Monday, August 28, 2006

Remarkable endgame from CTS

Today I failed a problem on CTS and as a result spent some time analyzing an endgame. At first I was convinced that my move, while not as crushing as the actual solution, was easily winning. While trying to win the position against a computer program I nearly convinced myself it was a draw until I found a rather neat resource. Since I don't want to forget the position, I'm posting it here where there will be a record of it.

Here is the starting postition, white to play:

Now that your 3 seconds are over, the real solution is 1. Bd6! There are two important motifs here, the pinned rook on e7 and the mating attack if Rxd6.

The move I played was 1. Nh7+. This leads to 1. ... Ke8 (Kg8 Rxe7) 2. Rxe7+ Kxe7 3. Bf6+ Kd7 4. Bxd8 Kxd8 (D)

In my few seconds of thinking time during a CTS problem I was sure that this was winning. However, when I tried to play it out against the computer, I didn't manage more than a draw.

I kept hunting for the right way to play this and eventually found a cute resource for white. One of the main problems for white is that black can try to trade off the kingside pawns and leave white with no mating material. Black tries to bring the king to e6 and play f6 and g5. For example, if White starts out trying to reactivate his knight with 5. Nf6 Black will play 5. ... a5 and now the king is too far away so the knight must chase down the pawn and play might go like this: 6. Nd5 a4 7. Kf3 a3 8. Nc3 Kd7 and black will play f6 next. (Note that it doesn't help white to play 7. f5 trying to get his king over quicker.)

Another way for white to fail to win this position is to leave the knight on h7 to prevent the f6 push: 5. Kf3 a5 6. Ke3 a4 7. Kd4 Ke7 8. Kc4 Ke6 and now the black king penetrates to f5 and takes the f4 pawn.

So far we have learned that (a) if we use the knight to stop the a-pawn, the king will be too slow to prevent the black king from supporting the f6 push and (b) if we use the king to stop the a-pawn, the knight can not be contented to stop the f6 push sitting on h7 because the king will come to f5 and take the f4 pawn. But it turns out the knight can defend the kingside all by itself! So here is the solution:

5. Kf3 a5 6. Ke3 a4 7. Kd4 Ke7 8. Nf6! Ke6 9. Ng4!! a3 10. Kc3 Kf5 11. Nh6+ Kxf4 12. Nxf7 (D)

The white knight now remarkably holds the g-pawn. If the black king tries to drive the white knight away, the knight will simply move to h6 causing a long trip for the black king to get back to the g-pawn. In the meanwhile, the white king is free to mop up the a-pawns before coming back to help on the kingside.

I doubt this endgame study will ever help me win a game, but at least it was entertaining!

Labels: , , ,

Friday, August 25, 2006

Friday progress report

Friday is here again and that means an update on what I've been doing. I continued the same 3 methods I've been doing, Chess Tactics Server, Chess Mentor, and Baburin's Winning Pawn Structures. Without further ado, here's the table:
Week Ending


Chess MentorBook
Aug. 4th1600(1624)-7000-76.3%59Baburin, Chap 1
Aug. 11th1614(1624)-7807-76.8%67Baburin, Chap 2
Aug. 18th1615(1634)-8770-76.3%72Baburin, Chap 3
Aug. 25th1623(1637)-9900-77.6%79Baburin, Chap 4+5

I managed to do 2 chapters in Baburin's book partly because they were shorter and partly because it was the first weekend I had in a month to work on chess. Two more chapters from that book will conclude the first section and then I will put it aside for a while.

At first I planned to to replace the book work from Baburin with Silman's Reassess Your Chess Workbook. However, after reading about Stean's Simple Chess on Blue Devil's blog, I decided to read a few more reviews and everyone raves about this book including more than one person in my rating class who attributes a lot of his recent success to reading and re-reading Stean's book. So, I've planned to read this book before Silman's.

I also have just one more week before I go, as Tempo would say, on holiday. Hopefully I will finish the first section of Baburin's book in that week. While I am away, I will take a book of checkmate problems with me to "stay in shape." It's a decent enough book for that purpose. It has mate in 2 through mate in 6 problems. I am currently on mate in 4. Once I get back from vacation I'll continue with CTS and start on CT-Art and Simple Chess. So now the plan is laid out for the near future and hopefully regular posting here will keep me on track.

Labels: , ,

Tuesday, August 22, 2006

Chess Tactics Server Thoughts

I mentioned that I'd write about this a little while ago and I don't want to let myself down. I've been enjoying the Chess Tactics Server (CTS) and my feeling via casual games I've played recently is that it is helping me see more tactics. The long term effects on my USCF rating are to be determined. When I started out at CTS a couple months ago my CTS rating was not much over 1500. I am now rated a bit over 1600 and so I believe that one can improve CTS performance through CTS study.

What is good at CTS

CTS is for learning quick pattern recognition. The time limits at CTS enforce recognizing the tactical patterns quickly in order to improve one's rating. It is not possible to calculate every possible move and response (computer search algorithm style) in order to determine the correct play under the time limit. Thus it is necessary to recognize the tactical patterns, see the moves and understand their consequences subconciously, and perhaps calculate only narrow forced trees.

When we get a problem wrong at CTS we must consider the source of error, otherwise we are doomed to this failure forever. Shortly after I began at CTS, I was missing or taking a long time on problems that were well within the scope of my chess understanding. In large part they were double attacks with the queen, most often checks. Why was I getting these problems wrong? I simply did not see the queen move when first presented the problem. Upon recognizing this, I made an effort to see all queen checks. Slowly I began to see all the queen checks without effort! This is pattern recognition.

As a second example, I also began to overlook defensive resources (the piece I want to move is pinned or my king is already in check!) for the opponent. Sometimes I would see a fork or skewer or something that appeared to win material and be stunned that my answer was wrong. I had overlooked a defensive resources thwarting my tactical idea. The correct tactic was first removing the defense or something else altogether! Thus I made a conscious effort to look for defensive resources for the opponent. Slowly I began to see defensive resources without much effort at all, and be more confident that there is no resource if I don't see one.

The fast time controls are important at CTS. They enforce the understanding of the solution without effort. Most of the problems that are just beyond my range in 3-10 seconds are still solvable with the right effort. It is important that I continue to put in this effort until they can be solved without it, only then should I move on. Thus, the time control ensures my rating will remain low enough that I will work on these problems until they are solved with less effort.

What CTS is not good for

CTS is not for calculation training. Calculation, as in a real game of chess, requires time. It requires considering many possibilities, some of which will include patterns like the ones learned at CTS. Thus, what is learned at CTS can aid in calucation, but it does not teach one to calculate.

CTS is not a game simulation. Every problem in CTS has a clear tactical resolution. I estimate that perhaps 10% of the time there is a clear tactical resolution during our move in a real chess game. The majority of the time, we are preparing our army to be ready for the tactical resolution. There will almost certainly come a time in our game where there exists a tactical blow, and certainly what we have learned at CTS will help us, but for the majority of the game, this is not as immediate as it is on the problems at CTS.

As always, we must have a balanced approach to learning chess. CTS alone cannot be enough. Studying deep tactical problems that require calcuation is necessary and studying positional ideas is necessary.

Rating and Percent Success

Some people are concerned with rating and percent success. My opinion is that in large part these things do not matter. Your CTS rating is not there for your ego. It exists so that the problem set you try to solve is taylored to your needs. It is not necessary to try to get a high rating, nor is it helpful to your chess playing ability. Your CTS rating will increase when your pattern recognition improves, presumably, so will your overall chess ability.

Of course we should strive to get a problem correct when it is put in front of us. Putting in this effort helps us learn the patterns. But we should not fear getting problems wrong. Often this most stern reprimand is the best way to open our eyes to our faults. Once our eyes are open, then we can work on the repairs.


Friday, August 18, 2006

Progress report

I've decided that Fridays mean progress updates. In addition to just spitting out the numbers this week, I'd also like to organize my future plans a little. To that end I've first organized my current work in a table format:

Week endingCTS
Chess MentorBook

Aug 4



Baburin, Chap 1

Aug 11



Baburin, Chap 2

Aug 18



Baburin, Chap 3

This table conveys what I've accomplished so far. My CTS rating, the number of challenges done in Chess Mentor and the book I'm currently working on, Baburin's "Winning Pawn Structures." The number of Chess Mentor challenges is deceptive because this week more than half of the challenges attempted were repeat challenges, so the number of new challenges didn't increase much.

While I think each of these things is benefitting me, I also think it will soon be time to shake things up a bit.

Sometimes in Chess Mentor I get a problem that I remember from a previous time. While I believe repitition is necessary for learning, I often score well because I remember the moves rather than having learned how to find them. I think I would benefit by putting CM aside and coming back to it later when my memories have faded.

Baburin's book is very well written and I think I learn a lot from his presentation. However, the book is somewhat narrow in focus. Every game presented involves the isolated d-pawn. Once I finish Part One on the advantages of the isolated d-pawn, I will put that book aside and come back to the second part later.

I am going to stick with Chess Tactics Server long term, so I just need to have something to replace CM and Baburin. Since I am always hearing so much about CT-Art, I've decided to try that. I don't know if I will do the 7 circles or not, I think it's best to see how quickly I work through the levels and make a decision then. To replace Baburin I am leaning toward Silman's Reassess your chess workbook. His book helped me improve quite a bit a few years back and working exercises in that style might be just what I need.

I plan to continue my current three methods through the end of August. I have a week vacation in the beginning of September and following that, I will make the switch to CT-Art and RYC-Workbook.

Labels: ,

Tuesday, August 15, 2006

2 positions from a recent game

In the last tournament I played, I lost a game that had some middlegame possibilities. I'm going to post 2 positions here from that game that I think are worth a look. Black to move:

My e-pawn is attacked three times and only defended twice. It cannot be defended again (Bd6 Qxd5+ or Nd7 Qxd5+), so I need to come up with another way to not lose the pawn outright. The game continued:

16. ... c4

17. dxc4 Na5

18. Qa4 Bxa3

19. Qxa3 Nxc4

20. Qb3 e4

There was better play for black (discussed below), but the e-pawn has been secured! White could have played somewhat better in that line, but it is the next move where he gives me an opportunity:

21. Nd2

At this point I returned the favor, making a bad move of my own. One of my first ideas is to play Ne5 and then Nd3, hoping to establish a powerful outpost. Unfortunately, Ne5 loses material to Nxe4. To alleviate the pin on the b2-g8 diagonal I played Kh8. I foresaw Nxc4 dxc4 and thought I was ok there. Unfortunately, I overlooked that Qc2 attacks the e4 pawn a third time and soon it will be lost.

On to the improvements! (Analysis from here out is Fritz assisted.) In the first position, black has a nice tactical possibility. It requires some deep calculation, but I'll admit that during the game, I didn't even really consider the idea:

16. ... Nh5!

Obviously Nxe5 is out due to Bxh4. Black needs a lot of faith to start calculating:

17. Bxe7 Qxe7

Where he's just placed his queen on the file with the doubled rooks!

18. Nxe5 Nf4

19. Re3 Nxe5

20. Rxe5 Qh4

Now black is mounting a strong kingside attack. The last piece can join in quickly with Rd6 and Rg6.

In the second position (after 21. Nd2), black also has a tactical continuation. I didn't want to trade the knights because I thought my Knight was better than his, but I overlooked this possibility:

21. ... Nd2

22. Rxd2 Qf4 (fork!)

23. Bxf6 Qxd2

24. Rd1 Qe2

I fear that even if I had seen this line, the resulting position would not have been clear to me. Fritz evaluates it as clearly winning for black (-1.50), but it looks like murky ground to me, especially in an over-the-board game.

Hopefully I can use this game to help me learn to look for tactical possibilities. The manoever of the knight to f4 in the first position and the queen fork in the second were totally overlooked!

Labels: , ,

Friday, August 11, 2006

Training progress

I'm headed out of town this weekend which means a mini-break from chess. Before I go, I just want to jot down my current training progress.

  • Rating: 1614 (current), 1624 (peak)
  • Tries: 7807
  • Success: 76.8%

Chess Mentor:

  • Challenges with scores: 67
  • Average of initial scores: 61
  • Average of latest scores: 83

Winning Pawn Structures

  • Chapters complete: "White advances d4-d5," "Attack on f7 (f2)"
  • Upcoming: "Rook lift," "Bishop sacrifice," "h-pawn battering ram," etc.

No real thoughts in this post, I just want to make sure I'm keeping a written record of what I've been up to.

Labels: ,

Wednesday, August 09, 2006

Chess Tactics Server oddity

I recently came across something a little odd at CTS. There are two problems there that are nearly identical, number 59505 and 06868 have the same starting position:

In one problem white plays Qc8 and in the other white plays Qc2, in each case defending the knight on f5. The solution to both problems is the same! You can take a minute to look at it, but here comes the answer: Qxg3+ Nxg3 Rg5#.

The problem with Qc2 is rated 1605 while the problem with Qc8 is rated 1679. Why is there a 70 point rating difference between essentially the same problem? One explanation is that on Qc8, the solver may immediately be attracted to Rxc8 and take time to realize that is incorrect before finding the correct solution; while on Qc2 the solver will have nothing to look for but checks on the king.

The less satisfying answer is that the rating system is flawed. While I do believe there are some problems at CTS where the rating system is not accurate, I hope it's not flawed here. The potential flaw is that the rating of a problem depends sensitively on which tacticians have attempted it and when. I have my fingers crossed and hope this isn't the case.

Labels: , ,

Tuesday, August 08, 2006

Tournament report: Broke Even

This week I had 2 wins and 2 losses. Certainly an improvement over my dismal performance the week before. On the other hand, except for one of my losses to a master, everyone was lower rated than me, so the performance was not all that spectacular as the rating change shows: 1685 -> 1683.

It turns out it doesn't take very long to mentally list all the legal moves of your opponents pieces and sure enough I didn't make any simple one move oversights. I figure I'll continue that practice in tournament games until something makes me change my mind.

Not sure when the next tournament will be. Maybe mid to late September or maybe mid October. That leaves lots of time for study/practice.


Friday, August 04, 2006

Tournament tomorrow and current progress

Tomorrow I will play in a one day tournament. Maybe playing two weekends in a row will be good for me. I sure need something to wash away the feeling from last weekend. Though I probably won't play another tournament until October, possibly at the end of September. I'll test out the idea of trying to at least list every legal move. The first two rounds are G/40 and the last two rounds are G/60, so I'm not sure how that will go with the short time controls. I may abondon it if I find myself in a large time deficit.

I thought I should also make a record of what I'm up to, so there is some progress to track. As I mentioned earlier, I do problems at CTS and use ChessMentor for more in depth (longer time, tactical + strategic ideas) problems. So I'll give my current stats for those.

  • Initial rating: ~1530 (Can't recall exactly, but I remember early on being psyched to get to 1550).
  • Current rating: ~1600 (I am now over 1600 all the time with a high of 1624)
  • Problems done: 7000
  • Success rate: 76.3%
Chess Mentor:
  • Challenges with scores: 59
  • Average of initial scores: 61
  • Average of latest scores: 79

Chess Mentor is a little odd because problems (challenges) that have only been done once have the same initial and latest scores. So the number of the latest scores doesn't fully measure the improvement due to repitition.

Additionally, I have started working through "Winning Pawn Structures" by Baburin. The book is mostly on the isolated d-pawn. The first section is on advantages of the isolated d-pawn. I have completed the first chapter, "White advances d4-d5." Following chapters are "Attack on f7", "the Rook lift," "Bishop sacrifices on h6," "h-pawn battering ram," "Queenside activity," and "Play on the e-file." I'll update my progress on these chapters as I go, and then the section on disadvantages of the isolated d-pawn.

The ratings are in from last weekends tournament: 1708 -> 1685. Maybe I can recoup some of those points tomorrow.

Labels: ,