Friday, June 09, 2017

Reshuffling the tree of scenarios

 I reshuffled the tree of scenario's
It is more logical now. Time to put it to the test.


25 comments:

  1. You will find that the PoPLoAFun system has been fully implemented here. Both in an explicit and in an implicit manner. The tree of scenarios implements very simple and obvious chess logic.

    ReplyDelete
  2. Good job of laying it out so clearly! I was a little worried that the ToS might be a little unwieldy, but it's not.

    ReplyDelete
  3. It needs fine tuning and more feedback from practice, of course.

    I didn't design it as a thought process, but as a framework to hang knowledge in, in order to make it retrievable. It is geared around the status of the target, point of pressure and attacker. Once the status is recognized, an action can follow, which is the actual scenario. I should have given the statuses a different colour than the action.

    The idea of statuses is that those can be seen in an instant with the vultures view. Which is something different than a sequential thought process.

    It works bidirectional. It can be used to help to recognize the state the position is in, and you can use it to categorize positions.

    ReplyDelete
  4. i think you should add an other category , the ones of error you do
    you can think of a puzzles as
    - a sequence of strongest moves and variations
    - as a status ( i call it tactical weakness(es) ) and action ( i call it (standard)method(s) )

    and additionally!

    -error you made ( not considering move/method xyz intense enough...)

    all this will make us spend more attention to the puzzle, guide it to the important aspects and will hopefully engrave it into our brains

    Still it will be necessary to make sure!! that we improve at least at this puzzle. If we do the same one again after months and we have the same performance again because we did forget everything.. shame on us

    ReplyDelete
  5. PART I:

    I think your distinction between a (sequential) thought process and the (parallel) vulture's eye view of statuses is very significant in terms of grasping the significant features in a position by intuitively drawing the attention to the available actions. A generalized thought process cannot get at the specifics in sufficient depth to apply to a specific position; a specific thought process cannot be applied except within a very narrow range of positions. Rather than search (perhaps fruitlessly?) for a "one size fits all" thought process, there is a better way. I think that "better way" is being developed here.

    I note that we have "seen" references to a "status examination" of every piece on the board from mr. Weteschnik and mr. Beim. I think your approach using PoPLoAFun is similar in that examination part of the process but is much better as a "coat rack" for anchoring our knowledge of specific areas (strategical, tactical, positional, combinational, etc.).

    It is difficult to remember all of the complexities of chess without memorizing a significant body of knowledge. (H/T to Aox!) I recalled (without using that specific technique) the Roman Road approach to remembering complex subjects. So, I did a Google search (using somebody else's memory and recall mechanism) and found the following very important articles (with regard to the current topic).

    (1) Art of memory

    (2) Method of loci

    (3) Creative Remembering techniques: Learn How To Remember

    I suggest exploration of those memory concepts with a vulture's eye view to "seeing" the connection between those ancient techniques and the current "coat rack"/”scenario” approach to remembering our chess knowledge AND HOW TO APPLY IT. Reference 3, Theory, summarizes remembering techniques this way:

    "Whatever you want to remember,

    you HAVE to change it into a picture

    and place it onto a journey.
    "

    ReplyDelete
  6. PART II:

    The "pictures" we need to remember are the pattern recognition "cues" that trigger the appropriate SPECIFIC action in a given situation. Are there 5,000? 10,000? 50,000?, 100,000? 300,000? “patterns” to be remembered? (I have seen all of these figures given for the number of "patterns" that a master [or higher] player has stored in LTM. My personal opinion is that the NUMBER of required "patterns" to be stored is considerably smaller than any of those values. Consider the PoPLoAFun approach as an example of how a very limited set of "patterns" can provide considerable insight ["cues"] into the relevant actions to be taken, PROVIDED we "remember" to apply PoPLoAFun thoroughly and consistently AND SUBCONSCIOUSLY. I digress.]

    I know I have resisted memorization in the past because I connected it with mindless “regurgitation on command” with no connection to USING the knowledge as SKILL. However, on reflection, that is EXACTLY what we want our pattern recognition “cues” to do – regurgitate on command when we say, “I need to remember the specific things I should do right now!”! By hanging our knowledge on the “coat rack” in specific “scenarios,” we enable our LTM to recognize the relevant patterns AND appropriate actions to be taken WITHOUT HAVING TO DO ANYTHING OTHER THAN BECOME “AWARE” THAT WE HAVE BEEN “TRIGGERED”! The “journey” (individual steps in the action) becomes easy to take because we have already cemented that “map” through memorization techniques which lean heavily on visualization. The “magic” occurs without having to laboriously work our way through a step-by-step sequential process. We fly like a vulture along the flight path, arriving at the destination as if we had used a transporter beam.

    “Beam me up, Scotty!” (There’s no sign of intelligent life down here.)

    ReplyDelete
  7. PART III:

    How could I "forget" Dr. Lasker?!?

    "The student should endeavor to acquire the habit of designating the squares and of VISUALIZING THEIR POSITION. There are many Chess-players who fail merely from their incapacity to master this geometrical task, not suspecting its value."

    "The motifs of a combination, in themselves simple, are often interwoven with each other. What is it that unites the multiplicity of motifs? We call it the "idea." Motifs, as, for instance, a simultaneous attack against several pieces or the encircling of the hostile King, are tricks of the trade, technicalities. The idea which links the motifs is artistic, it creates something that had never before been there. MOTIFS CAN BE TAUGHT, ideas must be discovered by original effort. Ideas come from nowhere [INTUITION], they are sudden inspirations; THE PLACE OF MOTIFS IS DEFINITE: THE MEMORY."

    Memory, oh memory, wherefore art thou when I need thee?!?

    ReplyDelete
  8. PART IV:

    Dr. Lasker (once more and then I'll stop):

    "Education is chess has to be an education in independent thinking and judging. CHESS MUST NOT BE MEMORIZED, simply because it is not important enough. IF YOU LOAD YOUR MEMORY, YOU SHOULD KNOW W-H-Y. Memory is too valuable to be stocked with trifles. Of my fifty-seven years I have applied at least thirty to forgetting most of what I had learned or read, and since I succeeded in this I have acquired a certain ease and cheer which I should never again like to be without. If need be, I CAN INCREASE MY S-K-I-L-L IN CHESS [considering his age and level of skill at the time of his writing this, that is a very encouraging statement for adult chess improvers!], if need be I can do that of which I have no idea at present. I HAVE STORED LITTLE IN MY MEMORY [whatever happened to those 300,000 memorized patterns?!?], but I CAN APPLY THAT LITTLE, and it is of good use in many and varied emergencies. I KEEP IT IN GOOD ORDER [spaced repetition, perhaps, since he played so infrequently in tournaments?], but resist every attempt to increase its dead weight. [QUALITY, not QUANTITY.]"

    "YOU SHOULD KEEP IN MIND no names, nor numbers, nor isolated incidents, not even results, but ONLY METHODS. THE METHOD IS PLASTIC. IT IS APPLICABLE IN EVERY SITUATION. The result, the isolated incident, is rigid, because bound to wholly individual conditions. THE METHOD PRODUCES NUMEROUS RESULTS; a few of these will remain in our memory, and as long as they remain a few, they are useful to illustrate and to keep alive the RULES which order a thousand [or 300,000] results. Such useful results must be renewed from time to time [definitely sounds like spaced repetition!] just as fresh food has to be supplied to a living organism to keep it strong and healthy. But results useful in this manner have a living connection with rules, and these again are discovered by applying a live method: the whole of this organism must have life, more than that—a harmonious life."

    The method under investigation has already been demonstrated to be (relatively) simple, and also applicable to a wide variety (all? most?) positions. The next stage is to bind the method to a small number of illustrative examples for memorization. I do not think even 1,000 "patterns" will be needed to considerably raise the skill level, perhaps even to the master level.

    ReplyDelete
  9. I have gone at great length to investigate at which things I spill my time when solving problems. I never found the memory to cause problems. Most things, patterns or chunks if you like, are already there, waiting to be used. The problem is for 95% that I focus my attention at the wrong side of the board.

    The PLF (PoPLoAFun) system helps to distinct important squares from unimportant squares, and between relevant pieces and irrelevant pieces. That leads to a dramatic pruning of focal points of attention. The gain in speed comes from exactly that.

    The discussion about the amount of patterns, or the amount of chunks, has always been confused by not making the distinction between the amount of patterns, and the amount of lookalikes of those patterns.

    A knight fork is just one pattern. Is has 26 forms to manifest itself which are merely mirrors of each other, and there are somewhere between the 1000 and 1600 lookalikes. I didn't count them exactly. So how many patterns do you need to memorize in order to recognize all possible knight forks? Certainly not all 1000-1600 patterns! Probably you need only 6 patterns or so. The rest is recognized by mirroring and translation.

    That is the miraculous magical power of our subconscience, it can work miracles when it comes to find the lookalikes of a pattern.

    ReplyDelete
  10. The PLF system is a set of pattern, you will need to memorise them. Then you will need to memorise many examples to handle this system fast and save and solve with it puzzles quicker.

    Or did you already improve in tactics?

    Your PLF system is a set of chunks.

    Not all chunks are made consciously, they are usually automatically created by the process of learning. In many positions you will make a move without being able to tell why that is a good move.

    ReplyDelete
  11. PART I:

    Presuming one is using the method of loci as the memorization technique,there are some preliminary aspects that must be completed before one can begin memorizing the desired subject matter.

    (1) An overall scaffold or framework must be erected in the memory upon which the items to be remembered will be placed. Traditionally, these have been either "rooms" or significant points of interest along the route of a "journey." One must have a rigid, easily retained order, with a definite beginning. Into this order one places the components of what one wishes to memorize and recall. In this context, that is the "tree of scenarios."

    (2) A vivid visually meaningful image to be associated with each location in the framework. The more vivid and distinct the image, the easier it is to remember it and its association. Additional items can be associated with a given image, but there is a rapidly diminishing recall capability, inversely related to the number of items attached to that specific image. I suggest just “solving” problems until you encounter a problem that you cannot solve. If you finally work through to a correct solution, then that is a prime candidate for memorizing. The difficulty and subsequent emotional experience will help imprint that example most clearly in your mind.

    As Temposchlucker noted above (regarding the "system"), the mind will automatically make a bi-directional association between any two (or more) things whether you intend to make the connection or not. Once a framework has been devised, any "memories" placed within the framework can be "triggered" whenever the relevant location within the framework is accessed. Corresponding actions (moves and variations of moves) can simultaneously be associated with the memorized piece of knowledge. When the visual "cues" from the specific configuration of pieces/squares are triggered, the corresponding actions are also retrieved for "free." MAGIC!

    The mind will automatically take care of "lookalike" (similar) patterns. There is no need to exhaustively catalog every possible position and try to memorize them. This is where capturing (in your own words) a general "rule" that summarizes the knowledge comes into play. It is NOT sufficient to just memorize an image by itself in isolation. If a "general rule" can be devised, it aids recall of several pieces of information which may be related to that rule.

    The mind will automatically make connections across concept boundaries. This is why it is not necessary to try to remember similar concepts discretely. When one concept is "triggered," in very short order the mind will often bring to consciousness and attention the other, similar concept(s). This is why it is so important to utilize as many modes of learning as possible to be able to readily recall the information when needed. Don’t just try to “see” with the vulture’s eye; try to “smell” with the vulture’s nose!

    ReplyDelete
  12. PART II:

    As Aox noted above, not all chunks are made consciously. I think this is especially true as the chunks increase in size and complexity. One starts with a "simple" alphabet, and then proceeds to "discover" and learn how to use words of ever-increasing length and complexity. (Think of the “definition” of the word as the corresponding action(s).) At some point, one stops "thinking" about the alphabet altogether. It is at that point that one can truly be said to have "mastered" the alphabet. At each higher "spiral" of learning, the same process occurs. The lower level(s) which have been assimilated become part of the subconscious UNLESS one consciously decides to "see" and think at that lower level.

    It is similar in nature to the layers of software that exist in any significant software package. Each layer is dependent on (but separate from) the layer(s) underneath the layer that we are interested in. In both cases, the underlying layer(s) (scaffolding; framework) are REQUIRED to be in place before one can work effectively at the higher level. In short, if there isn't an operating system which handles the low-level components (CPU, RAM, I/O) and provides another language layer above that for programming solutions, you are not going to support much higher levels of abstraction which are closer to the problem domain that you want to "solve." I've programmed at the machine code level (writing machine diagnostics for machines that provided no programming language support at all – and it is PAINFUL!) and applications up through the various language levels from assmebler to very complex systems based on non-procedural languages (like SQL). You DO NOT want to "program" your chess pattern recognition memories at the level of "Contacts" (A=Attack; B=Block; C=Constrain; D=Defend). If you do approach your training from that level, you will NOT improve your playing skill much, if at all.

    A cautionary note before jumping into developing a “universal” set of patterns to be memorized using the framework.

    Association was considered to be of critical importance for the practice of the art of memory. However, it was clearly recognized that associations in memory are idiosyncratic, hence, WHAT WORKS FOR ONE PERSON WILL NOT AUTOMATICALLY WORK FOR ALL. For this reason, the associative values given for images in memory texts are usually intended as examples and are not intended to be "universally normative." What this means for us as chess players is that example problems, positions, motifs, tactical themes/devices, combinations, and games that are helpful for recall will vary from person to person. In short, each person will have to find examples that "speak to him" rather than a universal set of examples for everyone.

    YOU MUST DO THE HARD WORK YOURSELF.

    ReplyDelete
  13. Here a great site about chunking,chess,expertise,cognitive science

    The different theories of expertise : http://snitkof.com/cg156/fourtheories.php .
    About the method of loci : http://snitkof.com/cg156/mnemonics.php

    The PLF system is in a way a ( translation of the ) "method of loci" , you make a connection to a "well known system". Well PLF is not exactly a image but the idea of the method of loci is : to make use of the method how the ltm store information : multidimensional associative and to make a connection of an information you want to store to a thing you know well to be able to find the new information more easy.

    I am often thinking about how to translate memory training like the n-back into chess.

    ReplyDelete
    Replies
    1. @ Aox:

      Thanks for the references!

      In the linked reference (1) [given above] to the "Art of memory," there is a discussion of architectural mnemonics, graphical mnemonics and textual mnemonics.

      IMHO, it is vital to eliminate (or drastically reduce to a minimum) the reliance upon word descriptions as the keys to memory. The ancient art of memory incorporated pictures for those who were uneducated. We obviously have a study domain that lends itself easily to "pictures" (a chess position on a board). The computer gives us the capability to make "selections" and modifications to that "picture" without ever using text.

      Since PLF is crucially dependent on the Contacts Method (H/T to Radovic and Averbach) at the lowest level, perhaps an "n-back" approach could be devised somewhat along the following lines. Please note this is off the top of my head, which means I am probably thinking "out of my mind." [Nothing new there!]

      Given a chess game or problem, perform the following sequence.

      (1) Set a value for "n" (the number of steps to "go back").

      (2) Play through the moves until one reaches the END of a local tactical/strategical sequence.

      (3) Immediately reconstruct the position which occurred "n" moves previously by changing ONLY the pieces that have changed since those "n" moves have been made. "Click and drag/drop" are an excellent tool for doing this exercise.

      (4) Ask your self: what did I miss? Pay very close attention to WHY you may have NOT been successful. Ask yourself: what did I miss?

      (5) Lather, rinse, repeat.

      (6) Pick another game/problem and repeat from step (1).

      There are any number of variations on this theme that can be tried.

      The point is to try to RECOGNIZE what is going on BEFORE it is too late to do anything about it. This is about pattern recognition ("cues"). The connection is made from the start to the end. The in-between moves come along "for free" by association, even though they are NOT included in the focus. This correlates well with the memory being able to recall easier the start and finish of a list rather than what is in the middle.

      Delete
    2. Say you have a trainingsset of chess positions and you solve them with a rating of say 1800 points and you analyse the right solutions.. 6 months later you tackle the same set again, what rating do you want to have now ?

      1800 again? more? less?

      If you have the same rating again.. you did forgett the right solutions, the training had no effect

      It is simply necessary to learn to become better. And when you forgett what you have learned.. the learning was useless. To gain and keep knowledge repetition of some type is necessary. The question is "only" what to learn and how to learn.

      Delete
    3. Say, you have a set of bad methods or patterns or whatever in memory that are related to tit for tat types of problems. To become better, you must:

      1. analyze tit for that positions
      2. find out why you fail them
      3. discover the good methods or patterns or whatever
      4. store in some way the new methods etc. in memory
      5. forget the old methods etc. and erase them from memory

      When you test it after half a year and you use still the bad methods, something has gone wrong with either 4 or 5.

      It reminds me of the methods of NLP that are used with phobia. Bad memories are ousted with good memories.

      Delete
  14. WOW! Fantastic and inspiring discussion...! Again!!! :) I love it!

    I want to read it more! :). Your last two articles (posts) are simply outstanding! You bring a new light into our "eternal chess problems" ;). This way you help me (and probably others, too!) to understand how [chess] mind works! Thanks a lot!

    I recenly did some tests. I created a dozen hundred positions into electronic format (PGN) from the books (tests and workbooks). What I discovered? I realized some of these positions I could solve within a few seconds or if they were a bit more challenging (difficult) within a minute. And after I repeated that few hundred times I noticed something strange. It turned out I KNEW most of these position not in an original setup, but with their "later" phase. In short - if there is a position #4 in the diagram I found out I have solved the same position, but a bit changed - I mean after two moves have been played (i.e. I knew the position of "mate in 2").

    This way I was really confused as at first we could say - you did not know these positions as you had not seen these before. But on the other hand - I KNEW these positions due to the "reflections" from the "later phase" or these positions. It is very similar to the stepping stone method presented by GM Jonatan Tisdall at his book "Improve your chess now!". And I think it may be a confirmation related to the patterns and chunks we are talking about now.

    If you create one important pattern and you know its internal (core) idea... there is no problem if anything will be changed unless the integrity is preserved. What do I mean? Let's take into consideration a well known chess motif of "smothered mate". No matter how many pieces will be added (or taken out) to the position you can quickly recognize the motif. The only problem may occur whenever the board (position) is rotated - it can create some problems (confusion) even if you know the position is a very known to your chess mind. The other problem - much serious - is related to the "mess to the original idea". It means if you have the original idea covered by the pretty unusual moves and you cannot decide what move sequence (variation) leads to win (or to be more precise: preserve the win, but shows the final motif in a clean/obvious way).


    What I like the most from Robert's last two posts (articles)? Here you have:

    1) "As Temposchlucker noted above (regarding the "system"), the mind will automatically make a bi-directional association between any two (or more) things whether you intend to make the connection or not. Once a framework has been devised, any "memories" placed within the framework can be "triggered" whenever the relevant location within the framework is accessed. Corresponding actions (moves and variations of moves) can simultaneously be associated with the memorized piece of knowledge. When the visual "cues" from the specific configuration of pieces/squares are triggered, the corresponding actions are also retrieved for "free." MAGIC!".

    2) "At some point, one stops "thinking" about the alphabet altogether. It is at that point that one can truly be said to have "mastered" the alphabet. At each higher "spiral" of learning, the same process occurs. The lower level(s) which have been assimilated become part of the subconscious UNLESS one consciously decides to "see" and think at that lower level."

    Simply OUSTANDING to me! I have seen (or even created) "magic" a lot of times in my games and I could not find any reasonable explanation for that phenomena. Now you helped me a lot my friend! Thanks a lot and keep doing great things (your posts/articles have been really very inspiring to me!).

    ReplyDelete
    Replies
    1. @ Tomasz:

      Thank you but I cannot take credit here; too many other valuable sources have provided the basic information which I have posted. Alas, I am merely a scribe, copying and pasting stuff together to (hopefully) make some sense.

      I think I have had a somewhat similar experience to yours. I have worked through several books of tactical problems. At times I have experienced a feeling of "déjà vu" - I KNOW I have seen an almost identical problem somewhere previously within the same book. So, I go "looking" and usually find one (or even more) positions that obviously are related, perhaps at an earlier or later stage of a problem or game.

      Here is an example from Lasker's Manual of Chess. At first sight, there is absolutely NOTHING in common between these two positions - except the possibility of a geometrical motif. The first position is (relatively) simple to solve. The second will take considerably more effort.

      Diagram 8, pg. 116 [5q2/8/8/8/bk6/1n6/1Q6/1K2N3 w - - 0 1]

      Diagram 12, pg. 119 [8/4q3/2p5/2n2Q2/p5N1/7k/8/1K6 w - - 0 1]

      Having that "Aha!" feeling is what intuition (pattern recognition) is all about.

      Delete
    2. Tomasz what software do you use for your training of your pgn trainingsfile?

      Delete
    3. Actually it is quite original way. I use Chess Intelligent Scanner to create PGN (FEN) files... and within the process I try to solve these (as the program sometimes processes one position up to 50-60-80 seconds).

      Most often when I want to see how much I can solve I use PDF with printed diagrams or PGN file with replaying the positions using ChessPad or ScidvsPC. And I replay the positions with short command (like CTRL+arrow down) or mouse click (to move to another position).

      I would not call my training this way. It is rather experimenting (or testing0 with familiar positions. Training is not a good word to describe what I often do.

      Delete
  15. Once again great discussion . I think Poploafun may be sufficient as tne best memory aid one can use. I at one point studied memory systems and wasnt a big fan of the loci system. I did use the peg system from The Memory Book by Harry Lorayne and Jerry Lucas which is a very readable book. I havent found a good way to apply it to chess improvement. Recently I saw a chess problem which had multiple loa to checkmate but couldnt quite succeed. The answer had to do with a deflection of a queen that was in a very cramped positon [weakness]. Does this weakness get discovered in the POA portion before Loa are realized ? Is sitting duck identification done before you are aware it can be attacked- exploited in your system ? Will send you a link to problem for clarity.

    ReplyDelete
  16. Felner Bancroft 1960 wtm from
    Alburt chess training pocket book #165. https://youtu.be/ZiEVSSMRfhU.

    ReplyDelete
  17. TOMASZ and others. I posted some 5 x5 ctart problems on takchess1 instagram account . Please request to be connected on instagram.

    ReplyDelete
  18. http://chessok.com/images/peshka_screens/CT-ART6-1.jpg Here is an image of a CT-art problem and it's associated 5x5 square . I believe there are about 1700 individual 5x5 squares which I see as as the alphabet of chess tactics. The chess tactic squeezed down to it's most basic form.

    ReplyDelete