Sunday, February 22, 2009

Lost Among The Trees

I've been taking some time off of chess, partly since I've had two consecutive bye weeks in the Club Championship Qualifier. My Pocket PC lacks a Scrabble anagrammer so I've been trying to write one of my own. The one I wrote in SQL Compact Edition is ok, but it chokes on the wildcard searches, taking about 150 seconds to return the anagrams. I ran across a neat data structure called a DAWG - Directed Acyclic Word Graph which seems to be an optimization of searching the dictionary. A guy named Sam Allen wrote a blazing fast anagrammer that I wished to emulate, so I've been building and scrapping data structures, testing and debugging algorithms. The DAWG is a bit like Kotov's tree of analysis in that there is a starting position with branches for every possible node and branches from there outward to an exponentially increasing number of nodes. What's interesting about the DAWG is that it is almost within the human brain's ability to grasp the magnitude. Words in Scrabble range from 2 to 15 letters and can only end on one of 26 letters, so the DAWG converges on itself like a gigantic geodesic lemon, tapered at two ends. The analogous structures in chess are that the opening diverges and the tablebases converge upon checkmate.

I started reading "Word Freak". In one chapter, the author quotes Joe Edley, many-times Scrabble champion, basically saying that winning takes concentration on winning; everything else is extraneous. I probably quoted this once before, but in "Stand and Deliver", Edward James Olmos tells the kids in his math class that "You got to have the ganas." Perhaps I'm getting old. Perhaps sitting on my rating floor is getting to be too comfortable. But I'm finding that my priority is shifting to having a good time rather than winning. Do I really want to win the club championship? It should be a rhetorical question answered with an emphatic yes, but somehow it's not. Continental Chess Association is running the new Western Chess Congress tournament in the East Bay within a couple weeks of the Far West Open. Chess is coming to my doorstep, yet I don't think I'll play in either tournament, not because of a lack of funds or time, but I just can't psych up for the battle.

On Facebook, my brother-in-law got me hooked on a silly pseudo-adventure game called Dragon Wars. There is the gaming aspect of beating up AI monsters and taking their treasure, but there is also an aspect where you match up with other Dragon Wars players and battle for glory and treasure. I enjoyed seeing the new monster quests, but when other players beat up my character, I only got slightly annoyed. There is a small voice that wants to be vindictive as the game seems to engender, but so far I haven't succumbed to the dark side of the force.

I've been helping direct some scholastic tournaments. It's funny watching children struggle to manage board, pieces, clock, scoresheet and rules. Sometimes, kids forget to punch the clock and end up taking two turns in a row because they think that the clock button being up means it's their turn again and nobody is paying attention to what's actually moving on the board. Many mates in 1 are missed. Kids' technique often requires an extra queen and rook for mating material. It's only been two tournaments and I've seen kids come back from being down a rook and a knight, queening a pawn because the opponent got careless, and winning the game. I've seen kids stalemate with two extra queens, a rook, and a knight against a bare king. The players barely know how to call touch move let alone illegal moves. 50-move draws and three-time repetitions are never claimed, but they could be useful because I've seen players just keep checking with their queens, leaving their extra pieces at home. We have an extra rule that checkmate and stalemate must be verified. One team had the match in the bag until one player proposed a draw with an extra queen because he was afraid he'd screw up. They ended up losing the playoff. One child burst into tears for the last ten minutes of the game and just let his time run out. I thought of Tom Hanks yelling "There's no crying in baseball!" in "A League Of Their Own". All in all an entertaining new dimension to chess.

Lest you get the wrong idea, it's not that my main reason for directing kids' tournaments is seeing all the errors. It's refreshing to see children in their naive states before the years of frustrating plateaus squeeze all the ganas out of them. The parents have been surprisingly free of any craziness so far.

Well, back to my DAWG. If I could just figure out the algorithm for an iterative node depth counter...

Friday, February 6, 2009

Chess For Tigers

Last night I played a Class C player in the club championship qualifier. I had a few jitters about an upset, especially because the week before, just such an upset had occurred. Alekhine said "During a chess competition, a chess master should be a combination of a beast of prey and a monk." The late Simon Webb devoted a chapter in "Chess For Tigers" to "How to catch Rabbits".

"Do you know how Tigers catch Rabbits? Do they rush after them and tear them limb from limb? Or do they stalk them through the bush before finally creeping up on them when their resistance is low? The trouble with the first method is that even Rabbits have sharp teeth and when cornered can be surprisingly ferocious. So a sensible Tiger takes no chances - he patiently stalks his Rabbit, and when the poor thing makes a bolt for freedom, he pounces and kills it swiftly and easily."

Webb goes on to basically recommend the Keep It Simple Stupid method of chess plus waiting for your weaker opponent to make a couple mistakes. For the most part it worked for me in this game, but there were still several good moves that I missed.

Plus, I struggled with several of Blue Devil Knight's coach's rules, namely #1, #2, #4, and #5. But it was fun to watch the woodpile imbalance grow. As a contrast, Drunknknite sought complications by violating #3. I wish I could play like a swashbuckler.

I'm afraid I have come to the admission that my style is quite boring and tame. I'm a mostly toothless tiger these days. But I'm trying to play sharper openings to get the Eye of the Tiger back.

Monday, February 2, 2009

Reno Blogosphere

Nathaniel has joined the blogosphere in Reno. You can check out his site at Eric Shoemaker has revived his blog again at The Wizard's Castle. I'm experimenting with the possibility of linking to ChessBase-generated games like Drunknknite has already done. Here's Nate's game in a simul with GM Kudrin.

Sunday, February 1, 2009

Canterbury Tale

Rowan Williams, the Archbishop of Canterbury

On Blue Devil Knight's resurrected blog, he posted Part 6 of his review of Rowson's Chess For Zebras. Because of BDK's discussion, I bought Chess For Zebras, mostly from the good things I was reading, but partly because I own Chess For Tigers by the late Simon Webb and I can't resist collecting a set of books on an animal theme. I also own The Hippopotamus Rises, but it's not quite in the "Chess for (animal name here)" format. The discussion turned to Rowson's opinion of how we think about a chess position and that Rowson doubted the utility of the natural language narrative. His point was that NARRATIVES correlate more with weaker amateurs while IMAGES about how the position is resolved and evaluated correlate better with how strong professional players process the position. Not that I wish to contradict Rowson or even BDK, but I found it useful to understand the following endgame position in terms of language. Otherwise it didn't make sense to me. Plus, my memory requires all kinds of verbal and nonverbal underpinnings these days.

Centurini 1856. White to play and win.

This position was first shown to me by a friend at the club. He whispered a couple sentences to a low-rated kid to defend the position. I floundered about for a while gaining nothing except that I noticed that the kid kept returning to attack the d6 square. Only by inference did I begin to suspect there was something crucial about d6. My friend showed me the solution, but it didn't sink in why the d6 square was so important until four years later (a month ago) when I could verbalize what was going on.

We begin with White's winning plan, Plan A: to evict the Black bishop from the h2-b8 diagonal by placing his own bishop on that diagonal. Since c7 is guarded twice and the Black bishop can mark time on the diagonal, then the route to victory has to go partly through...

Plan B: to maneuver the bishop through the g1-a7 diagonal to a7 and then to b8. Will Plan B work? The only way for Black to hang on once the White bishop gets to b8 is to move his own bishop to the g1-a7 diagonal, wait for the enemy bishop to move out on the diagonal h2-c7 diagonal so that the pawn can advance, and then post his own bishop on a7 to kill the newly born Queen. But, assuming White retreated his bishop to e5, f4, g3, or h2 can then put his own bishop en prise on the g1-a7 diagonal to distract the Black bishop at a7. The pawn can then queen by force.

So Plan B works if Black simply allows it. In a sense, this endgame helps support the endgame Theory of Two Weaknesses. If there are two weaknesses to spread the defense enough, the stronger side can win by quickly switching targets. But Black has a defensive resource in that his king at c6 can move to a6 whenever the White bishop gets close to the g1-a7 diagonal.

Centurini 1856. White to play and win.

From the diagram position, 1.Bh4 Kb6 2.Bf2+ (2.Bd8+ Kc6 simply repeats the position.) Ka6 and the a7 square is defended. If the White bishop tries to tack back to d8, the Black king comes back to c6 to prevent Bc7: e.g. 3.Bd4? Bd6 4.Bf6 Kb6 5.Bd8+ Kc6 and now the position only differs in that the Black bishop is at d6 instead of h2.

This opens a new possibility that is not quite enough. 6.Be7 tries to distract the bishop from its guardianship of the b8 square, but 6...Bh2 Black wants none of it and returns to a square where it cannot be chased.

Now notice that the White bishop at e7 could get to a7 in two moves if only the Black king at c6 wasn't covering the pivot square on c5. This provides the winning idea and is the key to understanding why d6 is so crucial to this endgame. So White can take another crack at it. 7.Bh4 Kb6 8.Bf2+ Ka6 9.Bc5! preventing Bd6.

9...Be5 10.Be7 Kb6 11.Bd8+ Kc6 12.Bf6 Bh2

Notice that this diagram after move 12 is different from the diagram after move 6 in that the bishop can now pivot through d4 to get to a7 instead of c5. The Black king is now caught with his pants down. 13.Bd4! Kb5 14.Ba7 Ka6 15.Bb8 Bg1 16.Bg3 Ba7 17.Bf2 and the pilgrims finally get to Canterbury.

My tale of this endgame is that White must maneuver his bishop through d6 on his way to e7 and d8 in order to prevent the Black bishop from being in the optimum square at d6. When the White bishop pops back out from d8, it can quickly pivot over to a7 without running into the Black king. Whether this information is "better" stored in my brain as an image or as a narrative surrounding the d6 square, I'm not sure. But now that I can say it aloud, I feel that I understand it or grasp it, which is better than feeling like it is a memory that will run away as soon as my hippocampus turns its back.