Choosing to Break 2200
By Matan Prilleltensky   
January 15, 2011
Matan325.jpg"Man is unlike beast because man can set priorities!" - me, citing Botvinniik

"He who makes a beast of himself removes the pain of being a man." - Anastasia Bailey, viewing the man-beast relationship differently

"You're both crazy."- Elizabeth Bailey


For me, wanting to break 2200 was not enough. I had wanted it plenty for years. At the 2005 World Open, I came =1st u2300, gaining 50 rating points and taking home $1k. After rattling off four straight master scalps, I felt untouchable: A strong master disguised as expert. But the stubbornly honest rating system disagreed. Despite the powerful lies we tell ourselves, truth is stronger than fiction. Still, after crossing 2100 in late '07, I figured the next step couldn't be far off. But my undergraduate years finished with 2200 no closer than it was two years prior.

It was finally time to revise my thoughts about chess strength. Many of us are dishonest with ourselves when we examine our ratings. Of course we are underrated - unlucky against lower rated players, perhaps, or frequently having opponents stumble onto good moves. Nonsense! There seems to be some confusion about the rating system's purpose. It is not a measure of your ‘chess understanding', future potential, or anything like that. It is a quantitative reflection of your results in tournament chess. As such, it is inherently accurate to a considerable degree. If your four digit number is an unsatisfying one, play better! I promise that will make it go up.

This realization prepared me to approach improvement differently. I lost a frustrating game in February of this year, to the co-leader of a Grand Prix tournament, without putting up a serious fight. Before heading back to Miami, I picked up a copy of "Interview with a Grandmaster", and was struck by Jonathan Rowson's thoughts on personal progress. The Scottish philosopher believes it is unhelpful to want to improve your chess - instead, you have to choose to improve it. During the long ride home with fellow Miamians Ernesto Alvarez and Mohammed Yosef, I made a belated new year's resolution: I was going to learn from every game I played in'10. More importantly, I chose to be a chess master.

Close to a year later, I can say I was true to my choice. It was a broad decision, leading me to prioritize chess study in my daily life. Things I deemed wasteful (parties,facebook, ESPN) were first phased out and then basically eliminated. There is no television set in my apartment - chess aside, my girlfriend wouldn't stand for it. Like everything I did, this type of constant prioritizing is not for everyone. It is simply what works for me. I have "all or nothing" personality traits and find moderate adjustments difficult, nearly impossible, to implement. (When I decided to eat healthier, I cut out all junk food in one swoop and accidentally lost 15 pounds in a week). The way I chose to break 2200 had to take this into account.

Naturally, I also had to change my assumptions about chess development. If nothing else, the old ones were demonstrably not working! In simplest terms, my program of study went from‘openings' to ‘chess'. Conventional wisdom says that learning reams of opening theory is almost never the most efficient way to improve, unless you are trying to go from 2500 to 2700. Yet a significant percentage of class players and experts devote most of their chess time to this stage! What's going on?

Dr. Rowson offers an interesting explanation: We exaggerate the role of the opening in determining the results of our games, because its relationship to study is evident and concrete: He played this move, I knew the theory, etc. It's less intuitive to connect your study of complex middlegames to more incisive attacking play. But this relationship is there, and for those of us without an international title, it's what helps win games! Do a little experiment: Look through your last 20 games and note when you left theory in each. I'm frequently looking for ideas around move 8, and you probably are too! The way we study has to reflect that.

With this in mind, welcome to my improvement journey. All the analysis you will see here was done by hand in a blue workbook I picked up in January. The notes date from February to November, and were not revised (i.e, altered to make me look smarter) in preparation for this article. Unless I say otherwise, nothing is computer checked - that's not how I study chess. The ‘notes to self', expressions of irritation with my play, rambling analyses, and musings about positions I don't understand were all part of the process. The lessons were written for my own benefit, but I believe other players will find them useful. I have selected the games to highlight important steps in my chess improvement - that's why they're almost all wins.

Let's get started! Here is a simple exercise I like. You can do it from any position where you think one side ought to win by force, but you don't immediately see how. An ICC Blitz game of mind reached this position with Black to move:
 
Matan1.jpg


Attacking with every piece

White should be winning. Prove it! The point isn't to get everything 100%. Really exploring the position (Move the pieces! Write your analysis!) will improve your ability to find candidate moves and coordinate attacking pieces.

Prilleltensky - Anonymous [A69]

ICC Blitz, 04.12.2010
19...Qa4
19...Qb5 20.a4 Qe5 21.Nf6+ Kg7 (21...Kh8 22.Nxe8 Qxg5 23.Rf8#) 22.Nxe8+Kg8 23.Bh6 is another brutal finish.; 19...Rf8 20.Rxf8+ Kxf8 21.Bh6+ Ke8 (21...Kf7 22.Rf1+) 22.Nf6+ wins the queen.
20.b3 Qb5
20...Qc6 21.Bf3 and White will have a decisive material advantage
21.a4 Qxb3 22.Nf6+ Kf7
For some reason, it took me ages to crack this position; I was fixated on discovered checks and forcing sequences. The right way is instructive: 22...Kf8 23.Bh6+ Kf7 24.Nd5+ Kg8 25.Ne7+Kh8 (25...Rxe7 26.Rf8#) 26.Rf7 Rg8 (26...Rxe7 27.Rf8#) 27.Bg7+Rxg7 28.Rf8+ Rg8 29.Rxg8#
23.Rac1
Brings the last piece into the attack (CIS students, are you reading this?) with the deadly threat of Rc7+. Black has no response.  
23...Rd8 24.Rc7+ Rd7
24...Kf8 25.Bh6#
25.Nxd7+ Kg8 26.Nf6+ Kh8 27.Rxh7#
Matan2.jpg

Along with a steady diet of Alekhine, exercises like this helped me internalize the value of attacking with all the pieces.
 
You can do the same thing while studying a GM game. A variation of Kotov-Keres reached the following position, with White to move:

 KotovKEres.jpg

The annotator, GM Yakovich, took the win for granted, but I didn't see it right away. Time to take out the notebook! I decided that 1.Nxe6! wins instantly. The idea is that the knight threatens mate, cannot be taken, and avoids perpetual/trades by covering both d4 and g5.

Suppose the position is made more difficult by sticking the black bishop on d3. How does White win?

1.-- Bd3 2.Bxd3 Qc3+ 3.Kf2 Qxd4+ 4.Kg3 and Black seems defenseless against Nxe6.  4...Rg8 (4...h5 5.Nxe6 h4+ 6.Kh3 Rg8 (6...fxe6 7.Qxg6+ Kh8 8.Qh7#) 7.Qxf7+ Kh68.Bg5+ Kh5 9.Qh7#) 5.Qxf7+ Rg7 6.Bxg6+ Kh8 7.Bf6 Rag8 8.Bxg7+ Rxg7 9.Qf8+ Rg8 10.Qxh6#]

I'd like to emphasize how important I found the process of looking for ideas. By trying to come to your own conclusions, your understanding of GM games will be deepened, even if you finish with more questions than you started with! Take this position, reached in a variation from Alekhine-Samisch:

According to Alekhine, White wins with Nxf8 (next move) followed by Rd8. I had some trouble seeing this and tried to demonstrate it.

Alekhine - Samisch
Matan4.jpg
1...g6 2.Nxf8 Nxf8 3.Rd8 Kg7
Planning ...Ra8 and ...Bb7, unpinning.
4.b4
4.Re8 Ra8 5.Rd1 Bb7 is also not clear.; 4.h4 Ra8 5.h5 gxh5 (5...Bb7 6.Re8 (6.Rxa8 Bxa8 Doesn't look like anything big for White.) 6...Bxe4 7.h6+ Kxh6 8.Re1 Bc6 9.R1xe5 Kg7 10.R5e7 Almost zugzwang maybe? 10...h5 11.Rc7 Traps the bishop and wins. Alas, this variation was quite cooperative; Black can do better.) 6.Rf3 Bb7 defends.
4...Ra8 5.c5 bxc5 6.bxc5 Bb7 7.Re8
This seemed like a critical position to me.
7...Bxe4 8.Rc1 Ra7
8...Kxf7 9.Rxa8 Bxa8 10.c6 wins.
9.c6 Rc7
9...Kxf7 10.c7 Rxc7 11.Rxf8+ Kxf8 12.Rxc7 also wins.
10.Rxe5 Bf5 11.Ra5 Kxf7 12.Rxa6 Ne6
12...Ke7 13.Ra8
13.Rb6 Ke7 14.a4
Matan5.jpg
Seems to give White excellent winning chances.

You're right, these lines are not computer checked! But they helped me construct some of the important ideas in the position.

Signs of Progress

After a couple months of doing this type of analysis, mainly with my losses, I was still hanging around the low 2100s. But I felt that I was on the verge of playing stronger moves - if nothing else, I started to see more ‘stuff' during my games. In a rated training match with then-IS 318 star Alexis Paredes, I began playing a bit better.

Prilleltensky - Paredes [A83]

5240


1.d4 f5 2.e4 fxe4 3.Nc3 Nf6 4.Bg5 Nc6 5.d5 Ne5
ParedesNe5.jpg
This was my third time playing the Staunton Gambit, and second time facing this response. Now the main line involves Qd4 followed by taking on f6 and e4, but I preferred to play a pawn sacrifice I found (of course, I did not introduce it) OTB in a prior g/30 game.
6.f3 exf3
In unknown territory, my opponent grabs the pawn.
7.Nxf3 Nf7 8.Bd3
This worked well in the game; by leaving the bishop on g5, I was able to capture on f6 only when Black needed to take back with the g-pawn, leaving his king permanently unsafe. Around this time, I was becoming more aware of the benefits of keeping the tension in certain positions.
8...c6 9.Qe2 cxd5
9...Nxd5 Tries to side-step White's plan, so I devoted some time to analyzing it after the game. 10.Nxd5 cxd5 11.0-0 d6 12.Rae1 Qb6+ 13.Kh1 e5 14.Bb5+ Bd7 This line is certainly not forced, but it looked plausible enough to me. Here I wanted to see if there was a forced win for White. 15.Nxe5 (15.Bxd7+ Kxd7 should be looked at.) 15...Nxe5 (15...Bxb5 16.Nc4+ Ne5 17.Qg4 Qc7 18.Qe6+ Be7 19.Bxe7 Qxe7 20.Nxd6+ Kd821.Qxe7+ Kxe7 22.Nxb5 Ke6 23.Nc7+ wins.; 15...Qxb5 16.Nxf7+ Qxe217.Rxe2+ Be6 18.Nxh8 wins.) 16.Qf3 (16.Qxe5+ dxe5 17.Rxe5+ Be7 18.Rxe7+ Kd8 might not be more than a draw.) 16...Be7 17.Rxe5 dxe5 (17...Bxb5 18.Rxe7+ is crushing.; 17...0-0-0 18.Bxd7+ Rxd7 19.Rxe7 wins. Black needs to deviate earlier.) 18.Qf7+ Kd8 19.Qxe7+ Kc7 20.Qxd7+ Kb8 21.Qxg7 Rc8 22.Qxe5+ Qc7 23.Bf4 wins. Based on these variations, I concluded that the line beginning with 15. Nxe5 is winning for White by force.
10.Bxf6gxf6 11.0-0-0 e5 12.Nxd5 d6 13.Bb5+
13bb5.jpg
Black missed how strong this check is. Things are now very difficult for him.
13...Bd7 14.Bxd7+ Kxd7
Here I thought for far too long, looking for a strong continuation of the attack. I was getting nowhere with queen checks, until my attention came to the last undeveloped piece.
15.Nh4 Rg8 16.Rhf1
Again, the last rook comes into play with huge effect!
16...Be7 17.Nf5 Bf8 18.Qh5 Ke8 19.Nxf6+
FinalParedes.jpg
Of course, this game was not a real reflection of Alexis' strength. Nonetheless, it gave me plenty of confidence heading into the Hartford Open.   1-0

Breaking the Drought: Consecutive Master Scalps

The following week, at the Hartford Open, I beat a master in classical chess for the first time in over a year.

Aaron - Prilleltensky [B33]

5241

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e56.Ndb5 d6 7.Nd5 Nxd5 8.exd5 Nb8 9.c4 Be7 10.Be2 a6 11.Nc3 0-0 12.0-0 f5 13.Kh1Nd7 14.f4 Bf6 15.Qc2 exf4 16.Bxf4 Ne5 17.Rac1
17Rac1Aaron.jpg
17.b4 Ng6 18.Bg3 f4 19.Bf2 a5 is given by Kolev and Nedev, "with good attacking prospects for Black".
17...Bd7
Right after playing 17...Bd7, I saw this: 17...Ng6 18.Bg3 (18.Bd2 Be5) 18...f4 and wished I had played it.
18.Nd1
I think Deepak avoided the immediate b2-b4 because he was worried about his c-pawn. Still, I think it was too early to commit this piece to d1, and it struggled to find an effective role in the game. 18.b4 a519.a3 would be my preference.
18...Qe7
Preparing a possible ...Ng6, preventing Nd1-e3 (not necessarily White's intention of course), and eyeing control of the only open file.
19.Nf2 Rae8
Black seems to be taking the initiative, while White still hasn't played b2-b4. This suggests his knight moves were too slow.
20.Bd3 g5 21.Bd2 Nxd3 22.Nxd3
This looked quite awkward to me. 22.Qxd3 b5 looks strong and is likely what put my opponent off this recapture.
22...f4
This was the idea behind ...g5 and ...Nxd3. The c8-h3 diagonal is opened for the light squared bishop, and ...Bf5 is a possibility.
23.Rce1 Qd8 24.Nf2 Be5
Intending ...g4. I also noticed in some of Gelfand's wins that the Israeli GM sometimes 'mysteriously' avoids rook trades when planning an attack on the king, even when it is not readily apparent how the rooks will be involved.
25.Qb3 b5 26.Qa3
I expected White to go pawn-grabbing on a6, taking his queen far from the kingside. That prompted me to begin going forward.
26...g4 27.Nd3
27Nd3Aaron.jpg
Strong move, preparing to exchange the important e5 bishop. This move also has a deeper purpose I did not see. After drifting into a difficult position, Deepak Aaron resists very tenaciously. 
27...bxc4 28.Nxe5 dxe5 29.Qd6
As we both approached serious time pressure, this move was an unpleasant surprise. It prevents ...Qh4 due to Qxd7 ...g3 Qh3. 
29...Rf6
The point of White's play: This is the only way to eject the queen from my position, but now the black queen's route to h4 is obstructed.
30.Qc5 Rh6 31.g3
Another surprise, more disturbing than the last. Now I won't be able to play ...g3 myself, and my king suddenly looks a bit exposed. I need to be more attentive to the defensive resources of my opponents. Time pressure was partially responsible for this, but White had less time than I did. When my opponent drifts into time pressure, I am overly likely to do the same. It makes no sense to follow the other guy's time management! My time management has improved, but it must get much better. One idea is striving to reach move 40 with 20 minutes to spare. For now, don't slow down when the opponent does.
31...Bb5
We were both well under a minute to reach move 40. Great.
32.Rg1
32.a4 c3 33.axb5 cxd2 34.Rd1 e4 35.Rxf4 (35.Rxd2is possible; no idea what is happening.) 35...e3 36.Rxg4+ Rg637.Rxg6+ hxg6 38.Qc6 e2 39.Qxg6+ Kf8 40.Qf5+ Kg7 41.Qg4+ Kh6 42.Qf4+ (42.Qh3+Kg6 43.Qg4+ Qg5 wins) 42...Qg5 43.Qd6+ Qg6 44.Qf4+ Kh7 45.Qc7+ Kg8 escapes the checks and wins. The fact that neither player had time to consider these variations highlights the randomness of time scrambles - and the importance of avoiding them!
32...Qd7
Guarding g4 and threatening ...Rc8, particularly dangerous now that Rg1 took an escape square from the white king. Now I think White is in trouble, so a2-a4 and 35.Rxd2 may have been necessary.
33.gxf4 Rc8
Black is now totally winning. I was guilty of 'chalking it up' at this point, thinking the game was almost over. This left me psychologically unprepared for the run up to move 40, when White will of course fight desperately.
34.Rxg4+
Otherwise White loses at least the exchange for nothing. This was a practical gamble that nearly turned the tables.
34...Qxg4 35.Qf2
My next move was total panic.
35...Qh4
Crazy. Now the h6-rook will be unable to block checks on the g-file, and the battle rages on! 35...Qf5 And White is just down a rook. If I had bothered to calculate after playing ...Rc8, maybe this would have been obvious!
36.Rg1+
Ah. Not only is my opponent not about to resign, he is almost checkmating me.
36...Kh8 37.Bc3 Re8 38.Qxh4 Rxh4
During the game, I thought White missed a chance here.
39.fxe5
Another very good practical choice, forgoing the win of the exchange and resulting lost ending. Alas, 39.Bxe5+ Rxe5 40.fxe5 Rd4. Shows he did not miss anything, since this wins easily for Black.
39...h5 40.e6+
With one second to spare.
40...Kh7
12 seconds left, although it didn't feel that way. It looks like Black is still winning. However, White has many tricks based on the inconvenient positions of my king and ...R/h4. 
41.Rg7+ Kh6 42.Ra7
Here I spent most of the sudden death time control, going into the tank for 40 minutes. Fortunately, I think I found a good solution. I looked, in many forms, at two ideas: attacking the white king, and stopping/winning white's pawns. Finally, I saw the two ideas should be connected:
42...Rg4
The threats to white's king require attention. But then this rook gains a crucial tempo to drop back to g8. Then black's rooks will be coordinated to stop white's passed pawns.
43.h3 Rgg8 44.Kh2 Rd8 45.a4 Be8 46.Rxa6 Rxd5 47.a5 Kg6 48.Ra8 Rg5 49.Rc8 Kf5
Again combining threats to the white king with stopping the passed pawns.
50.Bd4 Kxe6 51.a6 Bb5 52.Rc7 Bxa6 53.Rh7 Kd5 54.Ba7 R8g7
FinalAaron.jpg
And here White resigned. An intense struggle. Lessons:  - When attacking, look critically for the opponent's defensive resources! Remember he is desperatelytrying to beat you.- An attack on the king is not the only way to exploit an advantage; here the complications unleashed were difficult to evaluate.- don't play too slowly just because the opponent does! 0-1

Without leaving Hartford, Miami provides the background for this tournament's next game. NM Ernesto Alvarez (from the longcar ride earlier) wins an improbable number of games from positions where he is losing, lost, or completely busted. When I asked him why, he frowned. "I always know it's coming. I'm always saying it: He's going to mess up, he's going to mess up. You just have to be there."

Prilleltensky - Daftani [D43]

5242

1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.Nc3 e6 5.Bg5 h6 6.Bh4 dxc4 7.e4 g5 8.Bg3
Ok, first Moscow Gambit! I was fairly ignorant as to the theory.
8...b5 9.h4 g4 10.Ne5 Bb7
10Bb7Daft.jpg
Now I wasted some time looking at d4-d5, since I thought this might be the position where it was the right move (?!). I should just accept the impossibility of remembering (or knowing!) everything and be 100% fine with the moves I think are best OTB.
11.Be2
My comment about the alternative notwithstanding, this move is perfectly fine. But I may not have played it for the right reasons. Here, I instinctively wanted to play 11.Nxg4 a type of sacrifice I have seen Levon Aronian use to good effect. It's the type of move I like to play and has objective merit, so why didn't I do it? I think I was worried itmight not be "theory". Never mind that it is; if I think it's the best move, it's the move I should play!
11...Nbd7 12.Nxd7 Qxd7 13.Be5 Qe7 14.Bxg4
14.0-0 Looks like a more energetic approach, well used by Grischuk. White intends to play the position as a gambit.
14...Rg8 15.Bf3
I wanted to play 15.Bh3 But was needlessly hung up over its appearance. This is nonsense - chessplayers should have confidence in their ideas!
15...Nd7 16.Bg3 0-0-0
My opponent told me after the game he played this with his next move in mind. I had glanced at it but completely failed to appreciate its strength.
17.h5
17h5.jpg
Planning Bg3-h4 in some situations, but this is far too slow. This is an example of what Kasparov calls a "half-move": something plausible that does not answer the demands of the position. In hindsight, it is obvious something more direct is required. Furthermore, this completely misses Black's idea! It gives him a natural followup to ...Rxg3 by giving his queen the g5 square. Looking for black's forcing moves when considering h5 probably would have made this clear! "...Rxg3 fxg3 ok, now what are black's forcing moves? ...Qg5 looks strong." 17.0-0 Looks logical. The exchange sacrifice is still possible: 17...Rxg3 18.fxg3 Ok, what is black getting for the exchange? He eliminates a potentially dangerous attacking piece, weakens white's kingside structure, and may someday use e5 as an outpost. White's position nevertheless looks fine here. In the game. By contrast, the exchange sacrifice's power was multiplied by h4-h5.
17...Rxg3!
Extremely strong. My first thought was that my opponent had a rush of blood. Moments later I realized it wasn't so, and after two more moves Black was winning.
18.fxg3 Qg5 19.Qc2
Since my position became lost shortly after, this is an important moment to seek alternatives. [19.Qd2 Qxd2+ (19...Qxg3+ 20.Qf2 (20.Kf1followed by 21. Ne2 is an idea) 20...Qxf2+ 21.Kxf2 e5 22.dxe5 Nxe523.Rhd1 Initially looked like an improved version of 19...Qxd2+ for black, due to white's awkward king position. But white is compensated with faster access to the d-file. I have positional issues, but black's own structure is hardly perfect. So right now I think 19. Qd2! would have been a major improvement on the game and left white with a reasonably playable position.) 20.Kxd2 e521.dxe5 (21.d5 b4 22.Na4 (22.Ne2 cxd5 23.exd5 f5 looks very good for Black.) 22...f5 23.Ke3 fxe4 24.Bxe4 cxd5 25.Bf5 looks difficult forwhite.) 21...Nxe5+ 22.Kc2 Does not look terrible at all; I think this was thebest defense. 22...Bc5 23.Rad1 Rg8 24.Ne2 Shows White doing fine after some plausible moves. When looking at these lines during the game, I paid too much attention to the strong black knight on e5 and bad white bishop on f3, without sufficiently considering the position as a whole.
19...Qe3+
19...Qe3Daft.jpg
I had been planning to answer this with Ne2, but then saw what looks like a forced win for Black. So, I had to lose the critical d4 pawn and give black's knight the e5 square. Black is winning now. Looking critically at forcing moves down the line before playing Qc2 might have enabled me to spot this further in advance. I need to start calculating in a more structured way, stopping at each branch of each variation to look for forcing moves. Having resigned prematurely in round one of this tournament, I had resolvedto put up maximum resistance in bad or lost positions from then on. Now I remembered Ernesto's wise advice. Chess is played between humans, and tenacious resistance can save many half and full points. So, I gritted my teeth and prepared to fight like an animal. 
20.Qe2
20.Ne2 Ne5 21.dxe5 (21.Qd2 Bb4) 21...Bc5 22.Rf1 Bb4+ winning.
20...Qxd421.Rd1 Qe5 22.Rh3 Bb4 23.Qd2
Here I picked up the pace, since I had around 20 minutes to reach time control. Having a bad/losing poisition doesn't mean you should play slowly: opportunities will present themselves, and you are more likely to take them if you have enough time to think! 
23...Qf6 24.Rh4 Ba5 25.Qf4 Bxc3+ 26.bxc3 Qxc3+27.Kf1 Ne5 28.Rxd8+ Kxd8
Black has won a pawn, and intends to win the game by pushing the passer on c4. Makes sense. . .
29.Kg1 Qd4+ 30.Kh2 c3 31.Qxh6
I spotted a diabolical trap and played this quickly, hoping to avoid arousing any suspicion.
31...Nxf3+ 32.gxf3 Qd2+?
32...c2 It took me two pages of analysis to decide this move probably wins, but I'll spare you the details! Of course you can turn on your silicon analyst and find out instantly.
33.Qxd2+ cxd2 34.h6
34h6Daft.jpg
Forget wings, that pawn has a jet-pack!
34...d1Q 35.h7 Qe2+ 36.Kh3
The king is waltzing away from the checks.
36...Qf1+
36...Qxf3 demands attention. However, 37.h8Q+ Ke7 38.Qe5 Qf1+ 39.Kg4 Qe2+ 40.Kg5 Qe3+ 41.Kh5 (41.Qf4 Qxf4+ 42.gxf4 c5 43.f5exf5 44.Kxf5 c4 45.Ke5 Is a complex ending I don't think Black should lose.41...Qf3+ 42.Kh6 Qe3+ 43.Kg7 is winning for white. As in the game, if Black cannot advantageously trade queens, white coordinates his queen and rook with decisive effect. The king can safely run up the board in so many lines because black's bishop plays almost no part in the game.
37.Kg4
My opponent seemed shell-shocked and was either moving quickly out of inertia, or trying to exploit my time pressure. But automatically driving my king forward wasn't especially helpful:
37...f5+
I think Black is probably losing by now. Nonetheless, it's necessary to be aware of the trade off this move presents: The king is driven to a location where it can be checked, but the seventh rank is opened for White's heavy pieces. The latter proved to be an important factor.
38.Kg5 Qxf3 39.h8Q+ Kc7 40.Rh7+ Kb6 41.Qd8+ Kc5?
As often happens after a massive turnaround, the player experiencing the negative trend is unable to mobilize his resources. [41...Ka6 Is what I was expecting, and would have been far more challenging.  42.Qc7 a) 42.Qe7 Qxg3+ 43.Kf6 Qb8 44.exf5 exf5 and I don't see anything convincing.; b) 42.Qb8Qe3+ 43.Kg6 (43.Kh5 Has the idea of leaving the 6th rank for the rook.  43...Qb6 44.exf5 exf5 45.a3 Is this elusive zugzwang? 45...b4 Ok, it probably doesn't exist. This whole analysis skirts a fine line between usefulness and insanity.

43...Qb6 Here I became obsessed with trying to find a win based on zugzwang themes.  44.exf5 exf5 45.Rd7 b4 (45...f4 46.gxf4 Qg1+ 47.Kf7 Qb6 48.f5 c5 49.Rd6 Bc6 50.Qc8+ winning.) 46.Kxf5 Qb5+47.Kf6 Qf1+ 48.Ke7 Qe2+ 49.Kf8 Qf2+ 50.Rf7 Qc5+ 51.Ke8 Qe3+ 52.Re7 Qb6 53.g4 c5is annoying.; 42...Qe3+ 43.Kg6 Qb6 (43...Ba8 44.Qc8+ Ka5 45.Qxa8 Qxg3+46.Kf6 Qc3+ 47.e5 Winning for white) 44.Qxb6+ Kxb6 45.exf5 exf546.Kxf5 c5 47.g4 Bd5 48.Rh2 b4 49.g5 a5 50.Ke5 Bg8 51.g6 a4 52.g7 b3 (52...a3! "Seems to hold the draw, or even turn the tables!" According to my notes from a while ago. I'd like to say this 'conclusion' was reached after a cursory glance, because switching on Fritz tells me 53. Kd6 is winning for White. Evidently I need to hold my analysis to higher standards. . .) 53.a3 Kb5 54.Rh8 Bf7 55.Rf8 Bc4 56.Rb8+ Kc6 57.g8Q Bxg8 58.Rxg8 Kb5 59.Kd5 winning. Finally! There's something fascinating about searching for the truth, even when you know it is unlikely to be approached. 
42.Qe7+ Kd4 43.Qb4+ Ke5 44.Qc5+ Kxe4 45.Rh4+ Kd3 46.Qa3+

Lessons: - fight fight fight as hard as possible in bad/losing positions- be confident in your ideas and play the moves you think are best; don't worry about whether they are "theory"!-systematically examine forcing moves in every variation you calculate (this means all the branches!) 1-0

Fundamentals and Endgames
Attacking games like these helped push me off my low 2100s plateau, but my play lacked 2200-type consistency. Daaim Shabazz, whose interesting discussion of chess mastery emphasizes experts of African origin, likes to call relying on tactical trickery "fast food chess". (http://www.thechessdrum.net/65thSquare/65_marapr04.html) To break my temporary ceiling, I needed firmer grounding in chess fundamentals. Studying Mikhail Botvinnik, the patriarch of Soviet chess, continues to push me in this direction. His games have helped me broaden my middlegame repertoire and (sometimes!) win without throwing pieces at the enemy king. For example, this is the first game I can recall where my winning plan involved aiming for a certain ending.

Mooney - Prilleltensky [E63]

5243

1.d4 Nf6 2.Nf3 g6 3.g3 Bg7 4.Bg2 0-0 5.c4 d6 6.Nc3 Nc6 7.0-0 a6 8.b3 Rb8 9.Nd5 b5
Weak. "What's the threat" thinking would have averted walking into White's idea.
10.Nxf6+ Bxf6
10..Bxf6Mooney.jpg
As often happens, mistakes come in pairs. It's strange I didn't see/notice white's reply to my next move; maybe I so strongly associated b3 with a queenside fianchetto that I forgot it was by no means forced! I also didn't like Black's position after the alternative recapture, so I took with the bishop without too much thought. Obviously this was misguided;one move leading to a mediocre position doesn't mean the alternative won't be terrible!
11.Bh6
Naturally I saw this right after making my move. White gets a tempo for development and stops my bishop from dropping back to the normal g7 square without an undesirable exchange, weakening the kingside dark squares and leaving Black with a depressingly static position.
11...Re8 12.Rc1 Bd7 13.d5 Ne5 14.Nxe5 Bxe5 15.c5
This good move surprised me. The positional threat of c5-c6 compels me to take the pawn, after which White has a solid plus. 15.f4 Bf6 16.e4 c5 Actually looks fine for Black, so the velocity-first approach is not called for.
15...dxc5 16.Rxc5 b4
16b4Mooney.jpg
I think this move was good; I'm trying to get some prospects for my pieces, and ...Bb5 is now a possibility in some lines. There is also the long-term idea of playing against the a2/b3 pawns, which are now fixed. This proved critical later on.
17.Bf4?
Another move that surprised me, but this one is a mistake, giving away most (all?) of the advantage.
17...Bxf4 18.gxf4 e5
Aiming to free my position by latching on to the newly created f4 pawn.
19.dxe6 Bxe6 20.e3
I was half expecting 20.Qc2 during the game, and wasn't sure what I would play. 20...Qh4 21.e3 Bh3 22.Rg5 Rb5 23.Rxb5 Bxg2 24.Kxg2 Qg4+ 25.Kh1 Qf3+ with perpetual suggests White's decision to enter an ending was objectively good. Rather than go for this line, I would have had to find some way to keep the game going.
20...Qxd1 21.Rxd1 Rbd8
21...Red8 looks more natural, if the variations work: 22.Rd4 Rxd4 23.exd4 Rd8 24.d5 Bf5 25.Rxc7 Bb1 26.Ra7 Bxa2 27.Rxa6 Bxb3 28.d6 Be6 is good for Black. So I should have played 21...Red8, not fearing 22.Rd4. I did not see the ...Bf5-b1 maneuver during the game; it is instructive how the b4 pawn makes this possible.
22.Rcc1 Kf8
I spent an unreasonably long time looking at 22...c5, although I rather quickly decided it didn't work and didn't have new ideas in mind during the rest of my thinking time. Therefore 22...Kf8 should have been played much faster. I was nonetheless happy to have played what I believe to be the strongest move, rather than the pseudoactive pawn push I considered.
23.Bc6 Rxd1+ 24.Rxd1 Rb8 25.f3 Ke7
Ke7Mooney.jpg
Here I started to get optimistic, and I believe it was justified. My queenside majority can create a passed pawn, while White's kingside majority cannot. 
26.Kf2 Rd8
26...Rb6 27.Ba4 c5 28.Rc1 Kd6 29.Rd1+ Ke7 30.Rc1 would force a repetition draw. The way to advance on the queenside is to trade rooks.
27.Rxd8 Kxd8 28.e4 Ke7 29.Ke3 Kd6
Note to self: It would be good to have more time left to reach move 40 here, since the game is about to be decided!
30.Bb7 a5 31.Kd3
Here the white king ended up allowing its counterpart into c5 and being hit with ...Bc4+. 31.Kd4 c5+ 32.Kd3 a4 33.Kc2 Idea: The difference between this and the forms of passive defense in other variations is that here the c5-pawn obstructs the black king's route c5-d4-e3 into white's kingside. 33...axb3+ 34.axb3 c4 35.bxc4 Bxc4 36.Bc8 Be2 (36...Kc5 37.f5) 37.Bb7 Kc5 38.e5 Bc4 39.Be4 Bd5 40.Bxd5 Kxd5 41.Kb3 Kd4 42.Kxb4 Ke3 43.Kc5 Kxf3 44.Kd5 Kxf4 45.e6 fxe6+ 46.Kxe6 Kg4 47.Kf6 Kh3 48.Kg7 Kxh2 49.Kxh7g5 wins. Based on this line, it looks like 31. Kd4 would not have changed the result.
31...a4
31...Kc5 first looks more accurate, since ...a4 in any case cannot be prevented.
32.bxa4 Bxa2 33.a5 Kc5 34.a6 Bc4+ 35.Kc2
This definitely simplifies Black's task. 35.Ke3 looks more testing. 35...Kb6 36.Kd4 The point: Now there isn't time for ...c6. 36...b3 37.Kc3 c6 38.Bc8 Bxa6 39.Bxa6 (39.Bd7 Bc4) 39...Kxa6 40.Kxb3 Kb5 With what I eventually decided was a winning pawn ending for Black. It looks like 35. Ke3 (instead of 35. Kc2) would also not have changed he result.White's losing chances (and eventual loss) appear to stem from his acquiescence in trading off the rooks; by keeping them on the board, White would have had no problems whatsoever.
35...Kb6 36.e5
Hoping for ...Bxa6 Bd5; after the game, my opponent indicated he missed Black's next move.
36...c6
Now the a-pawn falls with a winning ending. The game ended just before time control:
37.f5 gxf5 38.Bc8 f4 39.Bf5 h6
Lessons:- one line looking unpromising doesn't mean you can just play the alternative; maybe the alternative is much worse!- If you know a move is not strong, stop analyzing it and look at a move you might actually decide to play- Trading certain pieces, entering an ending with a queenside majority: These plans, as implemented here, need to feature more in my games going forward 0-1

Losing a difficult ending to SM Yaacov Norowitz shortly beforehand also helped me improve my play in the ending. Analyzing with Yaacov is a fascinating experience; I got the impression he could grind out ‘simple' positions against masters in his sleep. Later on, I analyzed the game myself and came away with some interesting insights.

Prilleltensky - Norowitz [A51]

5244

1.d4 Nf6 2.c4 e5 3.dxe5 Ne4 4.Nd2 Bb4 5.Ngf3 Nc66.a3 Nxd2 7.Bxd2 Bxd2+ 8.Qxd2 0-0 9.e3 Qe7 10.Qc3 Re8 11.Be2 Nxe5 12.Qxe5 Qxe513.Nxe5 Rxe5 14.0-0 d6
"There's a full game ahead", Yaacov mused in the postmortem.
15.Rfd1 Bd7 16.Rac1
16.Bf3 Seems to improve.
16...Bc6 17.b4 b6 18.Rc3
Preparing f2-f4, although this plan may be overly ambitious. Here is how I was thinking about the game: I have a 4-3 kingside majority. Black has a 4-3 queenside majority. Majorities are meant to be advanced. Therefore I should play for f2-f4, g4-g5, etc. But this might just be misguided! How is either of us going to create a passed pawn? The pawn structure being *very slightly* asymmetrical does not automatically make the game a "battle of majorities" or something like that! I notice that in Endgame Strategy's chapter on majorities, there are no examples of "majorities" like mine. Also, since advancing pawns creates weaknesses, if white wants to advance on the kingside then first it should be prepared (i.e, put the rooks behind the pawns). 

When I asked Yaacov if he was intentionally provoking me to get aggressive, he smiled and said "put it this way. I wasn't provoking you to do it, but I knew you would do it." 18.Bf3 Bxf3 19.gxf3 Leaves things quite equal, according to my opponent. To me, this looked like a concession (I have doubled pawns), but FM Steve Stoyko agreed with Yaacov's belief it was irrelevant. My belief that trading the bishops would be a concession partially reflected the misguided "battle of majorities" idea. It also may be related to something like reverse egoism! Rowson describes egoism in chess as being so focused on your own ideas that you ignore those of the opponent. Here, when thinking (barely) about the possibility of the bishop trade, I disproportionately worried about the isolated h-pawn. But this pawn is not everything! If a black rook is over onthe h-file, Rd1-d5 will be an idea; if black then trades rooks, maybe c7 becomes a weakness. c4-c5 may also be dangerous at some point. The position is more than just the weak pawn on h2!
18...Kf8
A good improving move, playing the king closer to the center.
19.f4
Ambitious, though not necessarily bad. But without this, my opponent indicated there is no way for either side to create winning chances without undue risk. When I asked what White should do, the response came, "wait!" Interesting. 
19...Ree8 20.Kf2 a5 21.b5
Now the position is a bit more imbalanced, in that I have a backwards pawn on c4. Yaacov also pointed out an imbalance resulting from b4-b5 that can favor White in the long term: Some king and pawn endings may be difficult for the second player if White can walk his king to d5. I still don't totally understand this; he may have made the point in conjunction with a line where the bishops were traded, when it's a bit clearer. Still, what just happened favors Black: His queenside pawns are now all on dark squares, while my c4/b5 pawns are fixed on light. I think it might have been best to sit tight with 21.Rd2; although black can open the a-file, it's not clear how he makes progress.
21...Bd7 22.g4 f6
I didn't appreciate this move at the time. In addition to restraining g4-g5, Black also has ideas of playing ...g5 himself (as happened in the game!) It is easy to imagine a scenario where, after a couple pawn trades on g5, white's g4 pawn is fixed on a light square. Yaacov had ideas like this in mind when he said that "this is actually a very sharp position". The point is that superficially simple-looking situations have many hidden ideas and subtleties. By disturbing the equilibrium, I have given Black a very clear improvement plan when earlier it was very difficult to see one.
23.Bf3 Rad8 24.h4 Ke7
Critical moment: Black is preparing ...g5. Having gotten his king out of the way, he will be able to play ...Rg8 at the right moment to recapture the g-pawn, after which g4 will be isolated on a light square. Having started the plan of kingside expansion, it should be consistently pursued with 25. g5! with a perfectly ok position. So, it seems that the real mistake was not beginning a kingside advance: It was starting but failing to implement it consistently. Leaving the pawn on g4 just left it vulnerable to the ...g5 plan. I suspect I would have realized this against a weaker opponent, so what accounts for the sudden timidity? I'm not sure I really believed I could win against this opponent, especially in an endgame. This is a self-defeating consciousness; nobody is unbeatable in any stage of the game, not even close! 
25.Rd4
Vague. Now Black gets in his pawn break and takes the initiative.
25...g5!
25...g5.jpg
In time pressure I defended weakly, and Black won with ease:
26.hxg5 fxg5 27.Kg3 Kf6 28.Rd5 Rg8 29.fxg5+
Why? White should keep the tension.
29...Rxg5 30.Rxg5 Kxg5 31.Rd3 Re8 32.Rd5+ Re5
Now things are very bad.
33.Rd3 Be6 34.Rc3 Rc5 35.Be4 h6
finalyac.jpg
and Black went on to convert his advantage. Lessons:- This was not inevitable! g4-g5 would have been completely fine for White. Beginning to advance and then stopping was a fundamental strategic error, since the pawns were just left as targets. Playing g4-g5! would have gained space and stopped Black's ...g5 idea.- Always know you can and must compete for ideas with any opponent. Titled players must be beaten, and beating them starts with letting yourself persistently search for good moves. 0-1

Matan250.jpg
Photo Betsy Dynako Atlantic City 2010
The Final Push

A good result in the Atlantic Open in DC put me within touching distance of my goal, but there was a missing ingredient: Improved play in g/30 events. Enter semi-retired International Master Alan Stein. As a strong chessplayer interested in pedagogy, Alan generously offered to spend time analyzing positions with me and assisting my improvement process. Looking at tactical situations with him has been a highly instructive experience. It has nudged me away from "this has to win" obsessiveness toward broader thinking in the search for ideas. Our work has made me keenly aware of my tendency to spend my thinking time OTB hitting the same nail with the same hammer, trying to force the first line I looked at to work. Watching how he approaches these positions has improved my ability to order candidate moves (for that matter, LIST candidate moves), allocate time, and generally think more practically. This was well complemented by an unorthodox suggestion of his: That I memorize the solution to every single problem in Ray Cheng's Practical Chess Exercises (nobody said improving was always fun!) and develop the facility to repeat them in 1-2 seconds per problem. Immediately after doing this, I played a fast game that bore out Alan's prediction: I ‘saw more' at the board.

Hoffman - Prilleltensky [A45]

5245

1.d4 Nf6 2.Bg5 Ne4 3.Bf4 c5 4.f3 Qa5+ 5.c3 Nf6 6.d5 Qb6 7.e4 Qxb2 8.Nd2 Qxc3 9.Bc7 Na6 10.Ne2 Qxa1
11Qa1.jpg
An intuitive queen 'sacrifice' (the idea of 9...Na6) which worked well in the game. It enabled me to bring out my pieces and play chess rather than mine my way through possible preparation. 
11.Qxa1 Nxc7 12.d6 exd6 13.Ng3 b5 14.e5 dxe515.Qxe5+ Kd8 16.Be2 d6 17.Qb2 Bd7 18.0-0 Ne6 19.Nb3 d5 20.Na5 Nd4 21.Rd1 Rc822.Nf1 Rc7 23.Ne3 Bd6 24.Bf1 Re8 25.Nc2 Nxc2 26.Qxc2 c4
This unremarkable little tactic, keeping the knight sidelined on a5, is straight from the book. 
27.g4 h6 28.h4 Be6 29.g5 hxg5 30.hxg5 Nh5 31.g6 f632.Re1 Bb4 33.Rxe6 Rxe6 34.Qb1 Bxa5 and Black went on to win. My apologies for showing a very strong player and nice guy obviously having an off day. 0-1

I should add that Alan is interested in doing this type of coaching with other motivated, hardworking chess players.

Finally, I have come to the game that pushed me over the edge. To be perfectly honest, it was a product of a strong player avoiding an opening repetition at enormous risk to his position. Like Evan Rosenberg and Daniel Gurevich before me, I offer it for the record.

Prilleltensky - Bonin [D15]

5246

1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.Nc3 a6 5.c5 g6 6.Bf4 Bg7 7.e3 0-0 8.Be2 Nbd7 9.0-0 Re8 10.b4 Nh5 11.Bg5 h6 12.Bh4 g5 13.Bxg5 hxg5 14.Nxg5 Nhf6 15.Bh5
BoninBh5.jpg
Quite frankly, a draw with this difficult opponent (and splitting first place with him in our quad) suited me just fine. Alas, IM Jay Bonin, carrying a 5-0 record against me into this game, had other ideas.
15...e5
I was shocked, but quickly saw I would be doing very well in the resulting positions. The draw would result after 15...Rf8 16.Bxf7+ Rxf7 17.Ne6 Qe8 18.Nc7 Qd8 19.Ne6 White could also grab the rook on a8 and play for an edge, but that was not my intention.
16.Bxf7+ Kf8 17.Ne6+ Kxf7
Another big surprise. My opponent may have been aiming to outplay me in the materially imbalanced middlegame that arises. But objectively, this is excellent for White.
18.Nxd8+ Rxd8 19.a4 exd4 20.exd4 Nf8 21.Ne2 Bf522.Ng3 Bh7 23.Ra3 Ne6 24.Nh5 Bg6 25.Nxg7 Kxg7 26.Re1 Re8 27.Rae3
BoninRae3.jpg
and White went on to win after all the rooks were traded. I overheard TD Steve Immitt's characterization of the ending. "Jay had the Bonin knights protecting each other! But the other guy had a queen and three connected passed pawns." 1-0

As a reminder of how much hard work awaits me in 2011, the International Master efficiently took revenge when we met at the National Chess Congress. I look forward to more battles with this warhorse of northeast chess!

So there is my journey from 2100 to 2200. It benefited from a combination of hard work, prioritizing, and excellent advice. There is plenty of analysis and suggestion here, but no magic elixir. Many of our habits are so deeply entrenched that serious re-learning, pattern accumulation, and p-r-a-c-t-i-c-e are required to break them. So shake off that silver bullet mentality, grab your scorebook (you do own one, right?) and spend some quality time with your games. You never know what you might find.  

See Matan's previous contribution to CLO, A Disney Chess Weekend: Matan on the K-12 Nationals and watch a US Chess Scoop video with him.