More of a focus on playing

Again, I haven't posted for a while. Partly this is because I've been busy at work. And partly I just haven't felt so enthusiastic about MDLM training.

I have still been plugging away at CT Art. Running it on a mobile device has been a real benefit, as I can fit in the odd half hour at lunchtime etc, and if it wasn't for this then I suspect things might have ground to a halt.

So I have been struggling through the 30s. The 10s are now easy, the 20s are OK and I can usually find the answer fairly quickly, but the 30s are worse. Not that I can't solve them, but that it takes that bit of effort. I find it difficult to get the motivation to solve more than 5 or 10 of them a day, and as a result it is taking a while to get through them.

However, recently I have been focusing more on playing 5 0 blitz on ICC. My real goal is to get a peak rating of 1500 here, and I'm not too far off. Here's my latest ICC rating graph:



A - when I first started playing on ICC, and (rather optimistically) said my rating was 1400

B - brought down to earth with a real rating of 1000

C - 1300 peak, after my first session of training, which was mainly reading opening theory and solving puzzles. This is also when I first discovered - and started - the MDLM training programme

D - 1300 peak, still, the last time I posted this graph

E - 1400 peak, after more focus on playing

So I'm getting there, and it has been interesting playing. I think that tactics are a large part of the game, but are not the complete answer. I find that I also have to have a 'plan' - even if it is something as high level as 'attack on the kingside' or 'push the pawns in the middle'. I'm also finding that if I'm controlling the game more, and my opponent is reacting to my attacks, then I'm more likely to get a tactical shot ... I suppose this is what they mean by initiative. However, I have to be careful not to over extend, or to push an attack to far when it isn't completely sound ... something I can get into the habit of doing after too much CT Art.

That said, the main reason why I've gone from my all time low of 1000 to my all time high of 1400 is tactical training.

I've won my first tournament!

I've not posted for a while - been getting bogged down in CT Art level 50 problems, which I have now finished. I've started going through from level 10 again. Rattled those off in about an hour, and now working my way through level 20s. I can do these fairly rapidly, but not at the quick fire pace of the 10s. It will probably take me the best part of a week to get through them.

But more importantly, I have entered - and won - my first chess tournament! I feel a bit of a fraud though really, because it wasn't exactly a challenging tournament. I discovered it while wandering around one lunchtime at work - there was a notice in a cafe window saying "amateur rapid chess tournament". I phoned up and it was being run by this guy trying to start up a local chess league between cafes and bars in the area. Almost all the people who turned up hadn't played in years, and certainly weren't involved in intensive MDLM training! So I felt like I'd been cheating by using CT Art.

Thinking about it the day after, I concluded:
1) I'd forgotten just what it is like to be a novice. You miss all sorts of stuff and there are tactical shots all over the place. I now realise how far I have come since I first learned how to play, rather than looking at how much I still have to learn.
2) Most of the time I could tactically overwhelm people. I would just throw out a few opening moves and then win a tactical combination or two.
3) The most benefit I have got from endlessly repeating CT Art problems is from levels 10 and 20, and a bit of 30. If you get this stuff right, and rarely miss an opportunity when it arises, you can significantly increase your playing strength.

Slow progress through the 50s

I've been working my way through the level 50s. It is slow progress, but I'm enjoying it. I've completed about 40 of them, and I'm scoring about 60%.

They are definitely more difficult problems. I'm really struggling to try and see all the possibilities, but I'm improving a bit.

This made me think a bit about what I'm really learning from this tactical training. My feeling at the moment is that it could be two things:

1) My ability just to see more tactics in the position. By this I mean instantly realising that there is a concealed check, a pin etc. And that this is reinforced by iterating through the problems a number of times - especially the lower level ones

2) Increasing my calculation depth. Being able to see the result of a longer sequence of moves, and also being able to visualise the resulting position more clearly and spot tactics in that position. For example, being able to see a knight fork after four or five forced moves. I feel this is reinforced by trying to do the harder levels. Although when they get too hard it can feel counter productive.

The novice's view of chess

I overheard an interesting conversation between two people at work this week. One person was describing a technique a mate of his has for interviewing DBAs - he plays them at chess using yahoo instant messenger.

The other person questioned this technique. He said that if he were subject to such an interview, all he would do would be to google for grandmaster chess strategies. He would then be able to win easily by just following, presumably, the simple instructions he finds.

There is also a Alta Vista add featuring Kasparov where a kid playing a simul types How to beat Kasparov at chess into a search engine.

This seems to be a common misconception amongst novices. Googling for grandmaster chess strategies gets you a lot of links to Silman's books, and how to beat kasparov at chess gets you a link to IBM's deep blue page! So you either have to spend hours reading Silman, or build yourself a super computer ... neither easy options for passing that DBA interview.

Repeated level 40 until I got them all right

Although I completed level 40 in CT Art a few weeks ago, I decided to continue to solve the 'erroneous' problems until I had scored 100%. In other words, to redo every problem until I got them all right first time. This is the same approach that I did previously with levels 10, 20 and 30. Although it is nice to achieve, it does involve trying to solve the same problems over and over. And, in the end it seems more a case of memorising the move order rather than solving, and I wonder if this is a bad situation. Am I building up a generic pattern recognition memory that will improve my tactical play, or am I learning how to solve specific CT Art problems? Also, it takes a fair slug of time to complete, especially at level 40. On balance, I like it though, and I feel it is helping by repeatedly exposing me to patterns that I miss. I definitely have tactical blind spots, and certain combinations are much harder for me to spot.

I have now completed seven circles of levels 10 and 20, six circles of level 30 and five circles of level 40. Above this, I have solved level 50 twice, and levels 60+ once.

Even with level 20, I have not managed to complete them all in one day. I have with level 10, and got it down to about 1 hour. The best at level 20 is 3 days. I'm getting a bit better at level 30, and managed to knock off 30 odd without really trying this morning.

According to MDLM I should be aiming for completing everything all in one day. At the moment I think this is unrealistic. I also think there is limited value going above level 50 problems. So my goal for this year is to try to complete levels 10 to 50 only in one day.

I am not fully convinced of this tuning the problem solving down to one day final stage - what do other people think who have achieved this? Personally, I find the iterating through the erroneous problems until you achieve 100% approach is more useful for me.

On another note, is the enthusiasm waning amongst the knights? People have commented that we are not as enthusiastic at rallying round people who are struggling. And also, I feel there is a lot of divergence in methods and more of a focus on chess improvement rather than MDLM methods. Is this bad?

Progress

I have completed the level 40s, and scored about 64%. Not too bad. I'm now repeating the ones I didn't get 100% right, iterating through them until I do. I did this with the previous levels and found it a good way to improve.

Also, my 5 0 blitz rating on ICC has improved - my peak rating is now 1421. My goal is to try and reach 1500 which now seems in sight rather than a distant dream. I am certainly improving. Mostly this is as a direct result of the tactical training. After one year of tactical training this peak has risen by just over 100 points. Not the 400 points in 400 days though! But then I do have a job.

John Nunn was an early knight!

Still plodding through the level 40s, but it is going OK with my new approach. I'm taking my time and trying to think deeply about the problems - I even get a surprising amount right!

On a more interesting note however, I came across this bit in Jacob Aagaard Excelling at Chess:


It is a game John Nunn played against Portisch in the Reykjavik World Cup 1988.

Black played 31 .. c5? and Nunn followed up with a nice tactical finish:
32 Re4 Rg8 33 Qxh7+! 1-0

Aagard says:
"Nunn had obviously intended this combination. It is no coincidence that in his Best Games Collection he writes that, as a child, he had solved every combination (999 in all!) in a book and this very same theme had featured."

Sounds like knight training to me, and even a grandmaster level game can be decided as a result!