With what shall I fix it
dear Liza, dear Liza?
My little blitz test revealed me that I have about 30,000 holes in my bucket. Give or take a hole or two. Somehow, I must find the plugs that fix more than 1 hole at the time. I.e. I need multi fixing plugs. We are talking about transfer here.
So I'm investigating what the holes in my bucket have in common.
I did a quick search of your blog posts, and only found one other reference to "dear Liza." I also searched for references to GM Jonathan Rowson and/or his books, and only found one. (I have referred to GM Rowson's books in several comments.) Consequently, I'm not sure if you are referring to Chess For Zebras or to the song "There's a hole in my bucket".
ReplyDeleteIn any event, GM Rowson's conclusion (regarding Henry's 'holes' in his 'bucket') seems applicable.
"The reason our minds cannot be 'filled' is not because they have 'holes', but because they are not at all like buckets. . . . We construct our understanding of positions, which means using what we have, however imperfect, to make sense of what we are given. . . . real learning is often a painful process, because you are not just collecting new ideas and stacking them up in some sort of expanding cognitive warehouse. . . . Indeed, I have come to believe that the kind of learning that is most useful for chess improvement is actually 'unlearning'.
Perhaps a shift in paradigm might help. Instead of viewing your improvement 'problem' as 'holes' in your 'bucket', consider that your intensive training has created small 'screens' that obscure/block parts of your view of the big 'screen' (big picture; 'vulture's eye view; etc.). In short, a learned inability to see the 'forest' because there are too many 'trees' blocking the view. If so, adding more trees (or plugging 'holes' in your 'bucket') is unlikely to give a more complete 'forest' view (or a 'bucket' that will actually retain all that 'water').
As usual, that's sheer speculation, based on no actual 'facts' of the matter.
I know that by taking just a short number of seconds to let System 1 tell me the 'story' of a position, it is much more likely to point me in the right direction. When I grab hold of the first tree branch tossed to me, I often grab one that will not hold up my weight, and I fall down the cliff. Maybe that's just me. . .
I haven't read Rowson. So I refer to the song.
ReplyDeleteI think I understand what you mean. I don't see the bucket because of the holes.
In the forest metaphor, I have been chopping away trees for the past 22 years. And since a few months, I'm starting to see glimpses of the forest. The time comes near to stop chopping, and to focus on the big picture.
You might find GM Jonathan Rowson's Chess for Zebras: Thinking Differently about Black and White to be very informative relative to your current investigation. There are many similarities in your ideas and conclusions regarding adult chess improvement.
DeleteI'm experimenting with the omission of step 4, pre-digestion of the food by system 2 (conscious thinking).
ReplyDeleteMeaning, I solve blitz tactics, and afterwards I draw the squares and the arrows directly. Thanks to mfardal, I can do that directly now. Without worrying about any rote learning.
Another analogy (from drawing) that might prove helpful. Perception is keyed from “clues” that are right in front of our eyes.
ReplyDeleteLink: OPTICAL ILLUSIONS RELATED TO PERCEPTION
We "SEE" (perceive) what we expect to see. Look at the first picture in the link. If you are told to look for a young woman (in the first illustration), you WILL "see" a young woman. On the other hand, if you are told to look for an old woman, you WILL "see" an old woman. However, if you are NOT told which to look for, you may (initially) "see" one or the other - but you will NOT "see" both.
The perceptual "clues" which enable System 1 to "see" (and report as meaningful to System 2) one or the other are not mutually exclusive. It is the assigned meaning (based on previous experience, knowledge, etc.) to the components that makes the difference in WHAT is perceived. There are several differentiating "clues" in the drawing - the chin (young woman) or the nose (old woman); the ear (young woman) or the eye (old woman); the necklace (young woman) or the mouth (old woman); the nose (partial view of the young woman) or the wart on the nose (old woman).
It is the ROLE that these features play that determines the ease with which the two pictures can be differentiated. In the process of assigning MEANING and IMPORTANCE to the individual (ambiguous) features, System 1 causes other potential meanings and importance to be obscured/occulted. This makes it very difficult (if not impossible) to "see" the opposite "picture". Once System 1 has perceived ONE coherent "explanation" of what the picture is about, System 1 confidently asserts that it is the ONLY rational viewpoint; all other possibilities are eliminated SILENTLY (without conscious awareness or concurrence). System 2, once given a confident assertion from System 1, is satisfied that this ONE explanation represents "reality." Because System 2 is "lazy," it requires focused mental effort (an ability to “jump outside” of the overall System 1/System 2) to overcome that first impression/explanation.
In a lot of cases, it becomes impossible to "see" the alternative viewpoint. If System 2 can be used to change focus and interpretation of ONE of the significant (but ambiguous) "clues", then System 1 will take that "clue" and MAY reinterpret the remainder of the picture with a different plausible explanation. This presumes that we must keep an open mind during the initial search for meaning (i.e., what in this position is significant) during the "vulture's eye view." This is much easier said than done.
The purpose of coloring the squares and drawing the arrows (and using PoPLoAFun)is to change our interpretation of the perceived "clues" by changing the focus for System 1 to (re-)interpret the overall "picture" until the focused and engaged System 2 "approves" (verifies that the interpretation is the correct one).
A suggestion:
ReplyDeleteCAREFULLY look at the problems you got WRONG. (The ones you got right are actually unimportant for IMPROVEMENT.) Try to figure out WHY one (or more) of the "clues" that popped into your conscious mind seemed to be the most important indicators of the solution - but turned out to be WRONG. As you shift your viewpoint from the WRONG to the RIGHT “clues” and solution, determine if it was merely (HA! HA!) your interpretation of a specific "clue" that caused you to focus on the WRONG solution/variation. IFF you find this to be the case (and I think it highly likely), then that insight becomes a "trigger" to jump outside the system and reinterpret that "clue" as possibly ambiguous in ALL positions - and therefore something that needs careful interpretation whenever you "see" it.
There are 2 possibilities to lose points, you are too slow or you went wrong.
Deleteyou did not see the right weakness?
you did not judge it right?
you did play a wrong weakness?
you did not see the refutation?
you did not see the right method to make use of this weakness?
you did play the wrong method, what was wrong?
you made wrong thinking - reasoning
often the position is simply too complex, then you need to analyze it deeper
PART I:
ReplyDeleteOne of the potential problems with being of "two minds" (System 1 and System 2) is that it is likely to be a schizophrenic "view" of what actually happens inside our heads. Instead of the "two minds" working in opposition to each other (and thereby forcing us to choose one or the other "thinking mode"), it is much more plausible that BOTH "minds" are used during the process of choosing a move, a variation, a strategy.
Dr. Betty Edwards (Drawing on the Right Side of the Brain) provides a more nuanced view of the overall process, cooperatively integrating the “two minds” in the creative process. The essential idea is to “apply them [perceptual SKILLS located in System 1] to general thinking and problem solving.” The following discussion is based largely on Dr. Edward’s insights.
One of the things I got WRONG regarding System 1 and System 2 was the sense that the two System somehow operate in opposition, or at a minimum, in mutual exclusion. When System 1 is operating, System 2 is (or should be) shutdown and vice versa. In actuality, the two Systems work together concurrently. For example, during the "vulture's eye" mode, System 1 should be leading, trying to make sense of the "whole". System 2 should influencing (but not leading) by asking questions, such as "I wonder why. . .?" "I wonder if. . .?" or, "I wonder how come. . .?" However, it is very important to not let System 2 run amok and grab control. During this process phase, we have to allow System 1 to do its visual perception, only allowing occasional interruptions by System 2 with refocusing questions.
Once a "feel" for the overall position is acquired, then System 2 is brought to bear to the extent possible what is already known relative to the question(s), focusing on the puzzling parts, and trying to understand whether the problem has already been solved. In essence, this is a "research" phase, requiring saturation with information and data. System 1 operates silently in the background, viewing the uncovered information relative to the question(s) asked previously, looking for whether, or HOW, the information might fit the question(s).
The next phase is where most of us get "stuck." As the investigation continues and no solution is found, frustration rises. At least some of the question(s) remain unanswered, the gathered bits of information refuse to coalesce, and the key that would reveal the solution stays hidden. Logical links remain elusive, and exasperation, unease and anxiety set in. At this point, tired of hammering away with no result, most of us give up - we either look at the solution, or if playing a game, decide to just "go with the flow" and move whatever seems best at the moment. We have to resist this urge! Instead, we should switch back to System 1 and allow a solution to "incubate" and then to percolate back into our System 2 consciousness.
PART II:
ReplyDeleteTo shift the responsibility back to System 1, give it these instructions: "Please solve this for me! I need the solution in X minutes." Then we have to stop using System 1 altogether! [THIS IS VERY DIFFICULT, IF NOT IMPOSSIBLE, FOR ADULTS TO DO WITHOUT SPECIFIC TRAINING AND SKILL!] System 1 then rotates the whole in visual space, wordlessly envisioning missing parts, extrapolating from known to unknown information, seeking the PATTERN(S) where all the parts will fit together into a coherent whole with the right relationships of parts-to-whole.
In most cases, System 1 will present the solution IF WE WAIT FOR IT. System 1 and System 2 will then "celebrate" the Aha! moment of triumph together. This is a very short-lived phase.
The final phase is the stage that most of us don't do well, if we manage to do it at all. The solution must be verified and placed in final form: constructing a coherent model, using Popper's falsification, and delivering the final solution. This phase requires both Systems to work together cooperatively, with System 2 structuring the solution step-by-step and recording the solution, while guiding System 1's grasp of the whole and its well-fitted parts. This stage is simply a lot of hard work, and not nearly as enjoyable as the earlier phases.
Neural plasticity is generally accepted as being more prevalent in young children than in adults. However, I don’t think that is the key difference between youngsters and adults with regard to learning how to improve at chess. Over time, as we go through the educational system, we acquire an almost exclusive reliance on words and logical thinking (System 2); System 1 runs silently in the background, and we squelch it in favor of System 1’s step-by-step logical thinking.. Youngsters are much more flexible with regards to just accepting that WHAT YOU “SEE” IS ALL THERE IS. Where adults want to KNOW, youngsters just want to DO. Adults want to understand WHY; youngsters just want to play an interesting move or variation or combination. In short, youngsters have not developed the adult reliance on words and logical thinking, and are much more likely (especially during the early formative years) to be able to switch into System 1 mode as needed.
I know that’s not very encouraging when trying to use System 2 to “find a solution” to how to improve. I’m sorry, but I don’t think either the “bucket” or the “holes” exist; the problem lies elsewhere. I base this on my own experience learning and teaching a traditional martial art for many years. No emphasis was placed on KNOWLEDGE; every bit of training was oriented to DOING. After DOING something (a technique, a form) repeatedly under the watchful (and critical) eye of a master, eventually it just became part of you, not something you KNEW, but something you KNEW you could DO – without consciously thinking your way through it step-by-step.
Correction to the above comment:
Delete". . . and we squelch it in favor of System 1’s step-by-step logical thinking."
should have read:
". . . and we squelch it in favor of System 2’s step-by-step logical thinking."
It looks a lot like squelching. But system 1 (subconscious thinking) has no choice other than to go where system 2 (conscious thinking) goes. So the key to change is the disciplining of system 2. The weakness of system 2 is laziness during the stage of feedback.
DeleteWe must be very precise. There is not one problem to solve.
ReplyDeleteA. For sure there are holes in my bucket. Meaning, there are holes in my knowledge. When you have never seen a pattern before, you are not familiar with it. And there are patterns that I know, but which are almost forgotten. The holes need their own approach.
B. when the patterns are familiar though, you are right. Then there are no holes to be filled. Nor a bucket as placeholder for the holes. Then we have another problem at hand. Which needs a different approach than A.
C. system 1 goes where system 2 goes. If system 2 (conscious thinking) chases ghosts, then system 1 (subconscious thinking) will perfectly help him with chasing. System 2 must be disciplined not to chase ghosts for too long.
D. providing context helps. The implanted Robert Coble Chess Module can do that.
E. doubt is casted on the need for repetition. I had my greatest progress when I didn't repeat exercises. MDLM introduced repetition. There is no evidence that it matters.
F. I have problems to annotate during a game. Especially in time trouble. This means system 1 doesn't know the names of the squares well enough. While I probably have annotated a few thousand scores. Repetition doesn't matter. But conscious attention might fix the problem. Maybe I should test that. It would give more insight in how system 1 works.
I am very skeptical of any generalized process or body of knowledge which supposedly will improve everyone's skill. Each of us has our own unique set of problems to identify and overcome. The primary "problem" is that the vast majority of us do not know how to recognize the beam in our own eye while we pontificate about the splinters in other people's eyes.
ReplyDeleteI think there is a great need for repetition - but not a mindless repetition of specific exercises. Memorizing a given set of exercises (regardless of the size of that set) will NOT produce any significant improvement in SKILL. If you are "tested" using that set of exercises, you will most likely perform successfully at a high level. Yet when you attempt to solve a problem never seen before, it is highly likely that you will fail. Your pattern recognition mechanism will be highly tuned to the specific examples, but not the general method.
The repetitive training must be focused on the PROCESS, not on any specific examples; the examples chosen are not particularly important. Using the same process over and over to "SEE" what is there [WYSIATI - What You SEE Is All There Is] requires focused attention. Almost all of us fail at inculcating the method/process into System 1/System 2. The few who do MASTER the method/process are called "Masters" for a reason: they can apply their generalized method/process to ANY position.
PART I:
ReplyDeleteI received a copy of Gary Klein's book Sources of Power: How People Make Decisions, 20th Anniversary Edition on Tuesday; I finished reading (and highlighting) it yesterday. There is some interesting "food for thought" especially when contrasting it with Daniel Kahneman's Thinking Fast and Slow. Klein’s focus is on successful decision making under realistic conditions, whereas Kahneman’s emphasis is on heuristics and biases which lead us astray. Please keep in mind my skepticism regarding general “thought processes.”
The traditional multi-stage rational choice model (Peter Soelberg) is:
(1) Identifies the set of options.
(2) Identifies the ways of evaluating these options.
(3) Weights each evaluation dimension.
(4) Does the rating using comparative evaluation between the options.
(5) Picks the option with the highest score.
Klein: “The standard advice for making better decisions is to identify all the important evaluation criteria, weight the importance of each evaluation criteria, evaluate each option on each criterion, tabulate the results, and select the winner. . . . Again and again, the message is repeated: careful analysis is good, incomplete analysis is bad. And again and again, the message is ignored: trainees listen dutifully, then go out of the classes and act on the first option they think of. The reasons are clear. First, the rigorous, analytical approach cannot be used in most natural settings. Second, the recognitional strategies that take advantage of experience are generally successful, not as a substitute for the analytical methods, but as an improvement on them. The analytical methods are not the ideal; they are the fallback for those without enough experience to know what to do.”
Klein’s approach is a variation of naturalistic decision making under real world constraints. He recommends the recognition-primed decision [RPD] model for this type of decision making. It more closely mirrors the way that experienced people make decisions under various constraints such as limited time. The significant RPD difference is that a decision maker starts with a determination if the situation/problem is typical, and, if so, determines a viable option and explores it deeper (via mental simulation). If this option is found to “work”, the appropriate course of action is implemented. Other options are NOT evaluated, and never relative to each other. The problem is “recognized” [System 1], triggering the most likely option, simulated (using “look ahead”)[System 2], and the option and its subsequent course of action is either rejected (with the next most plausible option selected for investigation, starting the process over) or implemented. If the option is NOT found to work, anomalies are identified (if any), clarified and diagnosed to see if the situation remains typical. If not, a search ensues for more data regarding the problem. Otherwise, if the situation is now considered typical, the appropriate course of action is initiated. Lather, rinse, repeat until an option with an appropriate course of action is found and implemented.
PART II:
ReplyDeleteThe significant difference is that a Kotovian process is NOT followed:
(1) List all candidate moves FIRST, before anything else;
(2) Calculate each variation in turn;
(3) Visualize all possible lines as a "tree of variations"; and
(4) During a GAME, each branch of the tree must be examined once and only once.
Pattern recognition (typicality of the position) must occur first based on “cues” from the position. These cues in turn trigger options (concepts) and corresponding sequences of actions (devices/themes and moves). The options are explored in priority order, without comparing them to each other. The first viable option found to “work” triggers the corresponding course of action. [That kills mister Lasker’s aphorism, “When you find a good move, look for a better one.”]
The process is dependent on internalized background knowledge and experience. It generally takes quite a bit of time to acquire the appropriate kinds and amounts of experience. Once that expertise is developed, the decision maker can often make what seems like an intuitive decision, with little or no awareness of the process whereby he made the decision. This is why chess masters can often play at very fast time controls with very little decrease in playing strength.
Klein claims that this is actually the process followed by chess masters. The only comparable chess process that I can recall is GM Jonathan Tisdall’s [Improve Your Chess NOW!] analytical method called variation processing:
My theory contends that a combination of the natural human approach to the position, tempered with some of the discipline advocated by Kotov, [the Tree of Analysis approach] is more effective. The components of this technique are (in this order):
1) To aim towards the choice of a single critical variation (heresy!). Branches are dealt with when unavoidable, and primarily to navigate the chief variation.
2) The constant application of abstract assessment.
3) A scan for critical candidates.
Step 2 is a verbalization (silently, if actually playing a game) of the "cues" that one discovers while surveying the position.
And, like most processes, it does not tell us how to train ourselves to use it to improve at chess.
PART III:
ReplyDeleteActually, that last statement above is incorrect. There is no training program on how to "do" RPD making. Klein states that the RPD model is a description, not a prescription. As a decision maker gains knowledge and experience, the RPD model will be followed automatically (with some variation in the components of the model). It's just what experienced decision makers DO after gaining expertise.
What I'm pleaing for is to energize the moment of feedback with attention.
ReplyDeletePART I:
ReplyDeleteJust got my November 2020 copy of Chess Life, the monthly magazine for US Chess Federation members. (I'm a Life Member.) There is an excellent article in it by WFM Elizabeth Spiegel. WFM Spiegel is a middle school teacher in Brooklyn, New York, and a "leading scholastic coach". The title of her article is: Improving Your Tactical Sensibilities.
". . ., I have continually found myself returning to one question: what is the most efficient way to teach a group of students to improve their calculative abilities? Or, put differently, exactly what does a strong player think about when she/he tries to solve a tactics problem?"
"The first thing I teach is that it is your job as a chess player to look at every forcing move on every turn. Forcing moves are checks, captures, and major threats like checkmate or threats against the queen. You must calculate every forcing move to the end of the forced variation. [Emphasis added.]”
“I say this over and over to my students. It is your job to calculate every check and every capture, for both sides, on every move. This is admittedly a lot of work. Fortunately, there are more efficient thinking methods than sheer brute force. Over the years, I have found three questions that, I think, help to direct students’ thinking in the most useful way:”
PART II:
ReplyDelete“QUESTION 1: What enemy pieces and pawns are not protected [LPDO] or not well protected [B.A.D.]? ”
. . . [She defines LPDO and B.A.D., although using descriptions but not those terms.]
“QUESTION 2: What enemy pieces are on the same line as each other or as my piece?”
“This will help you find discoveries, pins, and skewers.”
“QUESTION 3: What’s your dream move?”
“This question actually has a few versions. The beginner version is, ‘If you could pick up your queen and put it anywhere, where would it be checkmate?’ And,, for such a highly specific question, it is remarkably helpful!”
“The advanced version of the question is “If you could pick up any piece and put it on another square, where would you put it? This will often give you great maneuvering ideas but will also draw your attention to tactics that might otherwise look impossible. The idea is to focus on what you want to achieve and not be limited by what you think your immediate choices are.”
“Let’s look at how we can ask our questions in a few positions. We will start by CIRCLING THE TARGETS: the unprotected and not-well-protected pieces. Then DRAW LINES wherever you notice pieces on the same line. [In her examples, she draws the line from the piece to the edge of the board; imagine THAT!] Finally, let your imagination run a little and ask yourself what would happen if you pick a piece up and move it anywhere. SEE WHAT IDEAS JUMP INTO YOUR MIND.”
“Next, [after working through her examples] try these yourself. [A series of 12 exercises/test positions follow.] Start by circling the targets, drawing any IMPORTANT lines you notice, and making note of any drop-queen checkmates or piece improvements.”
What she described is essentially PoPLoAFun, without any explicit reference to Functions.
It seems to work for her middle school students.
I wonder: might it work for adults seeking improvement?!?
The PoPLoAFun method needs an extension: the additional punch move. Some moves are gaining an additional tempo. I express that with a red arrow. (PoPLoAFunAp?)
ReplyDelete