH TIME AND CORRECTING IT. This is PURPOSEFUL PRACTICE. It does no good to do the same thing over and over again MINDLESSLY; the purpose of the repetition is to figure out where your weaknesses are and focus on getting better in those areas, trying different methods to improve UNTIL YOU FIND SOMETHING THAT WORKS."

      "TO EFFECTIVELY PRACTICE A SKILL WITHOUT A TEACHER, it helps to keep in mind three Fs: Focus. Feedback. Fix it. Break the skill down into components that you can do repeatedly and analyze effectively, determine your weaknesses, and figure out ways to address them."

      "Similarly, chess masters don't develop some incredible memory for where individual pieces sit on a board. Instead, their memory is very context-dependent: it is only for patterns of the sort that would appear in a normal game.

      The ability to recognize and remember meaningful patterns arises from the way chess players develop their abilities. Anyone who is serious about developing skills on the chessboard will do it mainly by spending countless hours studying games played by the masters. YOU ANALYZE A POSITION IN DEPTH, PREDICTING THE NEXT MOVE, AND IF YOU GET IT WRONG, YOU GO BACK AND FIGURE OUT WHAT YOU MISSED [and why you missed it]. Research has shown that the amount of time spent in this sort of analysis—NOT THE AMOUNT OF TIME SPENT PLAYING CHESS WITH OTHERS—is the single most important predictor of a chess player's ability.
      "

      "... this is exactly how chess players improve MOST EFFECTIVELY—by studying the games of grandmasters, trying to reproduce them move by move, and, when they choose a move that is different from what the grandmaster chose, studying the position again to see what they missed."

      "Despite the first word in the term "mental representation," pure mental analysis is not nearly enough. We can only form effective mental representations when we try to reproduce what the expert performer can do, fail, figure out WHY we failed, try again, and repeat—over and over again."

      "The plateau ... is common in every sort of training. When you first start learning something new, it is normal to see rapid—or at least steady—improvement, and when that improvement stops, it is natural to believe you've hit some sort of implacable limit. So you stop trying to move forward, and you settle down to life on that plateau. THIS IS THE MAJOR REASON THAT PEOPLE IN EVERY AREA STOP IMPROVING."

      Doesn't all of that sound very familiar to everyone here?!?

      There are some fascinating observations specifically regarding chess in the book. I highly recommend it if you are interested in the PROCESS of improving at chess.

      Delete
    3. this is a related but more chess specific paper : http://www.psych.utoronto.ca/users/reingold/publications/PDFs/Charness.Tuffiash.Krampe.Krampe.Vasyukova.2005.pdf

      Delete
    4. @ Aox:

      Thanks for the reminder - I had that paper downloaded back in 2013. The summary statement is interesting:

      There is no doubt that participation in tournaments and coaching sessions provide certain
      learning opportunities that are difficult to replicate in a solitary study environment.
      However, our data suggest that anyone with serious thoughts about becoming a title-level
      player will need to engage in several thousand hours of concentrated analysis and
      memorization of chess tactics and positions in order to build the knowledge base necessary
      to achieve regular success in highly competitive chess tournaments.


      Temposchlucker was right: we still have to do the hard work ourselves!

      Delete
    5. There is an other interesting statement in this paper: "In the older sample, only serious study alone was a significant
      predictor. "
      Meaning old folks dont benefit from instructions asf. so there is seemingly no secret knowledge to know.

      The question remains: what is to be learned and how. The term "deliberate training" is fuzzy
      What i did experience is that stronger player have a higher chess learning speed too. A 2200 semms to be 4 at least 4 times quicker in chesslearning than i. Of course the learning speed is related to the already existing number and quality of chunks. So i hope i will be able to work on that

      Delete
    6. Aox

      I disagree with your statement. Some people (from FICS server where I often play) says I am "a strong/pretty well player". And in practice I am quite weak, but I use some tricks other players do not use nor realize (I am legal tricks, not cheating). And I am chatting with some 2100-2200 players and they often say "I am quite slow and I avoid playing blitzes, because I need much time to play good moves". And the same phenomena takes place in my case. I like playing faster time control, but in reality I am quite slow player due to my very limited knowledge and chess understanding.

      However if you take into account the huge speed I could achieve while solviny your "salt minining" puzzles (mate in 1 easy), you could notice quite opposite picture of my person. I achieved even 50 MPS (mate per second), but only because I have repeated the easy puzzles about 120K times. And before that I had solved about 20K chess puzzles in a paper form (from #1 up to #3 and simple tactical shots as well). Do you think I am fast? What criteria do you use to evaluate my speed? There are many more questions to ask ;) :)

      Delete
  11. slightly off topic-- Shows the vision of Carlsen black to move. You might want to pass on solving it depending on time and focus.

    http://www.chessgames.com/perl/fen?fen=r2r4/pp3kb1/2p1n1pp/P4q2/1PQP4/2N1R3/5PPP/3R2K1

    How the game played out
    http://www.chessgames.com/perl/chessgame?gid=1477607&m=21.5

    Cheers, Jim

    ReplyDelete
    Replies
    1. Lots of "food for thought"! It is the Black Queen capture on f3 that shocks me! After that, the relevant lines crop up fairly easily/quickly.

      Carlsen has the amazing ability to use seemingly quiet, routine openings that (somehow) are "alive" with tactical shots - when he plays them. He ain't perfect, by any means, but he'll do as the closest human to approach that level so far.

      Delete
  12. Hi Tempo , want to try PLF on this one ? http://takchesschess.blogspot.com/2017/03/white-to-move-page-41-aagards-attacking.html

    ReplyDelete
  13. Somewhat more over the current target:

    The Tactical Detector - GM Daniel Naroditsky

    I found these assertions by GM Naroditsky to be interesting vis-a-vis our current discussion:

    "In my opinion, one of the principal reasons that players are unable to sharpen their tactical vision and calculation past a certain level is that they have a faulty tactical detector."

    "Your tactical detector sounds when you intuitively sense that there is some sort of combination or tactical sequence in the position."

    "The problem is that not all combinations spring up so naturally from the position. Hence, a weak tactical detector often prevents you from noticing latent characteristics of the position that make possible a hard-to-see combination or tactical blow."

    "'Alright, I get it. Tactical intuition is important. But how do I actually train my tactical detector?' [Y]ou can hone your detector simply by being mindful of its importance during the game. As soon as you are out of theory, you should spend some time on practically every move examining the position for hidden tactical elements. This might sound like a terribly cumbersome operation, but it will eventually become second nature. . . . Make no mistake: I'm not suggesting that you waste ten minutes on every move checking every conceivable sacrifice. Rather, you should simply be mindful of the features of the position that might give rise to tactics at some point in the future. This might be anything."

    "In conclusion, I'll mention that the notion of a tactical detector is my own invention, but the larger mechanism that it alludes to (i.e. the importance of general tactical awareness as opposed to concrete calculation) is an integral part of every grandmaster's thought process. Training intensively is one thing, but you can only unlock the fruits of your hard labor by adopting the right approach at the board."

    GM Naroditsky's concept of a "tactical detector" is a considerable part of the recent discussion here. [I'll note that Emmanuel Neiman's Tune Your Chess Tactics Antenna. Know when (and where!) to look for winning combinations was published in 2012. I suspect that GM Naroditsky was long past the point where that book would have made any difference in his rising skill, and therefore he was probably unaware of its existence and others like it.]

    We can't utilize tactical themes/devices unless we FIRST "see" the potential in the specific position for them to exist. IMHO, PLF and Chuzhakin's System are just two [similar] methods for constructing a tactical detector.

    ReplyDelete
  14. I think my vulture's eye is improving during actual play!

    I played the following game against the Thinking Machine 6 program. It's fascinating to watch the LoAs develop as the game progresses. TM6 covers the board with the potential LoAs in two colors, green for White, orange for Black. There was some "funny" business at the end - the program cheated to avoid a loss!

    I've played 3 games so far against this program, and have won all three games with strong attacks and lots of middlegame fireworks (which is NOT my normal style of play). I decided to start keeping the game score with this third game, and to use Fritz 11/Stockfish DD-64 to analyze, and show me what I missed. Surprisingly, I didn't seem to miss any major tactical shots!

    Link: Thinking Machine 6

    Robert Coble (1744 Elo; 1810 USCF) - Thinking Machine 6 [D02]
    Informal, 24.03.2017
    [Stockfish DD 64 SSE4.2 (60m)]
    D02: 1 d4 d5 2 Nf3 sidelines, including 2...Nf6 3 g3 and 2...Nf6 3 Bf4

    1.d4 d5 2.Nf3 Nf6 3.Bf4 Nc6 last book move 4.h3 Secures g4 4...e6 5.e3 Bd7 6.a3 Prevents intrusion on b4 6...Be7 7.c4 dxc4 [Better 7...0–0!? slightly better for Black and Black is still in the game] 8.Bxc4 Slight advantage for White 0–0 9.Nc3 Rc8 10.Rc1 g6 11.0–0 Kh8 12.Qd3 a5? [12...Na5!?±] 13.e4 White prepares the advance d5 13...Kg8 [13...a4!?±] 14.Rfe1 Rb8 15.Nb5 Rc8 [15...Ne8+-] 16.d5 exd5 17.exd5 Bf5 18.Qd1 Nb8 19.Na7 Re8 [19...Bd6 doesn't change the outcome of the game 20.Bd2+-] 20.Nxc8 Qxc8 [20...Bxc8 the last chance for counterplay 21.Bb5 Bd6 22.Bxe8 Bxf4+-] 21.d6 Bxd6 22.Bxd6 cxd6 [22...Rxe1+ doesn't get the cat off the tree 23.Qxe1 Nc6 24. Bh2+-] *** 23.Bxf7+! Decoy: d5 23...Kxf7 [23...Kxf7 10.22/10 7 ] 24.Rxc8 Bxc8 [24...Rxc8 doesn't improve anything 25.g4+-] 25.Qxd6 [Better 25.Ng5+ and White wins 25...Kg7 26.Rxe8 Nxe8+-] 25...Rxe1+ 26.Nxe1 Nbd7 [26...Nc6 does not save the day 27.Qc7+ Ne7 28.Nf3+-] 27.Nd3 [Better 27.Qc7 makes it even easier for White: 27...g5 28.Qxc8 b6+-] 27...Ne4 [27...b6+- hoping against hope] 28.Qd5+ Kf8 29.Qxe4 Nf6 30.Qf4 [30.Qc4 seems even better 30...Bg4 31.hxg4 Kg7+-] 30...Bf5 31.Qd6+ [31.g4 and White can already relax 31...a4 32.gxf5 gxf5+-] 31...Ke8 32.Ne5 [Better 32.Qxf6!? and White can already relax 32...Bxd3 33.Qe6+ Kd8 34.Qd6+ Kc8 35.Qxd3 a4+-] 32...Ne4 33.Qb8+ Ke7 34.Qxb7+ Ke6 35.f4 h5 36.Qd7+ Kf6 37.Qf7# 1–0

    Thinking Machine 6 actually "played" 37. ... Kxe5??? (Illegal move - game is already over!!!) 38. fxe5 (Capturing the King???) Game over - Draw??? I guess it is truly artificial intelligence: it has learned how to break the rules and cheat to avoid a loss.]

    My approach was to slowly build a classical center with more space under my control and then to look for tactical shots once development was completed. I began "seeing" some of the tactical possibilities prior to playing 15. Nb5. The BNc6 only has one square (b8) to go after the d-Pawn advances, and this entombs the BRc8, allowing me to win the Exchange. The Black Queen has the functions of guarding the c-Pawn and the BBe7. Black can cover the BBe7 with the Rook but there is the Pawn advance to d6 which uncovers the possibility of unleashing the WBc4 against the Black King on f7 with check, with a discovered attack on the Black Queen. After 20. ... Qxc8 I knew I was clearly winning in all lines.

    "Seeing" the PoPs, LoAs and Funs made it much easier to "see" how things might turn out!

    ReplyDelete
  15. Another indication (an estimate) of (possibly) increasing strength based on "solving" 76 problems at Elometer:

    Link: Elometer

    Based on your move choices, our estimate of your Elo rating is 2250, with a 95% confidence interval of [2112...2388].

    I didn't spend a whole lot of time on the individual problems. Some of them I could solve almost instantaneously; others took 2-3 minutes.

    I take these kinds of estimates with a 2250 lbs. bag of salt (and ASSUME that reality is probably lower than the lowest value of 2112), but it does seem to indicate that I MAY have improved as an adult (since I didn't start playing chess until my early 20s) and I'm now 69 years old. My last OTB Regular USCF rating from 1975(!) of 1810 (Elo calculated 1744) and current Quick Rating of 1477 certainly would seem to indicate that I have NOT improved AT ALL over the years.

    Maybe some of that PLF system is finally beginning to penetrate my hard head!

    ReplyDelete
  16. The tactical detector feels like marketing. I feel that these insights are most likely shared by others. It is an interesting question. Can we determine if their are tactical advantage to us in a position without full calculation ?

    ReplyDelete
  17. @ Jim:

    PART I:

    I don't think it's an "either-or" consideration. The purpose of the PLF approach (I hesitate to classify it as a "system" because it does not aspire to that broad of a classification scheme) is to provide hints ("clues"; "cues") of what MIGHT be available in a given position, thereby directing the attention toward favorable tactical possibilities. As Aox pointed out, it is a method for explicitly identifying weaknesses which may be capable of exploitation by standard tactical themes/devices, similar to the Chuzhakin System, but not intended to be as extensive in scope as that system.

    "Chess is 99% tactics" - Richard Teichmann Amended by Dr Dave at the Exeter Chess Club: ...Although it has often seemed to me that the remaining 99% must be all strategy :-)

    “In general I consider that in chess everything RESTS on tactics. If one thinks of strategy as a block of marble, then tactics are the chisel with which a master operates, in creating works of chess art.” – Tigran Petrosian

    "In [the Chess Master] this antagonism [between combination play and positional play] is completed by position play. By combination the Master aims to defeat the false values, the true values shall guide him in his position play, which in turn shall bring those values to honour. The Master is like a man in a learned dispute who knows sophistry but does not make use of it, except for the purpose of exposing the sly subtleties of an artful opponent who disputes a true, sound, vigorous thesis with mere trickery." - Emanuel Lasker

    Mister Lasker's observation was always opaque to me (at least until recently), in part (I think) because I could never "see" his point in a practical sense. (GM Petrosian and GM Teichmann allude to the same thing, but not quite as explicitly as mister Lasker.) In mister Lasker's view, everything in chess depends on a valuation. A true valuation, one that is accurate and corresponds to the requirements of the position, can be supported and "proven" (in a chessic sense) by the availability of tactical themes/devices for that purpose. By the same token, a false valuation, one that is inaccurate and does NOT correspond to the requirements of the position can be defeated and "proven" false by the availability of tactical themes/devices for that purpose.

    When pursuing the true values, there is almost an air of inevitability about the sequence of moves. On the other hand, when pursuing false values, there is a sense of trickery, "traps" if you will, that depend on the opponent overlooking something. In the final analysis, it all depends on who "sees" the most (or at least the last thing that causes the valuation to swing to one opponent or the other).

    ReplyDelete
  18. PART II:

    The winner is the one who makes the next-to-last mistake. - Savielly Tartakower

    In either case, it is the player who "sees" more of the available resources who is most likely (barring unforeseen mistakes later in the game) to prevail in the long run.

    IMHO, we CAN "see" more of the potential tactical possibilities (advantages AND disadvantages) by looking for the weaknesses FIRST, supplemented by short forays into the future (calculation) utilizing the various known tactical themes/devices. What is intriguing about the PLF approach is that those short tactical sequences often become readily apparent as soon as the PoPs, LoAs and Funs are recognized and enumerated. It's not calculation per se, but a variation thereon. GM Beim identifies this as an integral part of making plans in chess. I think of it more as becoming AWARE, sensing perhaps the critical points and lines, without indulging in the Kotovian (Pavlovian?) approach of first selecting candidate moves and then meticulously performing a tree search, visiting each node in the tree once and only once.

    I don't think like a tree - DO YOU THINK LIKE A TREE? - GM Anatoly Lein to GM Jonathan Tisdell in Improve Your Chess NOW

    NOT ME! I want to think like a hungry vulture!

    ReplyDelete
  19. Chess is the triumph of the intellect and genius over lack of imagination; the triumph of personality over materialism - Richard Reti - I becoming very interested in the Initiative and punishment of over materialism. The Steinitz use it or lose it law. Working on the second chapter in the aagaard book about this. Catching up to Tempos blog slowly so i am a little hazy on all the attributes of Plf as I am only caught up to may 2016 at this point.

    ReplyDelete
    Replies
    1. You might want to consider to skip a little by starting at december 3, 2016. That way you get the fastest complete picture of plf without missing anything important. Then you might come back later and read the skipped part, which gives you a picture of the preliminary struggle to get to the plf-system.

      Delete
  20. No worries, a new post is in the making. But it takes time to ponder.

    ReplyDelete