Saturday, December 03, 2016

What dit not work?

What didn't work out:
  • The board vision exercises as presented in my sidebar, AKA "the salt mines"
  • Decompose the solution for an hour and visualize the results for a few minutes
  • Imagine the future position with the aid of a verbal narrative
  • Identify the function of the pieces and visualize this
  • Identify the tactical themes of a combination and visualize this
  • Verbal formulation of the purpose of a move
  • Solving problems extreme fast
  • Solving problems extreme slow
  • Solving extreme many problems
  • Doing extreme many repetitions
  • Solving without repetition
  • Playing blindfold chess
  • Playing correspondence chess
  • Playing a lot of blitz
  • Playing a lot of rapid
  • Playing a lot of slow OTB
  • Stoyko exercises
  • Extreme difficult problems
  • Extreme easy problems
  • All kind of scanning techniques
  • All kinds of thought processes
  • All kinds of board vision exercises
  • Exercises for specific pieces
  • Themed exercises
  • Random exercises
  • With clock
  • Without clock
  • Visualization of covered squares
  • Imagining the aura of the pieces
  • Targeting my weaknesses in isolation
  • Targeting my weaknesses with an all out approach
  • Refusing  draw offers
And I exaggerated each exercise into absurdum in order to be able to draw a definite conclusion.
There are a few more thing I tried, but I don't feel like recherching that now. Let's focus on what works instead.

Only three methods showed any progress:
My efforts in the salt mines were inspired by the the result I had by playing Troyis. The problem with this method is, that it is too specific. I improved, but the improvements are irrelevant to the game. I analyzed extensively where I spilled points when solving problems at CT, and in only about 2% of the cases it had to do with the motor skills you learn while salt mining.

Mate in one
Working through Papa Polgars brick changed my vision. In stead of looking at the pieces, I started looking for the squares they cover, the "aura" of the pieces. I have done a whole host of vision exercises, and got improvement there. But again, those improvements are rather irrelevant to solving the kind of problems as presented by CT. It helps to visualize future positions, but that simply is not the main reason I spill points. Maybe in 3% of the cases a lack of visualization skills are the cause of spilling points at CT.

Steps method
The steps method just worked. But I couldn't find a follow up, and failed sofar to write my own. Everything I learned from it was relevant to the kind of problems CT presents to me.

The diagnosis
I know the tactical themes thoroughly. The problem is that they often are not popping up while scanning the board. The triggers do not fire. I have a lack of cues that activate the retrieval. If I guide my focus consciously, I'm able to retrieve the relevant tactical themes, but conscious thinking is extremely slow. In stead of spilling points by making errors, I spill then points by using time.

The remedy
The Steps method shows us the way. I have felt a long time that tactical prowess should be acquired in the same way I learned how to drive a car. The instructor provides the knowledge, and in just 50 hours or so, the unconsciousness works its magic, and internalizes it. That's how the Steps method worked.

What I'm going to do, is to write my own follow up. I have long neglected the knowledge I thought I already knew. But when you dive deep into a tactical motif, there is a whole lot to be discovered. I let me be guided by Martin Weteschnik's Chess tactics from scratch. When I studied his chapter about the reloader, I immediately recognized a few reloaders in the problems afterwards. That is how it should work. I will start with the study of the pin. Deep study of the pin should acquire me a cue or two, don't you think? That's the theory.

It must work. It is logical that it should work. It is the only thing I haven't tried yet.


  1. PART I:

    While I agree with your assessment that it is logical that your new approach SHOULD work, I am not as convinced that "it MUST work." [Emphasis added.]

    I've been working through various tactics books (again), trying to "cement" in the idea of performing a complete "survey" of the entire position before "jumping" to any conclusions about potential solutions. As a result, I have noticed that my failures usually are caused by failure to observe EVERYTHING before focusing in on the perceived critical piece(s)/motifs/themes. Here's an example, taken from 303 Tricky Chess Tactics - A great training tool for players, problem 121, in the section on Rook Move Discoveries. (When working on problems, I try to ignore the categorization scheme, simply because it narrows the "vision" down too much too soon.)

    FEN: 8/4N2k/2b5/P7/2r4P/6p1/5p2/3R1K2 b - - 0 1

    The given solution (in keeping with the classification scheme) is:

    "1. ... Bb5 (threatens 2. ... Rc1+) 2. Kg2 Rc2 (threatens 3. ... f1/Q+). It's two too many discoveries. White has no defense."

    My own analysis was to consider that Black can immobilize the White King and then threaten checkmate.

    1. ... Bf3 (The White Rook cannot leave the first rank, and any White Rook move allows 2. ... Rc2, followed by 3. ... g2#; a White Knight move allows capture of the White Rook and it should be relatively simple to eliminate White's a-Pawn and h-Pawn and then [eventually] to track down the White Knight.)

    Just out of curiosity to "see" what I might have missed, I put the position into Stockfish and let it run overnight. I was really surprised to find all the different tactical possibilities that I completely overlooked, given below in the order of preference by Stockfish.

    1. ... Rxh4 (Mate in 12)
    1. ... Rd4 (Mate in 12)
    1. ... Be4 (Mate in 16)
    1. ... Bb7 (-+ 80+)
    1. ... Bf3 (-+ 80+)
    1. ... Ba4 (-+ 80+)

    (There were other winning possibilities, much lower in evaluation, so I didn't examine them.)

    1. It must work because it is the last option. I tried everything I could think of, and this is the only thing I haven't tried.

    2. Yes, pins should (or must) work, but only for a small (but statistically important!) degree. The real breakthrought should be made with the analysis of tactical ideas and concepts based on hundreds of puzzles. In conclusion: the chess player has to build his own chess tactical system (radar) to recognize tactics and find out the best (or at least sufficient) option to continue from the specific position. It requires hard work, but in the end... it is THE ONLY way we can make significant change/improvement.

    3. @Robert

      You just showed the idea of PRACTICAL solution! Most players would play (and think) like that! Only (tactical) masters would consider such an "abstract" move like Bb5 (!). And to be honest - we simply have too little knowledge (conclusions) from the concept of interception and discovery. At least I have this feeling and problems related to this specific topic.

      We should (have to) answer to lots of questions like:
      1) When it is better to use/set up the discovered attack than normal attacking setup?
      2) Why the discovery (interception) should be set up?
      3) How to recognize if this type of attack is (much) better than the traditional one?
      4) When such an approach is not correct? What elements are critical when we want to distinguish between this kind of attack and the traditional one?

      There are many more questions, but we can start with these ones and the other ones will magically appear very soon...

  2. PART II:

    Why did I NOT "see" or consider all of those other possibilities? Because I stopped "looking" as soon as I found the first "solution" that I KNEW was winning AND that I KNEW I could win if I had to play it out, even against a master. Also, given the level of the problems in the book, I KNEW that I didn't have to find a mate in 12 or a mate in 16 in order to "solve" the puzzle. I quit looking once I found a satisfactory (to me) solution. NOT AN OPTIMAL APPROACH FOR LEARNING!

    I think (and will have to investigate my erroneous thinking a lot more) that I am simply "satisfied" with a superficial or "obvious" solution and do not force myself to investigate ALL possibilities before moving on to the next problem. As a result, I do not derive the maximum value from solving problems in this faulty mindset. I AM capable of "seeing" these various possibilities, once I step back into the position looking at ALL possibilities, not just the "obvious" ones.

    I could call it "laziness" but I think it is merely an indication that I do not discipline my thinking sufficiently. And why not? Because it requires an expenditure of time and energy that I do not want to invest. As a consequence, I don't get all that I could out of "solving" these problems, and I don't gain any significant improvement.

    I recall (and can relate to) Kotov's comment [Think Like A Grandmaster, pg 18] about his deficiencies (after playing successfully in two Moscow championships):

    "Most of all it became clear to me that my main trouble was not superficial knowledge of the openings or poor endgame technique, but my poor understanding of the middle game. My worst fault was inability to analyze variations. . . . It became clear to me that I had a lot of hard work to do on mastering the technique of analysis. . . . A lack of desire really to go into concrete variations thoroughly, a vague wandering about, those are my characteristic mistakes in my play in the 1936 Moscow championship. . . . I was not able to find a single one of the variations and combinations at the board. . . . To what a laughable extent my thinking is based on general principles and plans."

    I can certainly relate to that assessment!!!

    1. I have analyzed a lot of tactical problems before my break. It supplies just the kind of knowledge I'm after. Yet I got not much out of it. The knowledge didn't stick. Yesterday I discovered why. New knowledge needs a coat rack. So you can hang the knowledge there where you can retrieve it. What you can learn by analysis is so divers, that it isn't easy to decide where to hang it. But if you don't hang it somewhere, you lose it.

      By using the pin as a coat rack, I can hang all knowledge I discover from analyzing pins at the pin rack. For instance, there are two important winning methods associated with a pin.

      First: A pinned piece is a bad defender so attack the piece that is bad defended.
      Second: A pinned piece is immobilized, so attack it.

      If I recognize a pin, it can take a minute or more for I consider the second option. The knowledge is there, but not readily available. The cues for retrieval are weak.

      This means that the knowledge must become self evident. Seeing a pin = looking for option 1 and 2. There is a whole lot of knowledge associated with pins. Thing you want to pop up immediately when you see a pin, and not after a minute or so. The cues must be strengthened. I'm going to work on that. Step by step. First the pins. A post is already in the making.

    2. I have recently solved some dozens of puzzles related to pins. To me they are the most valuable tactical concepts (motifs). The other is double attack idea.

      I think pin concept is not learnt (studied) deep enough to bring real benefits. The pin motifs should be scrutinized (analysed) into pieces and for all the specific configuration - some dozens of puzzles are needed. If you look much closer to the positions you can convince yourself how many positions contain the idea of pin.

      And there are some kind of pins: absolute (against the King) and relative (against any pieces of higher value, but never the King). And we have pins and counrerpins - sometimes called "cross-pins". Most chess studies use the concept of "crosspins" to show amazing idea of double attack.

      I am looking forward to your next articles related to pins. I would like to make a small contribution to this topic as I think so far I have solved about 600 chess puzzles related to pins... and probably I can critically comment on your ideas.

  3. I (and several other avid followers of your blog, I'm sure) eagerly look forward to your new posts and how you approach "pinning" your tactical knowledge on the "coat rack!"

    1. It's not so easy to write about pins in a non boring way. Which is probably the reason why we avoid the acquisition of this kind of knowledge altogether. To prevent scruples we tend to think we already know it. But I will do my best. I'm struggling with the best approach though.

  4. Perhaps the needed "hint" of an approach is to be found in what you wrote previously. You already possess the generalized knowledge of pins: the definition of a pin (absolute and relative), the three components of a pin (target, pinned piece, pinning piece), ways to exploit the pin (add more attackers on the pinned piece, capture of or distraction of defenders of the pinned piece, utilization of the loss of function of the pinned piece by moving to unprotected squares). It seems that your primary concern is the "cues" that will trigger the "idea" of setting up or preventing or exploiting a pin which may be obvious (or hidden) within a given position.

    From your comment above, it appears that it is primarily a matter of exploring all possible means of exploiting a pin. Quoting you:

    For instance, there are two important winning methods associated with a pin.

    First: A pinned piece is a bad defender so attack the piece that is bad defended.
    Second: A pinned piece is immobilized, so attack it.

    I utilize two different sources for ideas regarding pins (and some other tactics as well).

    A. Nimzovich (My System, chapter 7. The Pin) identifies the defensive power of the pinned piece as imaginary. Consequently, the opponent can confidently place his other piece(s) en prise to the pinned piece. This is difficult to do for several reasons: (1) the pinning player "sees" the pinned piece as having the capability of capturing, even though it is merely an illusion; (2) the squares that might be otherwise occupied are empty, and it is more difficult to envision occupying empty squares (with purpose) than to capture on an occupied square. (3) the pinning player "sees" the possibility of losing a piece that is en prise to a pinned piece at some future point in the game, and so does not consider that possibility as desirable. Nimzovich states that every immobile, or even weakly restrained, piece tends to become a weakness, so it is most natural to "pile on" the pinned piece (the weakness). The primary problem in winning the pinned piece is to prevent the unpinning of the pinned piece. The primary approach is to first "pile on" attackers directly on to the pinned piece, with the ideal attacker (eventually) being a Pawn. The secondary approach (taken whenever there is an adequate number of defenders) is to thin the ranks of the defenders through direct capture or diversion of the defenders. The exchange on the pinning square is often facilitated by substitution of a fully pinned piece for a half pinned piece (one that retains at least some possibility of moving). Any relative superiority (determined by simple counting) remains in effect after such a substitution. Another motif is the gain of a tempo. It is often the case that one of the attackers on the pinned piece is itself counter attacked. In that case, it is useful to capture the pinned piece PROVIDED that the pin is maintained and fresh troops brought to bear on the newly pinned piece. I'll forgo Nimzovich's discussion of unpinning.

    Dr. Lasker describes the "encircling motif" as a combination (not in the chess sense) of two things: (1) immobility, and (2) superior force at a given point. That which is immobile must be attacked. He describes the obstruction motif as the removal (or reduction) of mobility through either obstruction or by pinning. Each side tries to maximize maximize activity; this is defeated by pinning.

    Both of those summaries are general ideas, more at the philosophical level than at the practical level. When trying to solve "THIS" specific problem/position, generalities are often useless or too vague to be really useful.

    1. The fact that I understand what Nimzovich and Lasker are saying makes me feel intelligent. It gives me the feeling that I know what a pin is. The problem is those situations where I recognize a pin, but it still takes a minute or more before the idea pops up how to exploit it. I consider that to be due to "weak cues". The cues that act as trigger to retrieve the knowledge how to exploit a pin takes a lot of time to fire.

      Somehow the connection between seeing the pin and the knowledge how to exploit it must become closer. Pin swoosh Exploit. Target defended swoosh Harass defender.

    2. The difficulty with the PIN idea lies in the fact - you have to understand the concept very well and practice it hundreds of times to make solid conclusions.

      The PIN may be divided into two parts:
      1. Easy - the PIN against the King. Here you cannot move away with the pinned piece (other then moving on the line of pinning), because it would be ILLEGAL move. Nobody can refute it, no matter if now or in hundred years.

      2. Hard - the PIN against any other pieces than King. Here you have to know if the pin can be broken (technically it may be ALWAYS broken) and what benefits (profits) it brings.

      And the biggest challenge is to merge these two parts into the SPECIFIC situations on the board. If is not that easy as it seems, simply because this motif (tactic) is connected with another ones and you have to evaluate if the pin idea may be possible to exploit or not yet.

      BTW. I have finally bought Martin Weteschnik's "Chess tactics from scratch" and I am going to study it. This way we can change our ideas, make observations and conclusions. In my view the pin concept may be a way more difficult (and deeper) than I have been thinking of it... all my chess life!

  5. How are you measuring progress to determine whether or not something worked for you?

    1. The rating at CT is a good objective indicator. Besides that, I look around in my own mind if there are changes. What I'm after is that knowledge about the position that normally pops up after a minute or so, will pop up in seconds. I'm talking about simple logical concepts here. Like when you see that a target is defended, that you should have a look at if you can harass the defender. That should pop up in seconds, not in minutes. There is no need to spill a minute while your mind is wondering around.

      Usually I start pretty euphoric when I try a new method, but after two months or so, reality sets in. Then I continue another two months while I know it is not working in the hope I'm wrong. Only when I'm totally convinced that another six months of training will not bring me any measurable progress, I abandon the method.

  6. I found the following position in Weteschnik's Understanding Chess Tactics, pg 183, in the chapter on Status examination. Weteschnik gives this game without any context. I found the entire game on (Oleg Romanishin vs Harold James Plaskett, Simul, 30b (1977) · English Opening: Anglo-Indian Defense. King's Indian Formation Double Fianchetto (A15) · 1-0 ) A comment on the game gives a little more background: "The game was played in a simultaneous display shortly after Romanishin had made a tournament appearance at (I think) Hastings." Haskett was only 17 at the time, and wouldn't get to the Grandmaster title until 1985. So, perhaps he can be excused for blundering against a seasoned veteran.

    FEN: r2q1rk1/pb1p1pb1/1pn1pn2/6Q1/2P5/1P4P1/PB3PBP/RN1R2K1 w - - 0 14

    The entire game is as follows:


    [Event "Simul, 30b"]
    [Site "London ENG"]
    [Date "1977.01.16"]
    [EventDate "?"]
    [Round "?"]
    [Result "1-0"]
    [White "Oleg Romanishin"]
    [Black "Harold James Plaskett"]
    [ECO "A15"]
    [WhiteElo "?"]
    [BlackElo "?"]
    [PlyCount "27"]

    1. c4 Nf6 2. Nf3 b6 3. g3 Bb7 4. Bg2 g6 5. O-O Bg7 6. b3 O-O 7. Bb2 c5 8. d4 cxd4 9. Qxd4 Nc6 10. Qh4 h6 11. Rd1 g5
    12. Nxg5 hxg5 13. Qxg5 e6 14. Rxd7 1-0

    Following Valeri Beim's suggested position examination procedure, the first thing to do is evaluate the material balance: Black is ahead slightly (Knight for two Pawns). White is slightly behind in development (noting the White Queenside pieces). NONE OF THIS MATTERS!

    At this point, we can move into the subject at hand: the pin.

    Weteschnik gives the following verbal "analysis":

    (1) The Qd8 has to defend the Nf6. If Black loses the Knight, he will be mated.

    (2) The Nf6 is protecting the Bg7 against mate. [THIS IS ANOTHER PIN WHICH IS N-O-T AS OBVIOUS AS THE PIN OF THE Bg7!]

    (3) The Bg7 is pinned by the Qg5 against the Kg8 and cannot recapture on f6 as long as it is still pinned.

    (4) The d7-Pawn is undefended as all its defenders are occupied with other important duties [preventing mate]. The d7-Pawn is attacked by the Rd1.

    "Taking a look at the list you will easily see that White can snatch the d7-Pawn with 14. Rxd7!"

    Prior to 13. ... e6?, it should have been "obvious" that the Nf6 is protected adequately by the e7-Pawn and the Qd8. By moving 13. ... e6?, Black removed one of the defenders, creating a situation referred to by Heisman as B.A.D. (Barely Adequately Defended). There are now two defenders, but both have the specific function (the function motif of preventing mate on g7. White can take advantage of both pins by attacking one of the defenders, Qd8 with 14. Rxd7!, leaving it nowhere to continue defending Nf6.

    So, in addition to "seeing" the obvious pin on Bg7, the astute player must see the more obscure pin of the Nf6. For ME, this is exactly what I was describing when I stated that the "auras" emanating from line-moving pieces (Q, R, B) must be followed from each piece all the way to the edge of the board in order to "SEE" the various connections and interrelationships that exist (perhaps several moves into the future). It also illustrates the importance of "surveying" the pieces, and observing the various functions that each piece performs in the specific position. Only then can we "see" what is really going on in a position and take maximum advantage of the available tactical opportunities.

    IMHO, it is only by "seeing" what is potentially available in addition to the actual situation on the board (the "hidden" pin in addition to the visible pin AND the functional relationships) that we can progress in our quest for tactical vision based on cues.

    1. I forgot to point out something that is not obvious about the Romanishin-Plaskett position. It ASSUMED that IN GENERAL the target piece is of higher value than the pinned piece. This is mostly true when the proper approach is to pile up on the pinned piece in order to capture it. However, if the target is unprotected (or insufficiently protected), then it does not matter what the relative value of the target is. In the example, the target of the obvious pin is the King; consequently, the approach is to (somehow) pile up on Bg7. However, in the case of the "hidden" pin on Nf6, the target is Bg7 (admittedly it would be mate if the Queen can capture on g7) but in cueing the appropriate signal, the value of Bg7 is NOT greater than the value of the pinned piece (Nf6) nor of the pinning piece (Bb2). Ergo, the point I have made in the past that when "seeing" what is available, the piece value(s) that are involved in the interrelationships is irrelevant. All that really "counts" is the function(s) which are involved. We apply the point value "counting" procedure only when the sequence reaches quiescence AND we are playing for gain of material. In a play for mate, the point values of the piece(s) involved are irrelevant.

  7. As another example from Weteschnik, consider this position:

    FEN: 4r1rk/2pn2b1/1p1p1q1p/3P1p2/1PP1pP2/7R/2Q1B1PP/2B2RK1 w - - 0 27


    [Event "Soviet Union"]
    [Site "Soviet Union"]
    [Round "?"]
    [Date "1946.??.??"]
    [White "Ratner, Boris"]
    [Black "Pokorny, Amos"]
    [Result "1-0"]

    1.d4 Nf6 2.c4 d6 3.Nc3 g6 4.e4 Bg7 5.Nf3 O-O 6.Be2 Nbd7 7.O-O e5 8.d5 Nc5 9.Qc2 a5 10.Bg5 h6 11.Be3 Ne8 12.b3 f5 13.exf5 Bxf5 14.Qd2 Kh7 15.Rac1 b6 16.Qd1 Nf6 17.Nh4 Nfe4 18.Nxf5 Nxc3 19.Rxc3 gxf5 20.Bc1 Qf6 21.Rh3 Rae8 22.Qc2 Kh8 23.a3 Rg8 24.b4 axb4 25.axb4 Nd7 26.f4 e4 27.Bb2 Qg6 28.Bh5 Qh7 29.Bxe8 Rxe8 30.Rg3 Nf6 31.Qa4 Rf8 32.Qa1 Qg8 33.Bxf6 Rxf6 34.Qxf6 Bxf6 35.Rxg8+ Kxg8 36.Rd1 Bc3 37.Rb1 Kf7 38.Kf1 Bd2 39.g3 Bc3 40.Ke2 Bd4 41.Rb3 Kg6 42.Ra3 b5 43.cxb5 Kf6 44.Ra8 Kf7 45.Rc8 Bb6 46.Rh8 Kg6 47.Re8 1-0

    The "obvious" pin is of the h6-Pawn. After 26. ... e4, the h6-Pawn is still protected by two pieces (Qf6 and Bg7). One of those two pieces has the function of preventing White from capturing the h6-Pawn WITH CHECK because Black has no other means of removing a check against the Black King on h6. (That's not very subtle, BUT it is "distracting" in terms of "seeing" the solution!) So, White can nonchalantly plop his Bishop on b2, attacking the Black Queen, then the Black Bishop, and finally the Black King - another kind of "pin." If Black foolishly captures 27. ... Bb2 (assuming via "counting" on that one square that he has the numerical superiority on b2), the "punishment comes as with 28. Rxh6+!, forcing the now overloaded Bg7 to capture, and thereby removing the protection of the Black Queen on b2. the "counting" must be done on BOTH squares, not just on one square. Black "dodged the bullet" with 27. ... Qg6, only to run afoul of 28. Bh5 skewering the Black Queen (g6) and Black Rook (e8). In spite of this, Black soldiered on until the 47th move.

    I'm not sure how one would analyze this position in terms of visual cues that would automatically trigger the appropriate play. Pinning is obviously involved, but it is hard (FOR ME!) to "see" at a glance (without logical reasoning concerning the functions of the involved pieces) how to connect the two different theaters on opposite sides of the board into a coherent approach.

    1. E-r-r-o-r: That statement above should read: "If Black foolishly captures 27. ... Qxb2. . .". It was the conflation of Black Queen capturing on b2 and the actual move; sorry about that.

  8. "What did not work" contains a lot of things that actually did work for some time.
    The surprising thing is rather: more of the same did not work at some point.

    Really, when you did your first few tactical exercises, with no special kind --> just any tactics, somewhere somehow --> really, didnt it work for some time?
    When I did my first 1000 tactics at chesstempo - I felt much sharper afterwards.
    Even Aox, mostly pessimistic about adult improvement, would not deny that the very first tactic puzzles wouldnt improve you.
    So, the thing is rather: why do trainings that helped us initially, didnt help us later? Deminishing returns everywhere. And then, even if we startet different, more sophisticated training methods, improving and refining on the training methods that once worked for us --> no further improvement.
    The "salt mines", they helped me during a training of about 6 weeks. After that, it took 1-2 years to get higher. The amount of salt mine was suddenly 10-fold. And I gained a bit, but not as dramatic as in those first 6 weeks when I did them.
    And after that gain - I really plateaued, too.

    There is one method, that helped me after the plateauing point, and it is not really listed in your list: "Winning more games without playing any better."
    Yes, it isnt really improvement in chessability, but... to win more games with the same ability is fun nevertheless.

    So how to do this?
    a) in my case, I had not so good openings (statistically proven by chesstempo database), and I could exchange my repertoire for a more promissing opening repertoire.
    b) I improved on my time management. In a tournament game, I now play the first moves within a minute if I know the opening. Thus, in one case, I needed 1 minute for the first 10 moves, while my opponent needed 20 minutes. The strange thing - he knew the opening, too. He had played it in games before, and at least in some of his blitz games he had the position on the board, too.
    And a lot of people waste time like this. I save my thinking time for when I need it.

    There are other little improvements: Offer a draw if you are slightly worse, dont be shy. Even if he declines, the word "draw" is spooking in his mind. It is like poisoning the objectivity of the opponent. When you improve a little bit, but are still a bit worse - suddenly he offers the draw himself.
    Or if you dont want your opponent to play the sicilian, take a big fat book about the sicilian with you to the club, place a glas of water on top of it: Chances are, your opponent might start thinking you are a sicilian player yourself, and thus decides to play his second-choice opening.

    It is the tiger-chess kind of thinking: "Win - no matter how!"
    The list of how to improve without playing any better is long, and I stop here. But: if you have not tried, then there is some potential here, too. And at some point there will be deminishing returns here, too, of course! I guess I gained 100-150 elo point just with issues like that. I am not playing any better, but I am better rated that equal skilled players.

    1. "So, the thing is rather: why do trainings that helped us initially, didn't help us later? Diminishing returns everywhere. And then, even if we started different, more sophisticated training methods, improving and refining on the training methods that once worked for us --> no further improvement."

      I have tried to categorize the problems in my database of failures, and I didn't come up with a useful categorization. The problems are so totally different. That means that it is quite difficult to learn something from one position what is useful in another position. The rate of transfer from one position to another is quite low.

      With the motifs of Lasker, we have for the first time a connection between one combination and another. Knowledge that we can hang on a branch of the tree of motifs can become useful in many positions.

  9. P.S. wanted to add: everything that "worked" was probably the initial improvement that happens when we start training. Instead of Polgars Brick, you could have solved with "salt mines" first, just to discover that afterwards you would not improve with Polgars Brick (because it would be a follow up of a similar training like "salt mines").

    I thought of one thing that improved me quite a bit: endgame studies. But not chesstempo´s basic endgames, but rather a good book, that explaines well. For a start I can recommend my endgame threads at chesstempo. There I put down what I learned, and what I see as common mistakes your opponents do all the time. Which means - if you come down to an endgame, then these kind of mistakes happen often. If you know you are good in endgames, you can often trade down more often so you see more often endgames. Then you win the endgames.

    Endgame knowledge doesnt only let you win (or rescue into a draw) more often, but I feel it improved my play in the middlegame as well. Because many considerations what kind of endgame might happen influence already your middlegame.
    We heard that Capablanca handled the middlegames already like endgames, and heard similarly often such things from other grandmasters. Nevertheless, I often never saw a middlegame like an endgame - until I knew more about endgames. Endgames is about handling your pieces well. Rooks are poor defenders but good attackers. You adjust already in the middlegame how you play because of the kind of material you are left with. 2 knights means defensive is better, 2 bishops means pushing pawns early is better, and that is not only true when the 2 knights are your last 2 pieces, but it makes sense in the middlegame, too. I often win middlegames by letting the opponent have the pair of bishops and then play defensive in a "having the knights"-manner. And suddenly, I win, because my opponents "can not tell a bishop from a knight", all they know is "pair of bishops" is better, but are not able to play accordingly. Many, many more issues like this.
    And why does it work? Well, because you are not really becomming better, but simply by acquiring more knowledge. And the endgame knowledge is an issue, most people are so poor in handling it, that already a little bit extra knowledge yields a lot!

    The previous trouble was, that acquiring more knowledge leads to cognitive overload (= you can not think of all rules at the same time, nor would you know when which rule has priority), but with endgame rules I have rather the experience - it cuts down a lot of thinking!
    I think (calculate) much less due to endgame knowledge issues.

    1. "P.S. wanted to add: everything that "worked" was probably the initial improvement that happens when we start training. Instead of Polgar's Brick, you could have solved with "salt mines" first, just to discover that afterwards you would not improve with Polgar's Brick (because it would be a follow up of a similar training like "salt mines")."

      You ar quite right. If the the knights errant have proven anything, then it is that any tactical untrained pupil will gain about 250 rating points, no matter the training method. That is by harvesting the low hanging fruit of knowledge that is transferable between one position and another. But when all this knowledge is assimilated and transformed into skill, we run out of transferable knowledge.

  10. Yes, seems to be quite right looking at my tactical training. Actually, if I look at it a bit more strict and eliminate the first 1000 chesstempo puzzles (getting used to the pieces, and the mode) like Aoxomoxoa suggests, and if I then look at the final average point I reached, I would say it is rather 150 tactics-chesstempo-Blitz-rating points only (and not 250).

    Richard, the owner of chesstempo, said that the ratings in the top range are a bit more compressed compared to fide elo ratings. So maybe my 150 CT-Blitz-rating-points-improvement is worth ~200 fide elo?

    Nevertheless, I improved more like 400 fide elo. But my tactics-ratings cant explain that. Doing some bullet and fast blitz games with Tomasz, we have seen that blitz and bullet is more about tactics (and maybe simple tactics) than longer OTB games. Tomasz won against me like 20:5 in blitz/Bullet games. Cant remember the result, but I was put down and destroyed convincingly.
    Nevertheless, Tomasz has a real rating of just about Fide elo ~1800?

    So my longer OTB thinking is somehow much better. Aox found I have a very high K-factor.
    But the trouble so far has been: what actually makes for a high K-factor?
    After self-obersavion and this and that, the conclusion is: it is heavily driven by knowledge.
    Instead of calculating (pattern recognition, tactical power), I follow general considerations. My tactics is just sufficient to not blunder so much.
    Other 2200 fide elo players are tactically much stronger than me. However, deep complicated tactics (= high rated chesstempo tactics) are statistically not very significiant. O.k., I lose a tactical battle if I would go into one. However, I can avoid too sharp positions most of the time.
    I do struggle against tactic-monsters, who like to play gambits. But the average player isnt recklessly sacrificing pieces, nor do my opponents really know that I am tactically weaker than them. So statistically - I do my average score, and get away with it. That is why I can compete against those master players, without having the tactic-skills.

    I have a good opening repertoire (chosen by their winning-statistics), and I do the most out of issues, that have nothing to do with skill (such as time management in the opening - simply play my first 10 known moves within 1-3 minutes in an OTB game).

    It is like you say, temposchlucker: at some point we can not improve further. At least our chessability wont become better. But - issues like time management, opening statistics, fitness, and using a lot of endgame rules instead of calculations - this boosted my rating another 200 points: about ~400 fide elo points altogether.

    Am I ever becomming a titled player such as at least FM? I dont think so. It is just another 100 fide elo points, but I am max-ed out, so to speak. I made the most of what is there to polish and to improve.