Pattern Recognition of Key Positions

Sort:
Benedictine

Wow thanks for letting me. It is surprising that such a tool doesn't really exist. It would seem to be a paper job then or sticking to winboard.

Benedictine

The best I have managed to come up with, which I'm quite happy with, is to use the games editor feature here and then save the file. I can create a games list and can then load it up through Winboard to keep reviewing each position. It's a shame the games editor feature here doesn't allow you to load in more than one position at a time though, as I have got used to the graphics interface which I think it better than the Winboard one. Still at least I have found something workable.

antonreiser

Hi Ben

 

some "here and there" ideas:

 

yes, i agree in hic et nuc's idea of storing 2000 patterns (tough i can olny talk from the "outside, of course, i am a rather weak player)...my point i s about how to get to it, and i think that that goal is better achieved mostly by doing thounsands of pulzzes and analysing many positions and not by picking some really nice positionsup and trying to learn them by heart as a recipe.

 

this is related with your comment on reading books; once Jorge Luis Borges said he was not sure to be a good writer...but , he said, was sure to be a great, great reader...of course if you read thounsands of books you instantanly  forget all about names, carachters and even plots and endings ("ooh, really.???.does Anna finally comit suicide in the last chapter?? uhmmm dont remember that, awesome..."-kind od story, yes), but in the by-process you get something different..you become a better reader, able to discenr subleties in a book that a non reader would never notice...E. Lasker said that too about chess...he said if you listen to a Symphony 10000 times that will not make a better musician out of you, but if you read 10000 games of chess , that will make you a better player.

 

i have read too tons and tons of literature, philosophy etc and can testify that that alone does not make your short term memory any better, BUT it makes wonders with the (sometimes subconscious) long term memory...the kind of memory you use all the time while talking, riding a bicycle or doing any other activity, (like chess i am pressuming, though i am not sure yet) requiring "practical knowleadge...

so..my point was not about not learning many chess, tactical patterns, but about how to get them usefufully stored in the brain for practical pourpouse. And Lev's idea of "the most 300 important" postions sounds to me ridicolus as no Gran Master in chess history has been sure about an objetive scale of important positions out of the thoushands of them, much less to pick uo the very 300 best. (and yet you know i love Lev's book as i love those positions, but just as i love any other good tactics book...well ok , Lev' is my best favourite, but the intro is more comercial than any other thing)

 

(i am really strugling with english trying to explain my view, so i am sure there is an aditional portion of surreal, nonsensical discourse in this lines,Embarassed, sorry).

zborg

Better to touch the pieces on a standard board, use a book for reference, and work through the problems -- in books by well established GM writers, like Nunn and Hellsten, to name just two.

Flash cards??  This is not about memory, it's about understanding and knowledge.

"Study Brings Wisdom, Practice Brings Perfection"  (back cover quote) Hellsten (2010).

If you insist on a digital solution, simply run every postion through Rybka, but don't expect to understand what it is telling you via the evaluation function.

You have to think, and touch, and move the pieces by yourself.  End of story.  Sorry.

P.S., 90 percent of active tournament players in the U.S. don't break USCF 1800.  So keep your expectations realistic, and judiciously consider the time you will need to devote to your improvement.  Buyer Beware.  Smile

IpswichMatt
zborg wrote:
Flash cards??  This is not about memory, it's about understanding and knowledge.

I think it is about memory. How do you know that 6 x 8 = 48 ? Because you memorised it. This enables you to do long multiplication - which requires understanding and knowledge, as well as instant recall of single digit multiplication. So it is that more complicated tactics are built on simple tactics. 

(I think that's the theory anyway)

VLaurenT

I think the dichotomy between 'memorizing' and 'understanding' is not really necessary.

Memorizing makes the material 'readily available' (think computer RAM) for further digging and more elaborate connections, thus leading to 'understanding'.

@anton : some GMs have actually selected their 'top positions to know' - and/or top games, the most famous example being Ziatdinov - please see my systematic training blog post mentioned before in this thread. Milos has also done something similar with his Chessimo tool, though he never stated it explicitly.

Benedictine

Hi Anton, no I know what you mean you explain it very well. The truth is I don't know one way or the other. It just seems a little impractical to go through thousands and thousands of games/positions/tactics etc and remember very little. As opposed to really committing to memory core principles/positions and then applying those to new situations.

Chess has unlimited possibilities but at it's core there are only a couple of dozen core tactical themes: remove the defender, pin, folk, skewer, etc, etc and a number of mating patterns: back rank mate, smothered mate, bishop and Queen mate, classic bishop sacrifice etc, etc the same with endgame principles and for positional play I don't know?

It just seems a little disorganised going through random tactics and random master games, 1000s and 1000s of them and hoping something sticks. Not that there is anything wrong with this additionally (and I still love 'Guess the Move' with the master games database on there) but I've craving and wondering about something more stuctured, along the lines previously mentioned. 

Kijiri

I just wanted to fill in 2 cents (keep in mind that I didn't read all the posts and that I'm "still" fairly new to chess having no otb tournament experience). 

You talk of key positions, I think pawn structures.

You talk of master players playing good in simuls without calculating much, I think pawn structures (and positional understanding).

Really, understanding the pawn structure and the strategy involved around that specific pawn structure, the important squares, the strong/weak squares that you create and your opponent has the oppertunity to create and being able to correctly evaluate how well placed your pieces are will do you so much good. All these things are easy to assertain without calculation. Based on those elements, you make logical candidate moves and then you calculate enough ahead to see that your position is sound (better infact in so far as your skill allows).

Either ways, my point is that the perimiters for choosing candidate moves that benefit the position is the very basis of positional chess and excactly how a good chess player can play a great move without calculating at all (or just calculating a few moves ahead). Obviously having a great tactical ability to see your own possibilities and the opponents threats are a must.

Now why pawn structures instead of openings? (because I could've said learn the strategic and thematic positions in your openings) To save time.  Learn to play an isolated q pawn position (both from black and white), because tons of openings lead to one. Learn to conduct a minority attack or deal with a pawn imbalance. Learn to evaluate a position positionally and figure out the weakness and strengths of both your own and your opponents position. Now that's knowlegde that you can apply to every single game of chess.


How? There's some excellent books that come to my mind. How to reasses your chess by Jeremy Silman, My system by Nimzowitch, How to play like a grandmaster by Alexander Kotov.. I'm sure there a many I haven't ever looked at. And then once you've learned the tools, study the games of capablanca, karpov, petrosian and start to notice how many moves they properly could make easily without calculating. 

lrmall01

The best free tools I am aware of for creating custom test sets based upon your own positions are Lucaschess - http://lucaschess.host22.com/ and Yatt - http://www.chess.com/download/view/yatt--yet-another-tactics-trainer

They both read a list of FEN positions and present the problems to you randomly, repeating the ones you miss.

I think other trainers at Chess.com and other locations let you have custom test sets, but using their positions. If you want to control your test set, these programs will help.

Good luck!

Benedictine

Thanks I've downloaded Lucaschess and like the simple layout. I've imported one position and like the analysis and book options. I want to be able to import multiple positions at once though, I'm just trying to work out how to do this. 

lrmall01

I can show you later when I have more time. It isn't very intuitive to get positions into the tactics menu. Lucaschess has utilities for converting pgn to FEN but you need to manually edit config files to manipulate the menus.

Anyway, don't want to hijack the thread but send me a PM and I can help.

lkmk

Benedictine/Irmall01, perhaps an easy way is : if you have a pgn with a lot of positions with their solutions, you could use Lucas Chess to read it, using Tools-PGN viewer-Read PGN, ... then option Utilities - Create tactics training. This creates a training accesible via Training-Training positions and via Training-Learn tactics by repetition.

Benedictine

Hi, thanks this is exactly what I've managed to do in the end, with help. I have set up the first 100 key positions from Lev's book (because that is where I am in it at present) and I have started to put in my own positions bit by bit. I'm also using it to help with opening preparation.

The only slight drawback with this is that with my own positions I have to manually copy and paste each one into LucasChess via Tools/PGN viewer/ Paste PGN. I then go to Utilities/Save/PGN format/Save to file.

I then create this file and upload the whole thing back the way you describe, by PGN viewer.

This is OK but it would be great if there was a way to get around having to do this manually. For example if i have 25 PGN positions all copied on a word document, is there a way I can save them all as one PGN file so I can just upload them using PGN viewer?

lkmk

Hi, if you save your word document with a pgn extension and with a type of format = txt, then you can read it directly with Tools, PGN viewer, read a PGN. 

If your word document have other elements than pgn code, copy and paste each pgn in a notepad txt file....

Benedictine

Hi, thanks. I've tried that but I can't save it with a pgn extention as that is not an option in the drop down box when I save it. I can save it as as txt extention but the Lucaschess only reads pgn files when you select the read pgn option.

It's not much of an issue anyway as it doesn't take that long pasted them in and then saving them to file. I'm getting quite quick at it.

lkmk

When you save from word, you can select type TXT in drop down and the name of file including extension, ie you write mygames.pgn .

When you read from LC (in case of saved with txt extension), you go to the folder , then write *.* and enter, and you can select file with any name.

Benedictine

No it's not done, never mind. I saved the file in that way and tried to read it in Lucas but it doesn't scan it. I can locate the file by doing *.* but it doesn't scan in from there, the bar stops about a quarter of the way across.

Knightly_News

I think your plug for the book sounded shilly.  Hard not to be suspiscious when you do a build up like that, then call out the book by name, then a ramp down.  Probably innocent enough, but still.  I would have, if I was serious about having people take my question seriously, explictly avoided putting an advertisement in the middle of my question.  I realize that you probably just like the book, and are excited about it or something, or want to share it because you think it is valuable.  But to avoid the appearance of impropriety,  I personally would have said said something like "I'm reading a book about..." but not actually name of the book and author.

Also, you refer to finding the optimal time to read. I believe the optimal time for many of us quasi-devoted players may be on the toilet. Sitting on the can is, in and of itself, a key position. Currently my toilet reading is gardening and landscaping books.  Why not chess books?  A man's home is his castle, the toilet, his throne, his kingdom is on the chess board.

Benedictine

Ha, ha I'm not getting royalites from the book. It was just an example and if it is a good book (which in my opinion it very much is) then I don't see the harm in naming it and the thought of not naming a book for advertising purposes hardly occurred to me on a chess forum thread! It's not as if I'm on live TV wearing a diet coke T-shirt, sorry a popular soft drink.

In terms of key positions sitting on the toilet is not the sort of position I had in mind, but thanks anyway.

lkmk
Benedictine escribió:

No it's not done, never mind. I saved the file in that way and tried to read it in Lucas but it doesn't scan it. I can locate the file by doing *.* but it doesn't scan in from there, the bar stops about a quarter of the way across.

If you want, send me the file and I convert it.