Friday, March 18, 2005

New Math?

Leave it to me to add up the number of problems incorrectly. It only took one complete cycle of problems longer than it should have for me to catch the error. I am actually doing 600 problems each cycle instead of the 500 I gave in the original layout. I was beginning to wonder why I was a couple of nights behind heading into the second weekend. Figured I had taken one too many nights off. Then it dawned on me that the original idea of taking a rest day each week wouldn't fit within the 2-2-2 plan, if I wanted to maintain the same start and stop days. I even have it on paper as not working in my sketch pad of ideas. In my haste and desire to allow for the club night I somehow managed to view the second grouping of 50 problems as just one pass when I was adding up the totals. That's what I get for trying to squeeze everything into a 2-week time frame.
Sorry for the confusion. So here is the corrected version which allows for time off if you don't mind starting and finishing on a different day with each complete set of 100 problems.
Really it hasn't been a problem to tackle the problems on club nights, and it actually seems to serve as a warm up. I think the only night I can remember not doing problems prior to playing at the club was this past week, which may explain why I wasn't seeing anything at the board.

(4 nights x 25 problems=100) 2 passes = 200 problems [200 completed total]

(2 nights x 50 problems=100) 2 passes = 200 problems [400 completed total]

(1 night x 100 problems=100) 2 passes = 200 problems [600 completed total]

Here are the results of the previous 2 cycles.

Cycle 1 96% success ratio

Cycle 2 97% success ratio

1,200 completed 4,800 to go.

How do I miss mate–in-one exercises? Well it’s easy if you start to push on how fast you can complete the task. Normally I can do them inside of 10 seconds, the fastest has been 2 seconds. The average is 5 seconds per problem. It takes about a second or so to get your bearings with the side move, and just locating the enemy King. Yes I know rushing is a bad habit, but it something I wanted to try just to see how fast I could solve these easier problems. Back to solving before moving now that I’m into the Mate-in-2 problems. I promise no more blitz!

Also I started working through Jacob Aagaard & Esben Lund’s “Right Decisions” CD.

This really is an excellent CD, I picked it up after purchasing Aagaard’s book “Excelling at Chess Calculation” which I’m currently reading. (I think I mentioned that I had just purchased the book during our first online meeting of the Knight’s Errant a few months ago.) This is the first time to my knowledge that this has ever been done. (Averbakh’s “Comprehensive Chess Endings” gets honorable mention.) Meaning a CD created by the author to further reflect his ideas of his written work. Where Aagaard differs from Averbakh is in the fact that he mentions the companion CD in his book, and not the other way around.

Just imagine if Kasparov published an interactive Chessbase version of his “My Great Predecessors” works showing all of his analysis while personally talking us through the games. It would be huge.

Wednesday, March 09, 2005

Thoughts continued...

First Welcome Back Jim! And I finally got around to updating the links of the ever expanding list of family and friends.

Here is something else to think about. When Don and I first started the Circles, my success rates were slightly higher. We never once looked at this as a competition, we only added our success ratios because one the readers requested that we do so, plus it made perfect sense to keep a record to use later for a reference point.

But as the Circles reached the higher levels Don's calculation muscle kicked into high gear. While I could still solve the problems matching or bettering my earlier success rates, my ability to solve them more quickly with each consecutive pass didn't improve as dramatically as Don's. He soon left me in the dust with his success rates when it turned to the critical portion of timed exercises.

In hindsight I believe it was because Don took great effort to use his full amount of time per exercise in the lower circles. I saw or guessed at the answer and moved on, all the while growing increasingly displeased at the random order of themes involved, thinking of CT-Art as a sloppy way to learn tactics. Of course I was spoiled by my earlier exposure to the excellent training CD made by TASC called “Chess Tutor”. I later found out that this was a highly successful 5 step method pioneered by Rob Brunia and IM Cor van Wijgerden. It breaks tactics all the way down to the fundamental elements. For example when describing and teaching pins it gives the themes of King + Material, Material + Material, Material + Square. I cannot give this CD enough praise, every beginning to intermediate player should own it, or at least be exposed to its methodology.

Anyway there I was, focusing on a pattern recognition/theme approach because at the time I thought that was the key to improved tactical ability and the basis of Michael de la Maza's plan. Now I realize that I couldn't see the forest because of all the trees. But until the Knights were formed no one had really thought about the difference between pattern recognition and "Calculation Muscle".

So really Michael de la Maza owes his rating increase largely in part to a heightened ability to calculate variations quickly and accurately. That is not to say his ability to find the correct move wasn't intact, because I have a feeling he was very good at seeing patterns, nor am I dismissing what he achieved as an easy feat of accomplishment.

While he stresses the benefits as tactical improvement, he really should be calling it an improvement in calculating accurately. Because CT-Art used as he prescribes certainly serves no other function better than a calculation exercise. If you wanted to use it in another capacity you should switch the setting to either 1. Tactical Methods, or 2. Combinational Motifs.

Once I reach that portion of my training where I want to work solely on calculation muscle I will certainly consider the MDLM version of 7-Circles. I will still be hesitant about using CT-Art but you never know...

Actually it would feel kind of good to slay that buggy piece of work. Technically I guess
I could finish my remaining circles but that would kind of miss the point. If I’m going to do something I want to do it properly.

I'm wondering if this post breaks the "obtuse" barrier.

Tuesday, March 08, 2005

Congratulations Don! and Hang in there J'Doube

Way to go Randy!
That was a pretty arduous task!
Cheers!



Jim,
I certainly understand your frustration with the program material. I certainly wasn't happy with it. I felt it was too broad in nature for actually mastering specific tactics. But what I have come to understand after taking a step back from the program and keeping a watchful eye on the other Knights/Non-Knights is what purpose CT-Art serves really well. It came to me while reading a post from Takches about pattern recognition versus calculation muscle. That really got me to thinking about the importance of each, and their dependence on one another. Mainly it got me to thinking about what would be the best way to train each of the disciplines effectively. This is where CT-Art in a sense is useful. Not the best of applications mind you, but used with MDLM's training plan it will bring about improvement results.

CT-Art and the 7 circles build calculation muscle, plain and simple, and of course it does help with pattern recognition too. (If one really wanted to train calculation muscle exclusively the study of endgames would be a great way to start.)

This is something Don and I differed on initially approach wise. I like you, found CT-Art full of holes with the programming errors, duplicate exercises, etc. Drove me batty. I had little faith placing that much effort in something so flawed and at the time seemingly haphazard in its ability to teach patterns. Impatient was I, to the dark-side I turned... oops, wait wrong story.

I simply chose to take the path of training pattern recognition first, and what turns out to be calculation muscle second.

This is not to say one way is better than the other, you have to have a solid grasp on both if you want to make it past a certain level of playing strength. Matter of fact I don't believe either one would last very long without the other.

I found starting with something simple as a basic forced mates CD and working through that also holds the ability to train calculation muscle.

Of course the variations are much shorter, and the number of candidate moves one has to list is much smaller. But is this really a bad thing?

How can one isolate and identify a thought process error if he/she is overwhelmed by the sheer volume of possibilities present. Likewise how can one build a check list of search patterns/to-do list unless you can clearly identify the strongest move or threat in a given position? Doesn't it make sense to hone one's ability one move at a time? Then expanding the depth of search move by move. Studying mates allows you to filter out the “noise” found in the position. Personally I don’t mind solving exercises with the same mindset as practicing scale patterns a million times, or shooting a million free-throws if it gives me the ability to do said exercise blindfolded and upside down. As long as I get the correct answer and without hesitation.

So anyway I hope my obtuse ideas about training methods will be enough to persuade you to stick around, or at least get you thinking about coming up with your own plan.

Sunday, March 06, 2005

Go Don!

Don is very close to the end of the 7-Circles.
It will be interesting to see what kind of experiences he has encountered over the last few days.
I'm hoping he will give us a detailed report of his mental state and thought processes.

Once Don completes this particular journey I say we all raise a glass and give a toast.

Wednesday, March 02, 2005

2-2-2

Two Pizzas, Two Toppings, and a Two liter beverage Mr. Pawnza?


No, actually it's my new method of working through tactical exercises.
I take a group of 100 problems total then I proceed to work through them in smaller blocks of 25 problems per night for two complete passes.
Then two passes at 50 problems per night.
The finale consists of two passes at 100 problems per night.
Simple and effective. I get exposure to the same problem 5 times, and I complete all of the groupings within 12 days. This allows for two complete days away from chess if I choose, or another form of chess activity. (I.e. Club night)
Another bonus is the final two days can be tackled on a weekend if one starts on a Monday. (Including the “off” days.)
Of course the starting number of 100 is one that I picked at random based off the material I am studying currently. (1000 x Checkmate CD)
I just filtered the games by number of moves, and created a database that I copied to my hard drive for ease of use. Plus it allows me to track my success rate by right clicking on the DB, selecting properties, and then the training tab.
Since I'm working through mate-in-one problems at the moment I figure 100 problems would be a manageable figure. The bottom line of course is the success rate, on which I have set 90% as being acceptable. If I fail to meet that target number I will simply have to repeat that particular set of problems again for the next 12 days.
Fortunately I met the minimum during the test run, so it is now on to the second group of 100.
I wanted to play guinea pig before I mentioned what I have been cooking up during my blogging absence.

Ahhhh! It is so nice to have a life again.