Thursday, May 29, 2008
n. pl. sco·to·mas or sco·to·ma·ta (skuh-toe-muh-tuh)
1. An area of diminished vision within the visual field, surrounded by an area of less diminished or of normal vision.
2. See blind spot.
I was a big, if not religious, fan of the X-Files. I can’t say I saw every episode, but I sure saw a lot of them. The stories were fresh and it was always fun to see an appearance of Cigarette Smoking Man and imagine that one shadowy cabal tied all the greatest conspiracies together. My favorite episode was “Jose Chung’s ‘From Outer Space’”. I usually bust out laughing when I think of the kid saying, “I didn't spend all those years playing Dungeons and Dragons and not learn a little something about courage.” The directing that goes into filming different versions of the same scene based on the perceptions of different characters is tremendously amusing to me.
I work in a profession related to what Dana Scully’s character does, so it occasionally amused me to hear her mispronounce things. In an episode about El Chupacabra, Scully said “aspergillus” with a hard G sound like in bo-Gus, instead of the pronunciation I’m used to like the letter J, as in Gillian Anderson. Maybe it was an inside joke. In one episode entitled “Unrequited”, there was a Vietnam veteran who through some magic could vanish while you were looking directly at him in broad daylight. Scully explained the phenomenon using the phrase transient scotoma.
The last game of the Club Championship Qualifier was a mixed bag. I went into it thinking that I didn’t even really want to advance any more: partly due to fatigue, partly because matches in the finals just don’t appeal to me that much. I had White against Jerry Weikel, who has been state champion before, but due to life’s intrusions has limited his chess tournaments mostly to the club championship and an occasional big tournament. This was our tenth meeting with my score +3 –1 =5 in our previous nine games. This draw made it =6.
Afterward, I felt a peculiar mixture of elation at swindling a draw out of a lost position and disappointment that my tournament had ended with an overall failure of missing the cut and underperforming my rating. Against Class B: one win. Against Class A: one win and two draws. Against Experts: three losses.
Two learning points: I am annoyed to discover that I seem to have a blind spot for backward queen moves. One should always be careful of knights because they can turn the tide of battle with a vengeance.
Since the first step in developing expertise in any field is to apply or invent new vocabulary words that only you can understand, I’m going to start referring to scotomata in lieu of blunders and oversights.
Tuesday, May 13, 2008
I consider my opponent the strongest player in the club. He apparently has some bad days like the rest of us, but on good days, he can beat regular masters and draw IMs and GMs. After a couple rating performances around 2350, he seemed genuinely surprised that his FIDE rating was so high and that he could just claim a FIDE Master title. There was some argument about what it takes to become an established rated player in FIDE, but the current rulebook says nine games is it. It’s going to cost him $105, but I think most of us untitled players would think earning any certified master’s title – and perhaps flaunting it a little - would be priceless.
Like my other two games against the 2100+ rated club members, I didn’t feel like I was in this game at all. When I blundered, I tried to hang on and fight for some kind of counterplay, but I could find none. I agree with drunknknite when he says you have to fight to refute the erroneous idea of the one-move turning point. But I didn’t feel like I went down fighting so much as I rolled over and piddled on myself in these three games.
Before the game, I had good reconnaissance that he would play a Four Pawns’ Attack against my King’s Indian. I tried to book up with my coach, but one two-hour study session can’t erase years of floundering through my openings. Afterward, my opponent and I discussed the theory of this opening as well as the Budapest Opening that I experiment with sometimes. It was disheartening to see that if I knew my lines up to move twelve or so, he knew them at least to move twenty.
The thing I find most discouraging about the quest for the master title is that there are people like Edwin and Dana who have made it, but in a sense, just barely. Edwin has never been a National Master in the USCF system. Dana has been a National Master, but his rating drifts in the 2100s these days. But they both seem light years ahead of me in their chess calculation ability, and their opening and endgame knowledge.
I guess the silver lining is that I'm finally learning about these openings work and I'm paying for the experience by losing. As for my intermediate-term chess plans go, we'll see what happens this Thursday and let the chips fall where they may.
Thursday, May 8, 2008
The title of this post sounds like a slasher movie or at least a Heuristic ALgorithm that becomes a killer. The fact that a lot of my posts are titled with violent words or warlike themes, however, tells me that conflict is not altogether a bad thing. Conflict sells advertising spots on the news and tickets at the theater. The two main contenders for this year's Academy Award for Best Picture were "There Will Be Blood" and "No Country For Old Men". I have seen neither, but from the trailers, they seem to have a heftier than usual helping of conflict. There is a lot to be discovered in the crucible of conflict as long as you don't lose too much of yourself. What is chess but a conflict between two people abstracted into a game? I tend to focus on the self-preserving first part of Nietzsche's quote (Conan the Barbarian translation) "That which does not kill us, makes us stronger." while the meat of the quote is the self-improvement second part.
A few years back, one of my projects was trying to program my own chess computer engine. I did it out of the challenge of the thing, and the fact that it married my two favorite pastimes. I had these grand schemes of having the program spar with computer programs I have at home and even releasing it from captivity into the wild and wooly world of FICS or ICC. But I didn't get very far because my perfectionist side began nagging at me before I got far enough to implement the things it was suggesting. "Have you considered taking advantage of the MMX? Why don't you optimize your slow algorithm? Bitboards or no bitboards?" I managed to make a program that could play legal, 1 to 2-ply chess, but it was still close to playing giveaway chess. The slow algorithm started to implement beta cutoff, but I never got around to implementing killer heuristic.
At temposchlucker's great blog, he's exploring the mysterious workings of the mind in the context of tactical ability and trying in the formalized manner of a methodical researcher to discover a new method of training the stagnant chess player. Currently he's elucidating and formalizing the step of "scanning" or finding targets, training his mind to use the method, and seeing where the research leads. Andres Hortillosa is supposed to give us some tips on analysis at his Monroi blog. So far it seems big on framework, but skimpy on detail of scanning improvement. While I wait, I think I will begin to construct my own scanning method, of course plagiarizing from the best so that I can stand on the shoulders of the giants.
I couldn’t find any particular chess personality who coined this phrase, but apparently some chess teachers tell their students to "listen to the pieces." Basically ask, "What's out there?" Scanning and target recognition are important because as many of us who Fritz our games know, many combinations lurk beneath the surface, obscured by features of the chess board, hiding in our blind spots. A saying I made up a long time ago is "Somewhere on the chess board, hidden among inaccuracies, weak moves, and outright blunders, the best move is waiting." I think I was single when I made that up.
One idea I suggested which temposchlucker either didn't see or dismissed without comment was the idea of the killer heuristic. It might be improper to put it in the scanning stage, but in the context of tempo's frustrating 1.5 hour-long search, I thought killer heuristic could lead to faster analysis. Many years ago, I wrote out my move selection method. At the top on the list was "Can I checkmate him?" Working my way down the priorities, next came "Can he checkmate me?" and "Can I win decisive material from him?" and its reciprocal. The two recent examples of problems that temposchlucker gave had the common element that high priority checking moves (Qxg7+ and Bxe4+) could be used as guidewires to find the key move of the combination.
Years later, I lumped my steps into "Forcing Moves First". From Googling this phrase, I see now that there is a book I was unaware of by FM Charles Hertan called Forcing Chess Moves. I'm not sure I'm even going to read this book, since the gist seems to be summed up in the reviews. Hertan recommends "use computer eyes and always look at forcing moves first," no matter how silly those moves look to our human bias. Kassa Korley remarked that GM Hikaru Nakamura calculates very similar to a computer. After seeing what he can do, I believe it. Kotov's tree method is similar to the computer's thinking method, adapted for us mere mortals. Arguments for and against Kotov's method seem to boil down to whether it's too hard for humans to apply. Despite anti-computer sentiment (and perhaps envy), I would not be unhappy if I "suddenly got good" and found that I could calculate like Fritz does.
It was only recently I began to equate the efficiencies gained from "Forcing Moves First" to the concept of the "Killer Heuristic". The computer algorithm employing the killer heuristic will keep one or two moves in memory to check first because the move contained the features of a very violent and decisive resolution to the chess position. If the killer move proves strong, all subsequent moves will achieve more efficient beta cutoff (computer jargon for "Stop looking, there's no contest between this move and the other one") leading to a quicker evaluation that doesn't suffer too much in terms of accuracy compared with full-width-depth brute force minimax.
What is a forcing move? In my estimation, the pecking order of moves is:
Moves that checkmate or threaten to checkmate
Moves that check or threaten to check
Moves that win or threaten oodles of material
Moves that win or threaten a little material
Toward the bottom of this list of priority are all the positional attributes that bounce upward like quantum irregularities and are rather difficult to constantly and accurately balance against the material. But evaluate we must because most of the time, at least in my games, the positions seem rather quiet and there isn't a forcing continuation at the head of the line.
For the same reasons as the Killer Heuristic, computer chess algorithms sort the remaining moves they consider by certain characteristics, probably by the material they win and the capturing piece value (e.g. pxQ ahead of QxN ahead of quiet moves). This is done at every node, because the time cost of sorting reaps downstream benefits in the efficiency of the overall algorithm. Prioritization leads to time savings and efficiency.
I'm sure I have only scratched the surface of this one, but likely I'll come back when I draft my scanning list.
Wednesday, May 7, 2008
Today, I’ll try something a little riskier, prompting a disclaimer. I do not intend to offend veterans or their families who sacrificed so much, the people and nationalities that suffered in World War II, or the players of this game. By recounting a war story with a chess game, I do not intend to trivialize the events of World War II or any other conflict. Through my research, I’ve come to appreciate that Douglas MacArthur was a controversial figure. I come neither to praise nor to condemn nor to bury, but merely to recount. Hopefully with artistic license and attempts to stay away from bad taste, I won’t invoke the wrath of my four-person audience.
This game was the top board during the final round in the Open section. Since the other players in contention had drawn or lost, the winner of this game would win clear first, a cool $2,000. IM Enrico Sevillano had White against GM Melikset Khachiyan. Although Khachiyan has the better title, Sevillano actually has the higher USCF rating, so he’s definitely no slouch. Fpawn has already annotated this game for the USCF website wrap-up on the 2008 Far West Open and even quoted me, but I wanted to embellish a bit more. Sevillano opened with his favored Alapin system against Khachiyan’s Sicilian. On moves 11 and 12, White offered a double pawn sacrifice which Black only partially accepted, but White succeeded in marooning Black’s King in the center. Black then offered three pawns of his own to get his rooks out. Through a series of inaccuracies and one blunder in moves 23-30, White's large advantage dissipated to nothing. Black seized the initiative on move 31 and returned his King to the kingside to participate in the final siege on White’s King.
Sevillano,E (2567) - Khachiyan,M (2556) [B22]
Far West Open Reno, NV (6.1), March 23, 2008
1.e4 c5 2.c3 d6 3.d4 Nf6 4.Bd3 Nc6 5.Nf3 Bg4 6.d5 Bxf3 7.Qxf3 Ne5 8.Bb5+! Ned7 9.0-0 g6 10.Bg5 Bg7 11.e5!? An interesting pawn sacrifice. 11...dxe5 12.d6 e4! [ 12...exd6? 13.Rd1 e4 14.Qf4 d5 15.Bxf6 Qxf6 16.Qc7+- White wins a piece and needs only to tame Black's pawn roller.]
In December 1941, days after their attack at Pearl Harbor, Japan invaded the Philippines and took control despite an American military presence based in Manila headed by General Douglas MacArthur.
13.Qf4 e6? Black's king is now stuck in the center helping the overloaded queen defend the Black Knights against White's Bishops. [ 13...a6!+/=] 14.Bxd7+!? [ Fritz prefers 14.Rd1! Nh5 ( 14...0-0?? 15.Bxd7 e5 16.Qh4 h6 17.Bxf6 Bxf6 18.Qh3 retains a piece advantage.) 15.Qxe4!! Qc8 ( 15...Qxg5? 16.Bxd7+ Kxd7 17.Qxb7+ mate in 6.)] 14...Kxd7 15.Nd2 e5? Black trades material to activate his rooks. 16.Qxe5 Re8 17.Qxc5+- White has all the cards: material, space, king safety, and development.
MacArthur had become a legendary hero during World War I, but his reputation became tarnished when he was forced to flee the Philippines, leaving behind his forces who were ultimately captured by the Imperial Japanese Army. Once he landed in Terowie, Australia, MacArthur vowed, “I shall return.”
17...Qb6 Surprisingly, although Black is behind in so many ways, offering a queen trade becomes a good way to blunt the disadvantage of the vulnerable king on this move and on move 24. White knows it and declines, but he must give ground. 18.Qc4 Re6 19.Bxf6 Bxf6 20.Nxe4 Rc8 21.Qa4+ [ 21.Nxf6+ Considering how the fortunes of the two minor pieces diverge, trading into an all heavy piece ending might have been a better choice for White. 21...Rxf6 22.Qg4+ Re6+-] 21...Rc6 22.Rad1 Bd8
There is a mate in chess called the epaulette mate, so named since the King, flanked by two Rooks in chess diagrams resembles a high-ranking military officer wearing shoulder boards, also known as epaulettes. In the position after 21...Rc6, the alignment of the epaulettes isn’t quite right, invoking an image of an officer with poor posture and shoulders slumped forward.
Fritz estimates that White has a whopping +3.5 advantage. Over the next four moves, White's advantage evaporates. 23.Rd4? +2.5 [ 23.Nd2!] 23...f5! 24.Ng3? +1.5. White's knight becomes rather useless here. [ 24.Nd2!]
The summer of 1944 saw Allied victories in the Marianas, Peleliu, and Morotai Islands, shrinking the Japanese Empire and giving the Allies strategic airbases. The Battle of the Philippine Sea crippled Japanese naval and air forces.
24...Qa6! Again, the queen trade offer is best for Black. 25.Qc2 [ 25.Qxa6 Rxa6+/= leaves the game nearly equal.] 25...Qxa2?! 26.Ra4?!+/- +1.1 [ 26.Ne2 rehabilitates the knight. 26...Rcxd6 27.Nc1 Qa5 28.Ra4 retains some advantage for White.] 26...Rxc3! 27.Qxc3 Qxa4 28.Qg7+ Kxd6 29.Qxb7 Bb6 30.Qf3?!= 0.0 Dead even according to Fritz [ Pawn hunting might actually have been better. 30.Qxh7 f4+/=] 30...Ke7 31.Rd1?! [ 31.b3!?=] 31...Qc2!=/+ The first advantage to Black in the game. Black threatens to overload the White Queen with Bxf2+. 32.Rf1 'A sad necessity.' - fpawn. 32...Kf6 33.b3 h5 34.h4 Qd2 35.Qa8 Qd8 36.Qb7? White's Queen loses active threats for a couple moves, enough time for the Black Queen to pick up h4. [ 36.Qxd8+ Bxd8 37.Rd1=/+ White should hold a draw despite Black's better King and pieces.] 36...Qd4!-+ 37.Qh7 Qxh4
In the fall of 1944, the Allies turned their attention to the Philippines retaking Philippine soil in The Battle of Leyte.
38.Qh8+ Kg5! Other moves draw or worse.
On October 20, 1944, MacArthur waded ashore on Leyte Beachhead and triumphantly proclaimed, “I have returned.”
39.Qc8 Re5 40.Nh1 White's army is in quite a sad state.
Although, the Japanese Army retreated to the Philippine hills and held out for a year, Manila became secure enough that MacArthur set up headquarters and began planning the invasion of Japan.
40...Qd4 41.Kh2 Qh4+
Operation Downfall became unnecessary following the atomic bombings of Hiroshima and Nagasaki.
42.Kg1 Qd4 43.Kh2 Re4 44.Qc1+ Kf6 45.f4 h4 46.Nf2? Rxf4 White resigned 0-1
On September 2, 1945, General MacArthur received the formal Japanese surrender aboard the U.S.S. Missouri, thus ending World War II.
What surprised me about this game is how large an advantage can be overcome at this high level. I’ve been putting together games bulletins for the Reno tournaments for four years now and I don’t remember too many times that turnarounds occurred among 2400-rated players. The evaluations over the course of the game usually move in a monotonic manner from edge, to advantage, to winning, and to won. It seems at high levels, strong technique prevents comebacks, but this game serves as a good counterexample.
Friday, May 2, 2008
Polly recounts titanic struggles against her nemeses, Kong, Jr. and Kong, Sr., two gigantic monkeys on her back. But I’m going to turn around the reference toward the tragic figure in King Kong, the movie. I’ll play King Kong, a strong but dumb animal who doesn’t really stand a chance against the forces of man and woman. The opening is my Scandinavian beauty whose allure is irresistible. My opponent is the ruthlessly efficient air force comprised of Curtis Helldiver biplanes.
Incidentally, the only “King Kong” I’ve seen is the 1976 version starring Jeff Bridges and Jessica Lange. I haven’t even seen Peter Jackson’s 2005 effort, although I heard it suffered from having too many T-Rexes. Black and white classics are just not my cup of tea.
Kong falls in love with a Scandinavian beauty he doesn’t understand, but he desires her anyway (move 3). Smitten, he’s easily captured and transported to New York (move 22). He breaks his bonds, grabs his femme fatale, and desperately runs for freedom in a strange land (move 23). Kong makes his last stand atop the Empire State Building (move 36). His lady having deserted him and surrounded by biplanes mounted with machine guns, he swats at them as best he can, but they just keep coming (move 41).
Finally, Kong falls to earth and dies partly from the trauma, but mostly from the broken heart.
Going into the game, my heart and courage weren’t in it. My coach and another chessplaying friend noted my psychological defect. I’m not sure how to break the bad cycle of morale. Perhaps I need to face it head on and play a six-game match against my nemesis, hoping that I’ll catch him in one of the games and rediscover my confidence.