Monday, December 18, 2006

CT-ART 3.0 And A Status Summary

I finally got CT-ART, and have been starting with the level 10 problems. As usual with any chess programs, the interface is quite amateurish and has a lot of room for improvement. But I'm getting used to it, and I guess it gets the job done... The problems are quite easy, of course, but I've still managed to get 2 wrong from the first 79. I'm going to concentrate on calculation with them, an area that I'm having serious problems with, despite the extensive drilling on CTS.

I also ordered Karsten Müller's endgame dvds 1&3, and will start drilling them as soon as they'll arrive. I'm also continuing the daily dose of blitz, and hopefully get the hang of it at some point. Now, I just keep doing ridiculous mistakes I didn't even think I was capable of anymore, and the embarrassing rating in the 1100's descibe exactly how much I suck at blitz. In CC I got swindled out of about 40 points by a 1900-2000 rated guy who mass resigned around 100 games at once, leaving me with a whopping 10 points for the two games I won against him. I would've hit 1800 otherwise, but the 40 missed points combined with a typical missplayed ending against a 1600 have dropped me back down to 1735. Well, live & learn.

CT-ART update:
day 1: level 10: 96%, level 20: 90% 61/286 done.

Monday, December 11, 2006

Mating With KBB vs. K

This one I never even tried until a couple of days ago, as I thought it would we quite trivial to line up the bishops and force the king out of space. But now that I've actually drilled it, I realize there was a good chance to stalemate in time trouble, and that the efficient mating sequence actually demands some technique.

The general idea is to drive the defending king into any corner, and mate him. It's easiest done setting the bishop pair side by side as close to the defender as possible, confining him into a triangular area against the edge. Then you just move your king up, take a couple of squares and squeeze a diagonal off every time the defender retreats, until it's against the edge. Then you pick a corner, and force the king towards it by blocking it from behind with the bishops.

In the corner, confine it against either edge, line up the bishops to take the final squares with check, lose a tempo if needed, then check and mate.

A blitzed example against Chessmaster 9000:

[White "wormwood"]
[Black "Chessmaster"]
[Result "1-0"]
[Setup "1"]
[FEN "7K/3B4/8/8/8/2B2k2/8/8 w - - 0 1"]
1.Be6 Ke4 2.Bf6 Kd3 3.Kg7 Ke4 4.Kg6 Ke3 5.Kg5 Kd2 6.Kf4 Kd3 7.Be5 Ke2 8.Bd5 Kd3 9.Kf3 Kd2 10.Be4 Kc1 11.Bd4 Kd2 12.Kf2 Kc1 13.Ke3 Kd1 14.Bb2 Ke1 15.Bc2 Kf1 16.Kf3 Kg1 17.Bc3 Kh2 18.Bf5 Kg1 19.Kg3 Kh1 20.Bd3 Kg1 21.Bd4+ Kh1 22.Be4# 1-0

Sunday, December 10, 2006

Mating With KNB vs. K

I thought I'd make a separate post on all cases I'm going to tackle, so here is the first one. Maybe it'll even help someone else, although I'm probably not going to get overly detailed where the matters are trivial to myself.

I've tried to learn KBN several times before, so I was not completely new to the ideas behind it. Drive the enemy king to the edge, force it to the corner the color of your bishop, and mate it. Sounds simple, but in practice the king just kept slipping away before I could get it into the right corner. Once in the correct corner, the mate is pretty straightforward after some thinking.

Also the driving to the edge -part is relatively easy, just 'oppose' the enemy king with your king, restrict with K&B until it has no legal moves but to retreat towards edge, and invade all the files/ranks it gives you until it's against the edge. It helps to remember that the enemy king's best defence is to first stay in the middle of the board, and that failing to retreat towards the wrong coloured corner. So these are the directions you want to block. In practice this also means that it's usually easiest to drive it towards the wrong corner to begin with. Another tip is to avoid placing a piece between the kings (sometimes it works though), so that you'll be able to snatch the files the enemy gives you without letting the him slip away. I didn't memorize any specific patterns for this phase, but after some drilling against Chessmaster it became instinctive and easy, and I'm now able to do that in a bullet game with no thinking whatsoever.

The thing that kept me baffled until I got some tips from Dragon Fire and Varenka on RHP, was the part where you drive the king towards the mating corner along the edge. The beginning is pretty easy, you just keep the enemy king against the edge with your king, build an additional 'wall' with the knight when needed, and block from 'behind' with the bishop so the king can't return. At some points you'll also need to lose a tempo with the bishop to get the enemy king where you want it.

But around the middle of the edge, you need to let the enemy king slip off towards the centre to herd it into the right corner. A sort of 'leap of faith' is required, which of course turns out to force the stray king back against the edge, when properly executed. The trick is to form a mating net from the king onward, only it's not possible to close the gap all the way to the edge, just up to a single square shy from it. It also looks like the net leaks badly at first glance, so it's a bit hard to visualize, but the bishop has time to block the remaining escape squares even farther to the middle, forcing the king back towards the edge. After that, it's again relatively easy to complete the journey towards the mating corner. -All this I pretty much worked out by myself, but there was one final trick that's needed for the net to hold: You need to start building it the knight one square from the edge, instead of moving it to the centre as seems intuitive. So Nb5! instead of Ne6 if you're driving the king from a1 to a8. After that, the king might try heading towards c8 which looks like a way out, but the bishop pounces over to the opposite flank to block it just in time. The net is still a bit sparse, and the king can turn back and try another break for the freedom via c6, but again the bishop is on time and the net closes up permanently. Nowhere to go but back towards the edge with all hope lost.

There is one more final little 'leap of faith' where you hang the knight for one move on Nb7+ to force the enemy king up, but it's easy to see and the attacking king simply moves one square up to defend the knight as the enemy king gets on striking distance.

Then it's only the mating itself left, so you just restrict the enemy king into two final squares, position the knight so that you can take the second to last square with check at will, lose a tempo with bishop to avoid stalemate if needed, then check and mate.

I personally like to set up a fixed position in the wrong corner to start the cattle drive, so I'll have a nice fixed series of 20 or so forcing moves up to the mate that I can blitz through. It's not the fastest way, and requires losing a tempo here and there, but it's systematic and reliable. There's really not many points in which the defending king has a choice, so it's pretty easy to master after the initial learning phase. maybe an hour or two of continuous drilling.

Here's an example I blitzed against Chessmaster, moving immediately after it did. It's not the fastest way to do it, but it works.

[White "wormwood"]
[Black "Chessmaster 9000"]
[Result "1-0"]
[Setup "1"]
[FEN "8/3B4/8/5K2/3k4/8/8/5N2 w - - 0 1"]
1.Ng3 Ke3 2.Bc6 Kd4 3.Kf4 Kc5 4.Bf3 Kd4 5.Nf5+ Kc3 6.Ke3 Kc2 7.Nd4+ Kc3 8.Be2 Kb4 9.Ne6 Ka3 10.Kd3 Kb3 11.Nc5+ Kb4 12.Kd4 Ka3 13.Kc3 Ka2 14.Bd1 Kb1 15.Nd3 Ka2 16.Bc2 Ka3 17.Nc5 Ka2 18.Nb3 Ka3 19.Bb1 Ka4 20.Nd4 Ka5 21.Kc4 Kb6 22.Nb5 Kb7 23.Bf5 Kc6 24.Be6 Kb6 25.Bd5 Ka5 26.Kc5 Ka4 27.Bc4 Ka5 28.Bb3 Ka6 29.Nd6 Ka5 30.Nb7+ Ka6 31.Kc6 Ka7 32.Bc4 Kb8 33.Nd6 Ka7 34.Kc7 Ka8 35.Bb5 Ka7 36.Nc8+ Ka8 37.Bc6# 1-0

Basic Endgames

Now that I have reached 1600 on CTS, I'm finally redirecting my main efforts from tactics to the area I have neglected most since I started chess 1½ years ago: The basic endgames. -The plan is to work my way up from the very basics to at least the fundamental rook endings, and probably a bit further. The study plan is to go through Karsten Müller's endgame dvds I, II and III, and drill the cases against fritz until I can blitz them through from any relevant starting position with no problems.

I'm not completely sure I'm going to get the first dvd though, as I already know a lot of the material. But not all, and there's probably quite a lot of useful commentary on it as well, so maybe I'll get it after all. But for now, I'm just gonna seek similar free material and work on that.

I'm beginning with the basic mates/draws I haven't yet mastered. KNB, KBB, KR vs. KN etc. The first one being the KNB, which may be quite irrelevant to practical endings as I might never face it, but I think of it as part of the craft and a useful lesson on piece coordination and board vision.

Monday, December 04, 2006

Blitzkrieg Pop!

I finally started playing blitz on FICS to get more experience and routine. I'm trying to play at least one 7 10 game (that's about 14 minutes total) a day, and go through them with light analysis. As before, I suck at blitz like an army of nuclear powered vacuum cleaners.

The openings go pretty well though, I usually get good activity and all the proper development moves seem to come quite instinctively. Clearly playing CC and using databases pay off. - Middlegame, I tend to reach crushing positions, only to throw it all away with horrible blundering. Endgames, well, I suck. But then again, I suck at endgames in CC as well, and get killed by people 200 pts lower, so that's not a big surprise. Actually I've done better than I had imagined in the few endgames I've had so far.

But it's all slowly coming together. After a couple of days blitz feels a lot less chaotic, and I've actually risen 100 points in that time. I even broke 1200 for the first time today. Got all bigheaded of course and took a couple of games against some 1400's. Needless to say, I got my ass swiftly handed back to me. It was clear those guys just had way more mileage under their belt, there really wasn't much of a contest there. But I also drew a 1300, with lots of missed opportunities to win. It's not like they played better chess, they're just making less idiotic blunders. So I'm sure it'll be okay after I've played some 100-200 games and gotten used to the fast pace.

No more 'fear of success' either, and I also think the losses have done wonders to the way I handle defeats.

FICS: br 1193 (1205), sr still 1380p as I haven't played standards this year.