Saturday, February 12, 2011

Dissection of a microdrill

Uwe asked:
You did 100+k tactical exercises. Why is your tactics-rating still "so low" ( meaning not at GM-Level )?

Indeed my exercises only gained me a measly 300 OTB points. And if my main goal wasn't to find out how stuff works I would have been a bit disappointed for sure. Luckily that isn't the case and most of the time I was happily aware of "spilling my time" for the sake of exclusion. On the road of exclusion chess improvement nonsense falls from the bandwagon every mile. The reason of wee improvement sofar: bad methods. The methods worked at first exposure but overtime the returns diminished gradually.

When one ages precision in exercise methods becomes paramount. High time to dissect the latest microdrills in order to find out what the decisive elements are. What is critical and what are frills.

I'm speaking of the drill of naming the squares. A little silly maybe, but we are looking after principles here and this is a nice and clearcut example.

First let's talk about what worked not.
I estimate that I have recorded games in a state of timetrouble for about 50 hours over the past twelve years. Especially when I played gambits timetrouble was the norm since you have to make the game. The past two years my time trouble has become a rare animal.
This 50 hours of "training" wasn't enough to convince my brain to transfer the task of naming the squares to the automatic part of my brain.

What are the adding elements of the drill?
The past 16 days I trained about 10 minutes a day pointing at squares at random and naming them. That is about 2.6 hours total. When I have spend 3 hours I expect to have mastered it to name the squares a tempo with no noticeable delay and without hesitation. If I call the name of 1 square a fact then I need 3 minutes per fact.
So for transferring a task from knowledge to automatic the following is needed:
  • total traning time 3 minutes per fact.
  • 120 repetitions per fact.
  • speed 1.5 seconds per fact
  • conscious feedback.
Compare that with the old "training method" of being in time trouble:
  • total training time 47 minutes per fact.
  • 100 repetitions per fact.
  • speed 30 seconds per fact.
  • No conscious feedback.
If you compare these two then there seem to be two critical factors: speed and conscious feedback. Conscious feedback I have found before. But speed has never been a serious consideration to me.

Critical factor 1: Conscious feedback.
That one can't learn or train anything on autopilot I have found often and I have written a lot about conscious feedback before. Simply seek for conscious effort, autopilot, feedback etc.. In this post I try to explain why when you are young everything goes but when you become older you have to add intensity to your consciousness in order to make progress. I assume that interference of knowledge plays a role too when aging.

Critical factor 2: Speed.
There are several reason why I always have dismissed speed as being critical. To name a few:
  • Blitz play seems to have no significant effect on improving slow games.
  • The saying: First you have to learn to do it slow. Speed comes with experience.
  • CTS training was all about speed. But it didn't work out.
When I reconstruct the name of a square it typically costs 2 sconds or more.
When I recall the name automatically it feels like "immediate".
The average speed of training is 1.5 seconds which lies somewhere in between these two. If you train too fast there is no time for feedback, if you train too slow then there is no need for the brain to transfer the task to a fster part of the brain.

Factors that are not critical.
  • Amount of repetitions. To make any fact an element of your knowledge 7-10 repetitions at a regimen adviced by the theory spaced repetition will suffice. But to transfer a task from knowledge to skill 100 repetitions can be insufficent when it is not done with speed and conscious feedback. As you can see above.
  • Spacing between repetitions. I don't know yet if there is any reason to spread the training of a fact overtime or that you can do it right away the full monty. I'll have to try.

I losely introduced the word fact as a means to compare different training methods. But when the skills become more complex it is easy to see that the word fact becomes rather meaningless. I mean, how many facts do you train when playing blindfold chess is an unanswerable question. Yet the term has served it's purpose.

Mutually exclusive tendency.
The two critical factors speed and conscious feedback have a tendency to be mutually exclusive. That's why it is so hard to improve at chess. Blitz and CTS don't work since there is no time for feedback. While Standard Chess Tempo is too slow to trigger the transfer. So the tasks and the speed must be chosen carefully. Speed is the trigger for transfer but conscious feedback is slow by nature.


  1. "if you train too slow then there is no need for the brain to transfer the task to a fster part of the brain."

    Some ideas:

    There is CTS with about 3sec, Blitz at CT with maybe 1 min and Standard CT with several minutes.. one should do?

    I see the at CT some doing an easy ( low rated like 1500-1600 ) trainingset over and over again, with low error-rate! and seemingly automatic increasing speed. This set has 3158 problems, more than you can easily learn by heart. Typical speed about 1 min. I think this method is requested by Dan Heisman.

    How about: "back to the roots".
    De la Mazza did a set of about 1000 "good" chosen problems with increasing speed.

  2. All the 5 idea's you mention I have done IN EXTENSO. All idea's fell from the bandwagon.

    Sofar I have found 4 methods that worked:
    First time exposure to new idea's.
    Blindfold chess (partly).
    Naming squares.

    The last 3 are microdrills.
    None of these 3 microdrills is decisive due to futility. But they point out the direction.

    So the quest is to find out which tasks are decisive and then develop a drill for them.

    The highest rated person in the world is 20. The dutch champion is 16. Ask yourself what that means for the ratio knowledge : skill.

  3. -"Ask yourself what that means for the ratio knowledge : skill."

    Jung people have more energy!, more willpower, flexibillity, are more awake, have more speed.. Older people have to compensate with experience, wisdom, calmness..

    "First time exposure to new idea's." helps improving in OTB. Khmelnitsky wants us to fokus our training at a catehory whe are weak in but
    -Without an improvement in tactics to masterlevel we can hardly get our OTB to masterlevel..
    - but maybe the improvement in Strategy, opening, endgame and so on can give a "sceleton" to improve in tactics again?

    -I watched a GM playing a Blitzgame with a member of my chessclub. The GM was so friendly to share ith us what he did see. Withot any doupt he did "see" tactics in a position like i do see a red spot on a white wall.

    how could microdrills help to see tactics, ..i have to think about this

  4. NM Dan Heisman said:

    Suppose you are an experienced player with good board vision, but
    have not extensively studied basic tactical patterns. Then you can see
    what is happening on the board, but instead of quickly and accurately
    recognizing basic tactical possibilities for both sides, you have to figure
    them out. This would be equivalent to knowing how to add, but not being
    able to recognize the answer to basic multiplication problems. For
    example, when presented with the problem 8x7, you would have to add
    up seven eights or eight sevens, instead of just answering 56. Therefore,
    you are not only slower, but less accurate, because knowing the solution
    reduces error in trying to figure it out.

    Most exercises are about 4521 x 9472. We have to drill the tables of multiplication though. What does that mean in chess?

    The tasks have to be simple. More like 7 x 8. But then by heart.

  5. "The tasks have to be simple. More like 7 x 8. But then by heart."

    "CT-Blitz preverence difficulty easy" and then extra cycles with the wrong-made-ones to learn them "by heart". ( Maybe to-slow-sloved-problems should be cycled to. )
    Thats what i try at the moment.
    You need to SEE the tactic and check it = about 1 min.
    To search for a CT-standard-tactic at OTB takes usually to much time because you have to look for so many things.

    Be aware of

    7*5 takes a little child 1 or 2 min before they learn it by heart. A Problem in Blitz at CT takes 1 min to.

  6. "Mate in 1" might be more in the sense of microdrill. At many Mate in 1 i have still to calculate. Next step maybe "short combinations"

  7. "The dutch champion is 16"

    As far a I know, Jan Smeets is 'already' in his early twenties :-). However the former Dutch champion was 15...

  8. I believe that conscious feedback is important. If you get the solution to a problem wrong, it helps to work out why you got it wrong, and what you could have done to get it right. (Being too slow can be “getting it wrong” in this context.)

    Finding simple chess tactics is mostly an automatic unconscious process (unless you are a beginner). If you get it right very quickly, so quickly that you did not have time to work it out, that proves that you are doing it on autopilot. If you do not solve a simple problem quickly, you have not demonstrated even temporary success at the skill.

    You need to find the solution quickly, and have as much time as you need to reflect if you do not. If you fail to get the solution, a fast time limit is bad, because finding the solution for yourself is much more instructive than being told the answer - but spending too long on a problem is bad time management.

    The number of repetitions is critical. If you do not constantly repeat a complex skill it will disappear. If you want to retain the skill to spot a tactical pattern, you have to practice that skill for the rest of your chess playing life (ideally every time you are just about to lose it). (Being too slow constitutes “losing the skill” here.) If you do not repeat a skill, you do not know whether you still have it.

    The spacing between repetitions is critical. If you space them too widely you will remember very little, and if you space them too closely you will forget most of what you have learned very quickly.

  9. Bright, you certainly have a point to work out. My experiences with the drills point in another direction than I first thought. I will elaborate on this in a new post this week.

  10. The SuperMemo people say that getting the right repetition schedule increases the rate at which you acquire knowledge by up to fifty times. This may be an exaggeration, but the basic message is right, according to the peer reviewed papers. See the early articles in my blog. The basic technique for chess tactics is simple. Pick problems that you can get right reasonably quickly. Practice them until you can do them very quickly. Redo them as often as you need to maintain that performance.