Pattern Recognition of Key Positions

Thanks I've pasted them in though. Maybe if I get a longer list though I might ask next time if I can't sort it, if that's OK?

Also if you want to create a tactics training by repetition the files have to include a fen string as well as a pgn.

I set up the 700 positions from "Manual of Chess Combinations 1a" into a SCID database. It has a "tactics trainer" like mode where it assumes each game in the database is a puzzle and feeds you one after another.

I'm using this for my de la Maza-esque tactics training but after reading this thread I will likely also do it for Lev Alburt's 300 Most Important Positions Pocketbook once I've went through all the book's problems once (I'm only a few pages into it right now).  

CT-ART is something that keeps cropping up, I'll pencil that one it and the SCID sounds useful too.

Yes the book is very good. Of course there are debates whether these are the "300 most important positions and ideas" as that's clearly a strong statement, they are very good though. I find the la Maza stuff interesting and such a thing clearly worked for him, proving that very strong tactical play will get you far.

There's also a debate it seems that whether going over the same tactics, key tactics, several times to remember them, is of any benefit than just doing lots of random tactics. I think it will be useful though, but I'll anyway I can only try it and see how I feel.

At first I didn't just intend to do with the positions in this book, but seeing as I have it and they are good ones I might as well. I think you can do it with just about any good tactics though.

I think it is easier to do it with less positions at first though. I'm doing the first 100 several times, then moving on the second 100 etc, and then periodically going back to the first set. I'm sure that I can easily memorise these 300 position by the end of the month, maybe sooner. Unlike say some tactics databases with 5000+ tactics in them. Of course I'm not just doing tactics, I'm doing other stuff, like playing through a lot of master games etc.

I cannot even remember the key positions on my keyboard!!

Ha, ha, I can, I got myself a teach yourself touch typing CD rom, works wonders, now I can waffle on really easily.

Hi Benedictine. I agree with you completely it's just like you wrote my thoughts. I was also wondering where there would be systematical approach to learning key patterns. I found a book "Understanding chess tactics by Martin Weteshnik" It breaks it down into categories..If you know any more sources of similar types of book.. Let me know

Cheers.

hicetnunc

Thanks for the informative blog.  And Benedictine thanks for the good post.  I think there is no question certain concepts come into play.  If you take the h pawn with your bishop will it get trapped?  This is just a simple idea that most chess players know about.  (Nevermind Fishcher appearantly flubbing this in his Match with Spassky) 

I tend to focus on the endgames with few pieces left because its often possible to simplify games to these situations.  The chess mentor courses are great for this. 

With time I would like to focus on the pawn structures of the openings I play.  Especially if they are more closed positions.  I read some good books on the benko gambit that I think improved my play quite a bit.  Now if I can only get someone to accept the gambit pawn! 

But in any event there are general ideas that should be learned.  Here is a good kindle book on this.  The author wrote a follow up that I bought but have not yet read. 

http://www.amazon.com/Fundamental-Patterns-English-Version-ebook/dp/B00AUWJEKA/ref=sr_1_7?s=digital-text&ie=UTF8&qid=1379354686&sr=1-7&keywords=chess

One thing I would like to mention is that the OP says that many of the positions in tactics books are just mate in three's...

The problem with this thinking is that there seems to be an assumption that mate in three's are standalone problems.

In fact I used to think that way myself.  I would be looking at getting a book like John Nunn's 1001 Deadly Checkmates and I would think "No, I better stick with something like Chess Tactics for Tournament Players by Alburt & Palatnik since I need real tactics and not just mates.

Then, when I finally started going through the checkmate tactics I learned that they are based on other tactical ideas as well. For example, a mate in two might involve a discovered check and a pin, etc.

The fact of the matter is that there really isn't a shortcut.  Solve tactics, play over games, analyze your own games deeply.  All of these will take time, but all of them will be helping you build up your store of patterns.

     I note that Amazon is currently offering pre-orders for Improve Your Chess Pattern Recognition: Typical Tools in Key Positions by IM Arthur van de Outdeweetering (a contributor to chess.com's magazine The Master's Bulletin). In the late 1930s, Dutch psychologist Adriaan de Groot recorded the thinking processes of many prominent players of that time: Alekhine, Euwe, Fine, Keres, etc. He also did this for Masters, Experts (candidate masters), and some class players. His objective was to compare their respective approaches to analyzing positions in order to glean insight into the differences that might account for advanced chess perception versus rudimentary sight of board.

     Dr. de Groot did not use composed problems, but instead chose complex studies (positions from actual games) that offered many possibilities. The goal in such cases is not to find a checkmate, but to find the best possible move - or even the second- or third-best possible move. His recorded "protocols" (de Groot's term) would then provide clues about the various levels of positional understanding of differently rated players. He made some interesting peripheral observations in the course of his experiments. He found that grandmasters do not think at greater depth than club players and that they take about the same amount of time to consider a position as players of lower strength. This indicates that their higher level of understanding must be related to something other than "move horizon" or investment of time. The logical conclusion is that pattern recognition must be a significant factor.

     Further corroborating this suspicion was the fact that de Groot's GMs fared no better at understanding than their lesser counterparts when presented with illogical board positions that make no sense with respect to established opening sequences and typical middlegame positional expectations. In other words, when the positions were wholly irrational, GMs did no better at analyzing than lesser players. Again, pattern recognition seems to be paramount in making the significant difference between higher and lower ratings.

     Pattern recognition naturally occurs after exposure to many similar situations, which in turn gives a player a kind of "mental database" of potential responses. This seems like a reasonable shortcut to tedious time-consuming if...then analysis. So it seems to be a matter of pattern-based intuition first with analysis afterward as a procedural measure to validate or dismiss the efficacy of the chosen candidate move. I suspect lower-rated players operate the other way round. De Groot theorized that masters have passing familiarity with something on the order of 10,000 positions, while grandmasters understand how to handle the subtleties of about 100,000 positions! De Groot did not go so far as to claim that GMs actually "memorized" positions, only that they had a superior "feel" for situations seen time and again in their chess studies and playing experiences.

     It seems reasonable that compiling archetypical critical positions into a logically ordered collection might accelerate and enhance one's pattern recognition abilities. Dan Heisman's revised Improving Chess Thinker 2nd Edition employs de Groot protocols to give lower-rated players an opportunity to analyze several carefully chosen positions, and to subsequently see how players at various higher levels appraise those same diagrams. This also seems to be a good idea - access to the actual recorded analysis processes of strong players to compare with one's own recorded analysis. Discovering in plain language how strong players think when looking at a specific position (especially after you have done so yourself) must translate into improved skills. Heisman has you record your analyses first, then skip to the chapter containing analysis by other players who are in your rating class. Then you read how the next class up sees the same positions, and so on...

     Meanwhile, I'm looking forward to IM van de Outdeweetering's book, which is due to be shipped by Amazon around 1 October 2014.

Not out yet - soon, I hope!

i think if you play the same openings then you will learn to remember the same pattern because you see it all the time

I like your idea.

As you become stronger you will better understand why you come up with better moves in five seconds then begineers sitting across from you thinking deeply for some time.

If you have seen strong players play bullet it becomes clearer as well.

Almost any player over class A can play chess blindfolded. Everyone is able to follow forcing lines

Stronger players know by experience which variations will be critical and what problems can develop for both sides. This is developed by thousands of slow games and much detailed concrete thought. There is really no shortcut to this.

Chunking happens as a byproduct not just because we look at a certain amount of positions and are shown the right path.

Thanks, yes I have just recently got this book and I agree it is very useful in this area, also recommended, good advice.

(General note: looking back over old posts is as usual a bit annoying...but in general now I think you can't really shortcut board experience so you just have to get out there and put in the time and effort and things will come.)

Hello,

I am new over here. I am doing a research at Federal University of Parana State, Brazil for my PHD thesis about pattern recognition in chess. I'm developing a formal language to describe patterns. This language will be used to teach/learn about patterns and also as a knowledge base. This knowledge base will be used during a simulated game. A specially engine is being developed to recognize that language. The idea is to develop a engine that plays at a strategic level, following a plan based on pattern recognition. The plan and patterns will be formalized by human chess players and will participate in a simulated competition. I have read the comments in this post, and I see there is a distinction among tactical patterns and strategical patterns. Where can I find strong strategical patterns? Anyone has a good example and could describe it for me?

Thanks,

@lcfb2003

Strategical patterns usually refers either to the position of the pieces (good position vs. bad position or good coordination), or to the pawn structure.

Here are a couple examples :

http://en.chesstempo.com/positional-motifs.html#knightOutpost

http://en.chesstempo.com/positional-motifs.html#isolatedPawn

You can find other on the same page, or buy a book on strategy like this one which describes and illustrates a bunch of positional patterns : https://www.amazon.com/Winning-Chess-Strategies-Everyman/dp/1857443853

 @hicetnunc

Thank you for your reply. Can you answer me another question: how would you describe mate pattern like Philidor's Mate, in such a way it could be found even in very different positions?

Perhaps it would be helpful to look at Improve Your Chess Pattern Recognition by Arthur Van de Oudeweetering

https://www.newinchess.com/Shop/Images/Pdfs/9006.pdf

and Train Your Chess Pattern Recognition by Arthur Van de Oudeweetering.

https://www.newinchess.com/Shop/Images/Pdfs/9021.pdf

Perhaps it would be helpful to look at 1000 Checkmate Combinations by Victor Henkin,

https://web.archive.org/web/20140708101520/http://www.chesscafe.com/text/review803.pdf

1001 Deadly Checkmates by John Nunn,

https://web.archive.org/web/20140708085825/http://www.chesscafe.com/text/review806.pdf

http://www.gambitbooks.com/pdfs/1001_Deadly_Checkmates.pdf

and/or Fundamental Checkmates by Antonio Gude.

http://www.gambitbooks.com/pdfs/Fundamental_Checkmates.pdf

start with the first move and go from there

I thought a lot about chess patterns and decided to build a course on it on Chessable. It is available here:

https://www.chessable.com/common-chess-patterns/course/13348/

My blog with further details can be read here:  https://www.chess.com/blog/Benedictine/common-chess-patterns