Friday, March 30, 2012

Adding intelligence


















.
.
.
.
I made a distinction between storage and retrieval for reasons of proper thinking. After some thinking I found the following: There are two methods to add intelligence.

Adding intelligence during the game.
The first method is adding intelligence afterwards. When there is allready a complex position we can guide our attention to prevent that we miss out crucial patterns.
You consciously think about the moves. This is what we all do when we try to solve a problem. This leads to a brainscanpattern that is typical for the amateur: we use the part of the brain that handles new problems. Allthough a thoughtprocess leads to improvement in this area this is not going to win the war.


Adding intelligence during study.
The second method adds intelligence beforehand. At the moment of storage.

I retrieve what is stored. I cannot retrieve what is not stored. So if there is something wrong with my retrieval, there is something wrong in my process of storage.

During the storage process, intelligence must be added.
During the 5 years of tactical training, I failed to do that in a proper way. The past week when experimenting with method A, I made the same mistake due to bad habits.

If you are able to store patterns with added intelligence,this will result in a better automatic recognition process during the game. This will show an entire different brainscan pattern which is typical for the (grand)master, you will use your long term memory. You recognize and remember in stead of seeing everything as new everytime.

Problem with CT.
The problem with CT is that you are not obliged to add intelligence. You can perfectly get away with just memorizing the patterns and the moves. That is not enough.

So I started to experiment with mnemosyne, which is a freeware program that works with spaced repetition. I store the position and a crucial question in a flashcard. The answer is usually in the form of a move or a square. The questions I make are targeting my main failure in the position. For instance:
  • What is the position about (mate, promotion, gain of wood).
  • Identification of the attackers.
  • Identification of the attacking squares.
  • Identification of the problems on the road from attacker to attacking square.
  • Identification of the targetsquares.
  • Identification of the targets.
  • Identification of the problems on the road from target to targetsquare.
  • Solving the problems on the road from target to tatgetsquare.
  • How can you falsify move x.
  • Which piece is pinned.
  • Which piece is overworked.
Usually one or two questions are enough to provide the cues for my main failures in the position. This way I hope to add crucial information at the moment of storage.

5 comments:

  1. identification the value of the sqrs?

    z

    ReplyDelete
  2. Nice graphic!

    I like flashcards. Twelve years ago or so, I invested in a set of rubber stamps created for postal players to create diagrams on postcards. I used them to make flashcards for tactics training on 3x5 index cards. I used these for several years, and then bought some software that let me create nice diagrams. Using Publisher, I created hundreds of tactics cards with diagrams on one side and solutions on the other. Now that ChessBase diagrams have improved in quality and flexibility, I use ChessBase 11 to create the JPEG files and create a text document with OpenOffice: eight cards to a page. I print these on sheets of card stock and then use scissors. I'm currently working with endgame positions, and no longer have the answers on the back. I still have the old cards and review them from time to time.

    ReplyDelete
  3. Sounds like a lot of work.
    Is it really worth it?

    If I repeat a failed puzzle at CT, usually I keep failing it if I did not understand it. Sometimes (but this is really not so often the case) I remember that I had the puzzle failed and I remember the right move - it just doesnt make sense to me. If this happens (but really, it does not happen often), then I move the right move and afterwards I go and analyse the position to see what bugs me with the right move. I read the comments, and I got more than once the surprise that I allready added a comment there. The puzzle needs a "needs different move" or a "more moves required", und usually is indeed not so easy, and I was right that there was s.th. disturbing about the puzzle.

    Anyway, my point is, that you dont "get away with it at CT" if you dont let it happen. You perfectly know when you understand the solution and when you dont.

    Again: It really hardly happens to find myself in such a situation.
    And so --> is it worth the effort to use an additional tool besides CT?
    I believe not. The extra effort is not "value for money" so to say. I would stick to CT only, and by sticking to CT only you are able to cover much more ground.

    But the idea is nevertheless interesting and worth to be kept in mind for later when you train s.th. else than tactics, such as positional questions. Tactics is not everything.

    ReplyDelete
  4. @Munich,
    with the low rated problems you use I can understand that. But I use problems that are rated about 1900. It takes a few minutes anyhow to work these failures out. To make one or more flashcards from it is about a minute extra work. The advantage is that I adress really the cues I need to learn and there is no temptation to just learn the patterns without adding intelligence.

    But of course practice will tell if it is a viable method. The upside is that it is not limited to tactics indeed, but I can put all my mistakes into flashcards.

    ReplyDelete
  5. Tempo, this is a good point. I am not doing "difficult" puzzles, and I could imagine my rare problem might happen more often at a more challenging range.

    I let you get away with it then, for this time! :-)
    Seriously, the 1 minute or 3 minute extra effort is not wasted. The process of making such a card is keeping your thoughts around the puzzle, and thus you will memorize it better even if you would never look at this card again (when you come across the puzzle again, cause I doubt you check every green solved puzzle if a flash card exists for it. But you solve it green because you invested time into making this card and keeping your thoughts about it for quite a while).

    Useless bit of information, but it might be of some interest:
    I suddenly added 36 CT Blitz rating points to my last ATH, which I achieved in the middle of February this year.
    So my latest CT Blitz ATH is at 1936.5

    I reached an ATH Fide estimate of 2031, which is +41 points above my last ATH which I achieven also in middle of february.
    For the feeled strength I believe I did improve all of a sudden. So my training outcome made a sudden step upwards. I was wondering why my chesscube rating made such a huge jump while my CT Blitz rating was just a little bit better. But now all goes smothly in line: I did get better.
    My ATH from February 2011 was 1854.5, achieved with a rather high RD, and I hardly could keep it at the 1850s.
    My 1936.5 ATH was achieved with a RD below 35. So it is at least a 82 rating improvement within about a year. If you take into account that my 1936.5 suffers from a reduction in received point reward for solving a puzzle I had before but cant remember it, then these 1936.5 are even more valuable.

    Lets make it a guestimate of a gained increase of 100 fide elo points then?
    I am of course very happy about it, but it is nothing compared with the "success" of DLM (which I dont take seriously), but the advantage is, that it is real, and everybody can copy "my" method (well, I took your ideas, emperical rabbit ideas and aoxomoxoas ideas, with whom I discussed these ideas a lot, before I discussed them with you or the bright knight).

    I am curious if taking a higher puzzle range is actually even improving this kind of training. It might eventually improve it indeed. Besides of "good" arguments for the easy range (which are not sooo good that I am totally convinced myself): In my case it makes sense to stick with the easy range, because I started it, and I cant go to a different range unless I am willing to forget what I learned so far. I need to do the repetition of "my" solving range. I cant change now (but at least I dropped the ultra easy range and cut my range down from 1150-1475 to 1300-1475). But after I am through with them (in maybe 3 more months to come), than I intend to step up with a more difficult range. I havent made up my mind how difficult I want it to be, though. It probably makes sense to continue with 1475-1600, not leaving a gap between the trained ranges? Or shall I leave a gap and make it a 1600-1700 range? Dont know yet.
    click on my name for my stats or follow this link:
    http://chesstempo.com/chess-statistics/munich

    ReplyDelete