Wednesday, August 31, 2005

Goals

If I'm going to do this, then there should be a goal in mind at the end.. otherwise how can I measure if it was a success or a failure?

So the goals I've choosen are:

1. Higher performance rating than in my previous tournament
2. Higher percentage result than in my previous tournament
3. Higher position than in my previous tournament

This means I am aiming for:
1. Performance rating higher: was 1473
2. Percentage result higher: was 4/6 = 66%. Since it will be a 5 round tournament this means 3.5/5 or better.
3. Higher position: was 4th=

It is probably wishful thinking to expect that I will achieve all 3 goals so lets define my level of success as:
3/3 goals: Very successful
2/3 goals: Successful
1/3 goals: Somewhat successful
0/3 goals: Fail

Now let the training begin, and we will see what the result is in a very short time....

I'm not going to aim specifically for a higher rating, the rating system here is too slow (new lists every 6 months) - a better performance is a good goal. My rating is 1400, and should go up slowly as a result of better performances. Both my previous tournament and my next tournament are U1700, but the next one should be harder as it is more popular.

A list of my previous results, most recent at the top, oldest at the bottom:

(Performance rating, result, score)
1473, +2=4 -0, 4/6
1688, +3 =2 -0, 4/5
1580, +1 =1 -1, 1.5/3
1473, +2 =0 -3, 2/5
1249, +2 =3 -2, 3.5/7
1311, +3 =0 -2, 3/5
1370, +2 =0 -2, 2/4
1477, +3 =0 -1, 3/4

Tuesday, August 30, 2005

Going for broke...

With just 38 days to go until my next weekend tournament, I wondered how much training can you squeeze in if you really tried??! And what would be the result? Would my brain be a lean mean chess playing machine, or would it just turn to mush?!!?!!

I've always liked the idea of the Michael de la Maza method, but havent really tried it, instead I've just slowly plodded thru a couple of tactics collections, doing a few problems every few days. This has definately helped - I just scored an undefeated 4/6 (+2 =4 -0) for 4th= in a weekend tourney a couple of days ago.

So if slowly doing tactics helps, what can results might I see from a tactical training blitz ala MDLM? Its a good question, but there are some immediate obstacles ahead ... I only have 38 days, not 400, and its quite possible my initial burst of enthuisiasm will fade when I really realise what work I've gotten myself into!

First I worked out how many days are available: 38 actual days between now and the next tournament on the 8th of October. Now minus chess club nights between now and then 38 - 5 = 33. Now round down so that there is some free time available just in case: leaving 30 days available for training.

Next thing is to work out how many tactics puzzles to try to solve? A quick blog search reveals someone who tried 256 problems in a month, but I want to try to at least 500(!). It took me many attempts of writing out numbers on a scrap of paper before I figured out a possible method.

The main problem was that, for the first time thru the puzzles, I can realistically only solve 25 per day. And 500/25 = 20 days, just for the first circle! That is most of my 30 days gone already....

In the end I came up with this following plan:

Circle 1: 25 per day times 20 days = 500
Circle 2: 125 per day times 4 days = 500
Circle 3: 166 per day times 3 days = 500
Circle 4: 250 per day times 2 days = 500
Circle 5: 500 per day times 1 days = 500

Phew! I feel tired just thinking about it...

Now the next problem is that 20 days of 25 per day is too long - the repetition should start earlier. By day 20 I will have completely forgotten days 1 to ? 19 maybe, 15 at least! So I had to devise a schedule using my new invention (I havent read it anywhere else regarding MDLM method, maybe someone else has invented it seperately) ... I propose using overlapping circles (tm)

This would make my schedule look something like this:
25, 25, 25, 25, 25, 25, 125
25, 25, 25, 25, 25, 25, 125
25, 25, 25, 25, 125
25, 25, 25, 25, 125
166, 166, 250
166, 250,
500

By mixing circle two amongst circle one, I start the repitition earlier, which should help the pattern recognition start quicker as well (or at least that is my theory!)

Ok, that was longer than I'd expected :) I will post updates ...