Pattern Recognition vs Calculation
I find myself once again reflecting upon the pseudoscience that is Chess Improvement theory. The ability to see tactics where one gains material and to checkmate your opponent and to retain one's material and avoid being checkmated is necessary to play a good game.
The ability to solve a chess problem greatly depends on two things: Pattern Recognition and Calculating Ability. Although you think it is clear cut I find myself wonder about what exactly the difference between Pattern Recognition and Calculation. So I came up with this chart that speaks to what I see as the nature of Pattern Recognition and Calculation. Please note this is rough ideas on the subject and will change as my ideas evolve.
Pattern Recognition
Can be Extremely Quick
Pattern Recognition is certain
Intuitive/subconscious
Tendency to be Visual/Nonverbal
Extemely Important in Blitz Games
Doesn't alway consider counterplay (and can be dangerous to totally rely on)
Improved through practice of repetitive positions (flashcards)
Pattern recognition does not rely on calculation
The more one has seen a position the more likely pattern recognition is used
Calculation
Tends to be slower
Calculation may involve trial and error,what will happen if I play this?
Intellectual Questioning/conscious
Tendency to be Verbal with some visualization
Takes into account Strategy and other factors
Extremely Important in Long Games
Often times looking to prime the pump of pattern recognition
More in practice in complicated positions
Important in considering counterplay
Improved through practice of many different positions
Calculation relies more on pattern recognition than PR relies on calculation
The less one has seen a position the more calculation is used
This chart speaks to what I think tendency of these two elements and is somewhat simplistic and arbitrary. In a game in looking at the same position one uses elements of calculation and pattern recognition. Much of what I have defined as calculation is looking for elements of recognized paterns in a new position and shifting and reshaping the position in ones mind. So perhaps we should think of pattern recognition as layer within Calculation. Pattern Recognition directs one to
consider what to calculate and how deeply.
Both are important in Tactics so I look to strengthen them both
___________________________________________________________
Circle 2 continued
Sat Aft l20- 216
Sun m -249
Sun -269
mon -290
tues 308
The ability to solve a chess problem greatly depends on two things: Pattern Recognition and Calculating Ability. Although you think it is clear cut I find myself wonder about what exactly the difference between Pattern Recognition and Calculation. So I came up with this chart that speaks to what I see as the nature of Pattern Recognition and Calculation. Please note this is rough ideas on the subject and will change as my ideas evolve.
Pattern Recognition
Can be Extremely Quick
Pattern Recognition is certain
Intuitive/subconscious
Tendency to be Visual/Nonverbal
Extemely Important in Blitz Games
Doesn't alway consider counterplay (and can be dangerous to totally rely on)
Improved through practice of repetitive positions (flashcards)
Pattern recognition does not rely on calculation
The more one has seen a position the more likely pattern recognition is used
Calculation
Tends to be slower
Calculation may involve trial and error,what will happen if I play this?
Intellectual Questioning/conscious
Tendency to be Verbal with some visualization
Takes into account Strategy and other factors
Extremely Important in Long Games
Often times looking to prime the pump of pattern recognition
More in practice in complicated positions
Important in considering counterplay
Improved through practice of many different positions
Calculation relies more on pattern recognition than PR relies on calculation
The less one has seen a position the more calculation is used
This chart speaks to what I think tendency of these two elements and is somewhat simplistic and arbitrary. In a game in looking at the same position one uses elements of calculation and pattern recognition. Much of what I have defined as calculation is looking for elements of recognized paterns in a new position and shifting and reshaping the position in ones mind. So perhaps we should think of pattern recognition as layer within Calculation. Pattern Recognition directs one to
consider what to calculate and how deeply.
Both are important in Tactics so I look to strengthen them both
___________________________________________________________
Circle 2 continued
Sat Aft l20- 216
Sun m -249
Sun -269
mon -290
tues 308
posted by takchess @ 9:49 AM
3 Comments:
At 5:53 AM,
Temposchlucker said…
If the chess game was "cracked", study of it would be solely based on pattern recognition. Calculation comes in where pattern recognition has reached it's boundaries.
The problem with calculation is that it makes use of short term memory which is errorprone and VERY limited. Even if I calculate for weeks I cannot come up with the moves that Topalov finds within minutes. Pattern recognition is the only way to break thru this.
At 1:38 PM,
takchess said…
Pattern recognition is vital and one could not calculate properly if the pattern recognition was poor. I will further refine what I mean in this post what I refer to the elements of Pattern Recognition vs Calculation.
For ease of differentiation in this post, I picture Pattern Recognition as viewing a tactical problem and immediately knowing the answer ie what to play. The position is known and it is memorized and a part of you.
In this post I view Calculation outside of its normal definition. I am viewing as the ability to solve a tactical problem that one does not know immediately through P.Recognition. So perhaps a better term might be tactical problem solving ability:the ability to solve over the board.
So when I first see a tactical problem and
walk through possible solutions. I may be doing some pattern recognition (hmmm that queen is on the same file as the king, how can I take advantage of that once I remove the pawn in my way?) That is worthwhile skill to have and as I am building pattern recognition the fact that I am solving the problem I believe helps build pattern recognition. Although I am not sure I can prove this. Tempo and others what do you think?
So in some ways Calculation relies on me making connections with Patterns I have seen in the past from problems, It can be a connection that this feels like a specific position in a classic game.
Pattern Recognition knowing the answer is the ideal for me. So I will pose this question to you chess people. What is the ideal for building Pattern Recognition is is an endless loop of video showing pieces moving solving chess problems ? Or is the ideal given a chess problem and solving it on it's own does that better build Pattern Recognition?
One of my first posts when I became a Knight asked is Calculation muscle necessary ? Can one progress by just studying the answers? Is anyone actively studying by doing that directly or aware of programs that teach that way?
At 1:49 AM,
Temposchlucker said…
You touch the core. I'm searching for the best method for long.
If you don't start with recognizing a pattern in any given position, you start with trial and error. [You hardly can call this calculation.] Until all of a sudden you recognize a pattern. Most of the patterns don't work right away. Then starts a process of logical thinking. "If I do this then he does that." Which is again a form of trial error. You try to look in the future from this position. You have to assess any future possibility. For unknown reasons we call this "calculation".
I don't think very high of men's possibility to "calculate". If you take the mystification from it, trial and error remains. Plus "assessment", which basically is a comparison with past similar situations. The word similar indicates pattern recognition.
I think of calculation as being very finite. You reach your top after a year of 3 intensive training. After that there is nothing to gain there. The efficiency of calculation is totally based on the patterns and chunks that are used.
I'm experimenting with different drills and flashcards-like approaches (trial and error:)
Until now I haven't reached definite conclusions. Yet.
Post a Comment
<< Home