The Axiom System - Part 4: Justification in Chess

@Toscani said in #47:

Principles are not meant to be rigid rules, but rather flexible guidelines. When a person lives by principles, it does not imply inflexibility or lack of adaptability. This concept extends to the principles of chess as well. These principles serve as tools to generate or guide our thoughts in a constructive manner, rather than being strict commandments.

Thank you for your comment. This is very similar to what wodjul wrote in comment #40. The key here is further exploring this statement 'these principles serve as tools to guide our thoughts in a constructive manner'. You can see more details on this issue in my comment #44.

@Toscani said in #47: > Principles are not meant to be rigid rules, but rather flexible guidelines. When a person lives by principles, it does not imply inflexibility or lack of adaptability. This concept extends to the principles of chess as well. These principles serve as tools to generate or guide our thoughts in a constructive manner, rather than being strict commandments. > Thank you for your comment. This is very similar to what wodjul wrote in comment #40. The key here is further exploring this statement 'these principles serve as tools to guide our thoughts in a constructive manner'. You can see more details on this issue in my comment #44.

FM DailyInsanity

#53

@GnocchiPup said in #48:

Thank you for your comment and your thoughts. I appreciate your perspective, but I believe you might be giving my argument less rigor than it was intended. The key to the dichotomy is whether consideration of principles does or does not influence the decision-making process. For example, if you have an initial move preference, after considering principles, your preference either changes or remains the same as a consequence of that consideration.

It seems you don't disagree with this approach but are suggesting that principles can influence the decision-making process through an unconventional method. You seem to imply that the quality of a move is proportional to the number of principles supporting it. While this might hold in a general, academic sense, or within rule-based computer evaluation (which differs significantly from human evaluation), it still lacks the precision needed to explain how it helps us make decisions.

Should we count the number of principles supporting each move and use this to decide which move to play? If we intuitively sense that a certain move is good, but this method suggests a different move, would we follow the method's suggestion? If not, how is this method useful?

As I see it, there are two problems with this intuition approach to chess:

In an actual game, how exactly does one select candidate moves with an intuition only approach?
In teaching, how can concepts be taught to others without resorting to principles?

Here, I don't see how the first point is a problem. Players simply select a candidate move based on intuition (combined, of course, with search, which is described by the 'vision' component of the decision-making process). The lack of an explicit, external reasoning process to dictate this decision is more or less the definition of intuition.

Regarding the second point, I should note that the conclusion of my argument is "Chess principles are not useful when selecting between a set of candidate moves," not that they shouldn't be taught to others. We might indeed question whether they should be taught, but in my view, even if we assume they have zero practical utility, they might still be worth teaching for enjoyment and satisfaction reasons. We should, however, be aware of their relatively minor importance in directing in-game decisions. Additionally, as discussed in other comments here, awareness of chess principles might help with intuition development, even if they don't directly assist in a game scenario.

@GnocchiPup said in #48: > Thank you for your comment and your thoughts. I appreciate your perspective, but I believe you might be giving my argument less rigor than it was intended. The key to the dichotomy is whether consideration of principles does or does not influence the decision-making process. For example, if you have an initial move preference, after considering principles, your preference either changes or remains the same as a consequence of that consideration. It seems you don't disagree with this approach but are suggesting that principles can influence the decision-making process through an unconventional method. You seem to imply that the quality of a move is proportional to the number of principles supporting it. While this might hold in a general, academic sense, or within rule-based computer evaluation (which differs significantly from human evaluation), it still lacks the precision needed to explain how it helps us make decisions. Should we count the number of principles supporting each move and use this to decide which move to play? If we intuitively sense that a certain move is good, but this method suggests a different move, would we follow the method's suggestion? If not, how is this method useful? >As I see it, there are two problems with this intuition approach to chess: 1. In an actual game, how exactly does one select candidate moves with an intuition only approach? 2. In teaching, how can concepts be taught to others without resorting to principles? > Here, I don't see how the first point is a problem. Players simply select a candidate move based on intuition (combined, of course, with search, which is described by the 'vision' component of the decision-making process). The lack of an explicit, external reasoning process to dictate this decision is more or less the definition of intuition. Regarding the second point, I should note that the conclusion of my argument is "Chess principles are not useful when selecting between a set of candidate moves," not that they shouldn't be taught to others. We might indeed question whether they should be taught, but in my view, even if we assume they have zero practical utility, they might still be worth teaching for enjoyment and satisfaction reasons. We should, however, be aware of their relatively minor importance in directing in-game decisions. Additionally, as discussed in other comments here, awareness of chess principles might help with intuition development, even if they don't directly assist in a game scenario.

jdannan

#54

The choice of which move to prefer, which may boil down to which principle we prioritise, cannot be made by mere consideration of these principles. There is no hierarchy or strict ordering of principles (other than perhaps, getting checkmated by force is always the worst choice, and checkmating by force is always the best). Rather it depends on which principle we choose to prioritise in a give position, together with any concrete calculations regarding which candidate moves actually achieve their strategic goals without a tactical refutation.

I agree with the premise that the principles themselves do not determine the evaluation of the position but they do provide useful guidance. When I look at a position, I see backward/isolated pawns, outposts, bad bishops etc. Sometimes these things are over-ridden by an immediate tactic but most of the time they provide useful guidance. In theory chess is a game of pure calculation, "strategy" is a construct that we only use because we can't calculate adequately. But it's necessary for precisely that reason.

I think principles have more of a role to play in the selection of candidate moves than in the choice between them. Applying principles, we can think of developing pieces, controlling squares, files, creating and using outposts, etc etc. Exploring these ideas may lead us to looking at several candidate moves that achieve (or at least support) such goals. The choice of which move to prefer, which may boil down to which principle we prioritise, cannot be made by mere consideration of these principles. There is no hierarchy or strict ordering of principles (other than perhaps, getting checkmated by force is always the worst choice, and checkmating by force is always the best). Rather it depends on which principle we choose to prioritise in a give position, together with any concrete calculations regarding which candidate moves actually achieve their strategic goals without a tactical refutation. I agree with the premise that the principles themselves do not determine the evaluation of the position but they do provide useful guidance. When I look at a position, I see backward/isolated pawns, outposts, bad bishops etc. Sometimes these things are over-ridden by an immediate tactic but most of the time they provide useful guidance. In theory chess is a game of pure calculation, "strategy" is a construct that we only use because we can't calculate adequately. But it's necessary for precisely that reason.

Wodjul edited

#55

My (self-concerned) position is that I just don't know what would work for me. Chess Principles come in lists. An online search will easily find a list of ten principles. Then there are the chess sayings like "a rook belongs on the open file" or "for the rook the seventh rank is heaven" or "a rook belongs behind the passed pawn". Are any of these principles and aphorisms always accurate for every position? I would guess not. Then come the lists for candidate moves (checks, captures, threats) and lists for blunder checks. And lists for tactics to look for. Who can keep all these sayings and lists in their mind at one time or even memorise them and then run through them for every move (other than for memorised opening moves)? Certainly, actually running through memorised lists would only be possible in classical time limit games and even there the routine would become time consuming and onerous.

Based on what @jdannan wrote, and what I know from understanding the necessity for tree searching (calculation) in chess it would seem that calculation of combinations is the ne plus ultra for chess. Without some of that, all principles are certainly useless. There is probably a "theoretical minimum" of calculative ability which needs to be attained before chess principles would help (if they help). The bottom part of combination ability is "automatically seeing stuff" which flows from pattern recognition, motif recognition and the like. The top part of calculative ability is true and exacting calculation of combinations.

Rapid and Blitz chess have changed the chess equation. Very few play even "fast classical" (30+20) online. Rapid and blitz put a premium on "automatically seeing stuff" and having an automatic inbuilt "anti-blunder reflex". Raipd and blitz also put a premium on memorising openings, unusual and gambit openings and opening traps. Except among higher rated players, endings are reduced in importance. More of the low rated games never reach endings and time flagging rises in importance. Being well ahead on the clock can end up being worth a minor piece or thereabouts.

Having said all that, what should the adult improver do? For adult improver read "slow learner". Adults, especially older adults, notoriously have trouble learning and adapting to new stuff. As an aged geezer I can attest to the truth of this. It has to do with the continuous loss of brain plasticity (making new synapse connections).

DailyInsanity's recommendation for tactics improvement (one recommendation) is the repetition learning of same puzzles and like puzzles (not streams of unlike puzzles). I agree in one important sense. Adult improvers especially will need multiple reinforcements to make perception of the right moves automatic or near automatic. My concern is with the number of patterns and motifs which will need to be reinforcement learned like this. Ten reps (say) of each pattern and motif times the potential number of patterns and motifs could multiply up to a huge number and huge time investment. Then there will be revision needs. Approximately how many patterns and motifs will need to be fully learned to automaticity to make a real difference in most games? Would it be (for example) 100 or 1,000 or 10,000?

We need to ask these pragmatic questions about modern chess learning and I think DailyInsanity is asking such questions or strongly implying the need to consider them.

I see a strong concern with teaching in these @DailyInsanity essays and at his site. As a start, how do we effectively teach a beginner or a rookie to play better moves? How does a person effectively learn or even teach themselves to get better? Do "chess principles" really help them? I won't presume to ask if principles help the IM, FM or GM. That is out of my orbit. My (self-concerned) position is that I just don't know what would work for me. Chess Principles come in lists. An online search will easily find a list of ten principles. Then there are the chess sayings like "a rook belongs on the open file" or "for the rook the seventh rank is heaven" or "a rook belongs behind the passed pawn". Are any of these principles and aphorisms always accurate for every position? I would guess not. Then come the lists for candidate moves (checks, captures, threats) and lists for blunder checks. And lists for tactics to look for. Who can keep all these sayings and lists in their mind at one time or even memorise them and then run through them for every move (other than for memorised opening moves)? Certainly, actually running through memorised lists would only be possible in classical time limit games and even there the routine would become time consuming and onerous. Based on what @jdannan wrote, and what I know from understanding the necessity for tree searching (calculation) in chess it would seem that calculation of combinations is the ne plus ultra for chess. Without some of that, all principles are certainly useless. There is probably a "theoretical minimum" of calculative ability which needs to be attained before chess principles would help (if they help). The bottom part of combination ability is "automatically seeing stuff" which flows from pattern recognition, motif recognition and the like. The top part of calculative ability is true and exacting calculation of combinations. Rapid and Blitz chess have changed the chess equation. Very few play even "fast classical" (30+20) online. Rapid and blitz put a premium on "automatically seeing stuff" and having an automatic inbuilt "anti-blunder reflex". Raipd and blitz also put a premium on memorising openings, unusual and gambit openings and opening traps. Except among higher rated players, endings are reduced in importance. More of the low rated games never reach endings and time flagging rises in importance. Being well ahead on the clock can end up being worth a minor piece or thereabouts. Having said all that, what should the adult improver do? For adult improver read "slow learner". Adults, especially older adults, notoriously have trouble learning and adapting to new stuff. As an aged geezer I can attest to the truth of this. It has to do with the continuous loss of brain plasticity (making new synapse connections). DailyInsanity's recommendation for tactics improvement (one recommendation) is the repetition learning of same puzzles and like puzzles (not streams of unlike puzzles). I agree in one important sense. Adult improvers especially will need multiple reinforcements to make perception of the right moves automatic or near automatic. My concern is with the number of patterns and motifs which will need to be reinforcement learned like this. Ten reps (say) of each pattern and motif times the potential number of patterns and motifs could multiply up to a huge number and huge time investment. Then there will be revision needs. Approximately how many patterns and motifs will need to be fully learned to automaticity to make a real difference in most games? Would it be (for example) 100 or 1,000 or 10,000? We need to ask these pragmatic questions about modern chess learning and I think DailyInsanity is asking such questions or strongly implying the need to consider them.

#56

dboing

#57

The 2 things I strongly agree with, and it would not change if I did not agree on everything else once I would have understodd correclty the current version of the author intent (yes, behind all solid shiny construct there is the fallible human still, and its grasping at verbal languages constricted logic communicationj tool to share it with us).

The value of from scratch propositions.
the separation between seeing and evaluating.

This may still be semantically slipped in each readers mind, even possibly to mean something I might not agree with. But there seems to be a world where my meaning is the dominant one in all reades too, and hopefully the author.

Talk to you later, all of you. if reading this. Keep filtering each others intended ideas through the sliperry approxijmating tools that those keyboard and the internets can still let us exchange in somewhat syncrhrniy.

i do not have energy to follow here for a while. I just wanted to say what I find refresshing and agree with. even now in spite of my other concerns, some of which I am going to be working on with jomega, whom I can read behind my prose, thanks for past mutilple discussions, we know our agreements and disagreements about theory of learnings, and that has been our priority from different end point of the problem, I would say very pragamatic approach through debate. but i don't have the energy or intensity per day to do all I want to do.. And thanlks the others for their helping the blog authors test its hunches made theory. The 2 things I strongly agree with, and it would not change if I did not agree on everything else once I would have understodd correclty the current version of the author intent (yes, behind all solid shiny construct there is the fallible human still, and its grasping at verbal languages constricted logic communicationj tool to share it with us). 1) The value of from scratch propositions. 2) the separation between seeing and evaluating. This may still be semantically slipped in each readers mind, even possibly to mean something I might not agree with. But there seems to be a world where my meaning is the dominant one in all reades too, and hopefully the author. Talk to you later, all of you. if reading this. Keep filtering each others intended ideas through the sliperry approxijmating tools that those keyboard and the internets can still let us exchange in somewhat syncrhrniy.

FM DailyInsanity

#58

@Wodjul said in #55:

Those are all very important questions you have raised. Yes, I suppose the overall concern would be with teaching/chess improvement. Many prevailing, general assumptions should be questioned, which might lead to 'general progress' in the chess world. However, I'm most interested in how this progress relates to chess improvement strategies and teaching, as I think that's what most people are interested in.

Approximately how many patterns and motifs will need to be fully learned to automaticity to make a real difference in most games? Would it be (for example) 100 or 1,000 or 10,000?

I recently read that some research has argued that chess masters have acquired ~10,000 'chunks' as the cognitive scientists call it. Not sure if this would be exactly the same as '10,000 motifs', but is interesting to hear nonetheless.

@Wodjul said in #55: > Those are all very important questions you have raised. Yes, I suppose the overall concern would be with teaching/chess improvement. Many prevailing, general assumptions should be questioned, which might lead to 'general progress' in the chess world. However, I'm most interested in how this progress relates to chess improvement strategies and teaching, as I think that's what most people are interested in. >Approximately how many patterns and motifs will need to be fully learned to automaticity to make a real difference in most games? Would it be (for example) 100 or 1,000 or 10,000? I recently read that some research has argued that chess masters have acquired ~10,000 'chunks' as the cognitive scientists call it. Not sure if this would be exactly the same as '10,000 motifs', but is interesting to hear nonetheless.

Wodjul edited

#59

@DailyInsanity said in #58:

Those are all very important questions you have raised. Yes, I suppose the overall concern would be with teaching/chess improvement. Many prevailing, general assumptions should be questioned, which might lead to 'general progress' in the chess world. However, I'm most interested in how this progress relates to chess improvement strategies and teaching, as I think that's what most people are interested in.

I recently read that some research has argued that chess masters have acquired ~10,000 'chunks' as the cognitive scientists call it. Not sure if this would be exactly the same as '10,000 motifs', but is interesting to hear nonetheless.

The average adult English speaking person has about a 20,000 active word vocabulary according to wordcounter.io. We can assume these exist as 20,000 chunks for comprehension and composition. Then there are the sound chunks. This is for words alone. People know a lot of other stuff too, obviously. They can learn a second language in adulthood. I am aware that learned, brain active people can learn a second language after age 60. (Not me, I haven't tried that.)

So brain capacity and brain activity alone are not the problem for adults without cognitive decline. It will be depend on the years of life left to live and learn in and then the time that can be allocated per day at chess learning and the speed of learning with proper retention. It will also depend on the individual's motivation and energy. Stamina decline, physical and mental, become an issue as we age. Dealing with frustration at a slow learning rate (compared to a healthy, intelligent child), life responsibilities and motivational justification (is chess worth all this effort?) will be what the adult improver will contend with too. For these purposes, I would call the adult improver someone who is only getting serious about chess after say the age of 30. Under 30s (young adults) are still in a good position I think unless they have a demanding (non-chess) career or coursework and/or marriage and kids.

If you want to be a GM start before age 10 I would say and make sure you are born with a darn good IQ, at least for spatial, temporal and symbolic thinking / visualization. That's just my non-expert opinion of course.

"In cognitive psychology, chunking is a process by which small individual pieces of a set of information are bound together to create a meaningful whole later on in memory. The chunks, by which the information is grouped, are meant to improve short-term retention of the material, thus bypassing the limited capacity of working memory and allowing the working memory to be more efficient." Wikipedia.

@DailyInsanity said in #58: > Those are all very important questions you have raised. Yes, I suppose the overall concern would be with teaching/chess improvement. Many prevailing, general assumptions should be questioned, which might lead to 'general progress' in the chess world. However, I'm most interested in how this progress relates to chess improvement strategies and teaching, as I think that's what most people are interested in. > > > > I recently read that some research has argued that chess masters have acquired ~10,000 'chunks' as the cognitive scientists call it. Not sure if this would be exactly the same as '10,000 motifs', but is interesting to hear nonetheless. The average adult English speaking person has about a 20,000 active word vocabulary according to wordcounter.io. We can assume these exist as 20,000 chunks for comprehension and composition. Then there are the sound chunks. This is for words alone. People know a lot of other stuff too, obviously. They can learn a second language in adulthood. I am aware that learned, brain active people can learn a second language after age 60. (Not me, I haven't tried that.) So brain capacity and brain activity alone are not the problem for adults without cognitive decline. It will be depend on the years of life left to live and learn in and then the time that can be allocated per day at chess learning and the speed of learning with proper retention. It will also depend on the individual's motivation and energy. Stamina decline, physical and mental, become an issue as we age. Dealing with frustration at a slow learning rate (compared to a healthy, intelligent child), life responsibilities and motivational justification (is chess worth all this effort?) will be what the adult improver will contend with too. For these purposes, I would call the adult improver someone who is only getting serious about chess after say the age of 30. Under 30s (young adults) are still in a good position I think unless they have a demanding (non-chess) career or coursework and/or marriage and kids. If you want to be a GM start before age 10 I would say and make sure you are born with a darn good IQ, at least for spatial, temporal and symbolic thinking / visualization. That's just my non-expert opinion of course. "In cognitive psychology, chunking is a process by which small individual pieces of a set of information are bound together to create a meaningful whole later on in memory. The chunks, by which the information is grouped, are meant to improve short-term retention of the material, thus bypassing the limited capacity of working memory and allowing the working memory to be more efficient." Wikipedia.

#60