Saturday, November 08, 2014

to AOX: i've reached 6 hours on "attackers"

after lots of discussion with AOX about chesstempo and how to gain real improvement, i offered to be a guinea pig on visualization practices.

he said the first step is to get about 6 hours of practice in for "attackers" on

i just surpassed 6 hours today.  i averaged about 32 pieces of minute.

AOX - now what?  should i be trying other visualization exercises that you list on your blog?


  1. I see, you did gain some nice rating here. I had problems to get further. I did reach a plateau in rating soon.. My experience was that the implimentation of a thinking process did help me to overcome my plateau in improvement. The thinking process was: look first for the queens : what is attacked by queen, is the queeen attacked . Then for the rooks then for the bishops, ...
    When you reach the knight and you see the knight is attacked by the queen and i missed that then i found an "error" and the place where i made the error
    I did this for 2? hours and after that i was instantanious able to reach my present level ( without usuing the thinking process )

    I did wonder if others experience the same problems and hit a plateau early, and if they do, if they can overcome the plateau by a thinkingprocess too. That would have been a hint for me that "its all about" thinkingprocess

    But it seems as if you dont need such a method, seemingly you did not hit a plateau? Or did you use a thinking process right at the beginning?

    So i have to release you from your "job" as quinea pig.
    Seemingly "anyone" can reach GM-level at such a board vision task just by exercising long enough

    The implementation of a thinkingprocess might not be necessary, but you still may like to try?

    I still hope that the implementation of a thinkingprocess might help to improve in tactics. With such a tp you can find the moment where you made the error and try to make this better

    For example:
    When i see a CT tactic puzzle i cout material and then i make the decision: Mate / Gain Material / Advanved Pawn
    Then i have a quick view : which weaknesses did the opponents move leave behind and only then i start to go ahead.
    So if i judged: the puzzle is about Mate, but it was, say, a fork, then i can analyse further why i did make that error

    But ... i am still sitting at my plateu at CT... :/

    I discouvered a strange phenomenon: while it is seemingly impossible to improve in tactics, its more or less easy to improve at such easy drills. So i was wondering: how "easy" should an exercise be to be "improvable"

    So i made a "Mate in 1" trainer which is seemingly at least very hard to improve. Maybe you like to test this tool to see if you can improve.
    Tomasz and Munich try to improve by "brutal force" they solve hundreds and thousands of puzzles a day, Tomasz might have a small improvement, Munich not.
    I try to improve at "Mate in 1" by doing related board vision exercises..

    1. It looks like I can efficiently solve just the most trivial puzzles (mate in 1 easy). All the rest requires CALCULATIONS and it means - some (hard) work. I am a very impatient person and I have to force myself "not to guess" the solutions.

      Anyway I am very curious if doing other kind of excercises will influence my speed at solving "mate in 1 easy" puzzles.

  2. what's the difference between your mate in 1 problems on your blog and the mate in 1 problems on chesstempo?

    1. IMHO - these are the best "mate in 1 puzzles" I have ever seen or solved. They should be OBLIGATORY to all the kids who wants to see how well they understand and recognize the mate in 1 idea. After solving 20-30K of such puzzles - you should posses the feeling what kind of mate is possible. At the meantime it may help you "to trigger" some mating and attacking patterns when you learn or meet the "sharp positions" (or these where the King is attacked and may be easily mated).

  3. At the Moment my Mate in 1 trainer has almost 100 000 puzzles taken from real games and CT has 3721.
    These mate in 1 are so easy that it is no problem to do 2000 attempts a day ( as Tomasz does, i usually do 300-600 a day ). At chesstempo you have enough different puzzles for 2 days.
    An othher difference is: at chesstempo you have to wait a while for the next puzzle and you have to click at the piece and the destination square. At my trainer you click only the destination squres and the next puzzle is displayed instantly.
    The nest difference is: at chesstempo you need a custom set to do mate in 1's but customsets dont have a timed rating. But at such easy problems you score almost 100% anyway. Here its better IMO to measure the speed how quick you can solve these puzzles. As quicker as better

    See for example:

    or our discussion here