Wednesday, March 30, 2011

Tricky Tricky Chess960 - part 1

Wooow. Every now and then a mid-game position arises that is totally Chess960 specific and is soooooo perculiar but kinda beautiful. Keep in mind that most of the time Chess960 principals are no different to standard chess. It' just that every now and then something crops up in 960 that cannot happen in SP518. When that happens, sit back and enjoy the show.

The main problem I ask quite often in Chess960 is:

When is a Bishop = Rook ?

I've talked about before and that I've also talked about at as glider:

It's a great topic headlined:
"Does Chess960 change the way we play chess?"

Nonowho says:
".....I see people giving up rooks for no apparent reason....."

Softbabe says
"....Chess960 confuses me..."

Glider says:
"Great question nonowho....."

Atos says:
"I think it's just that people tend to blunder more in 960 because the positions are unfamiliar"

Well here is one of those bizarre Chess960 positions! An exchange sacrifice has taken place and white is up a rook but down a bishop....exactly what Nonowho has noticed. The point about this position is that it shows us how the Chess960 jungle sometimes confuses the relative strength of pieces when pieces get blocked up, hemmed in, rendered useless or just can't get into the thick of it in time. This position is totally unique and cannot happen in standard chess. You would think that black's knights blocking the centre pawns is bad. You would think that white could castle g-side and release the h1 rook. You would think that it is reasonable for white to get rid of black's dark bishop and simplify.

So on and so on! I won't paste the whole game because it's not that instructive. However the way that black's rook got nuked was that it captured on the c-file, got captured itself by a bishop and that bishop was cleaned up by one of black's knights!.....

SP136 - Black to play
Firstly just to keep it as simple as possible here are two questions that are answered at the end of this blog post:
  1. Write down the coordinates of a dark square diagonal that is weak for WHITE
  2. Write down the coordinates of a light square diagonal that is weak for WHITE
There are dozens of amazing lines in this position, but I'll give just the main lines analysed with Rybka 4-960. I break it into slabs of five moves because I think that most of us amateurs should be able to keep five moves in mind and in any case it is good practice to do so:
1.      ... Bg6   attacking e4 and indirectly attacking Qc2
2. f3   ... h5!   take the dark bishop I don't care!
3. NxB? ... Qb6+! the only move and black is winning
4. Kf1  ... fxe5  white had no real choice (Nd4?! ...)
5. Re1! ... Nxe4! white's Re1 is necessary and takes even Rybka4 a while to find....
    6. Rxe4  ... Qc6 white had no choice
    7. Ke1!  ... d5! best moves!
    Now one of the most remarkable tactical sequences I've ever seen!
    8. Rc4!! ... Qe8 The rook moves into the pawns attack! (...QxR??)
    9. Rc8!! ... BxQ An amazing chase sequence!(...QxR??) 
    10.RxQ   ... RxR Black has a pawn but for how long?!

    The rook became a "suicide" piece chasing and attacking black's queen without care. Optically it looks like the black queen should capture the rook and kill it mercifully! But the point is that white's active rook is essentially a zombie piece. So in this variation overall:
    Bishop = Rook!?

    (  Thinking about the "zombie" rook a bit more, the actual reason why the tactic works is because black needs to continue to direct the affairs. If the black queen captures the rook, she gives white the move. So the queen prevents that by withdrawing to the protected square of e8 knowing that the white rook will chase her down there (white doesn't have an option). However by arriving at her e8 home under guard by her f8 rook, black is then able to be the first to capture a queen with BxQ and in so doing eliminating the problem of that undefended bishop on g6. Chess tactics fundamentally are about achieving more than one thing at a time. What is so beautiful about this tactic is the way that the black queen must die deep in her own territory on the royal e8 square.  )

    That line is just one tiny variation that gives you some idea of the bizarre features of this specific Chess960 scenario. You can load the position into the computer yourself by pasting this FEN code in: 3qbrk1/pp1pp1pp/3nnp2/4b3/4P3/1NP1N3/PPQB1PPP/2R3KR w HC - 0 11

    Hope you enjoyed that rook chases queen tactic! Do not fret if you don't understand (I don't either it's just that I am using a computer like a microscope to gaze down into the position). Like I said the Chess960 jungle can be bizarre at times. Ok the answer to the two questions above are:
    1. Answer: The g1-a7 dark diagonal is weak for white
    2. Answer: The b1-h7 light diagonal is weak for white

    Monday, March 28, 2011

    Chess960 Opening Survival Guide: The Edge Pawns Are Safe!?

    After the last post on SP362, I began to think just a little bit about the "prickly" pawn. It is not poison to swallow, but just how unpleasant are the edge pawns to capture across all nine hundred and sixty Chess960 starting positions? The theory is that white can forget about capturing the edge pawn on the a/h files. If it is grabbed, it will prickle and cause slow bleeding. Out in the Chess960 Jungle, that is not a good way to start!

    Why are the edge pawns safe then? The proof should be pretty simple but is still in the works. Here is a really quick knock up:

    The theories:
    1. All arrangements of back rank pieces directly or indirectly defend the edge pawn
    2. A knight may attack the edge pawn, but it's opposite always defends it adequately.
    3. Tactical captures of the edge pawn that lure a defender onto it's square for subsequent capture, do not work.
    Proof of 1:
    What are the combinations of back rank pieces that leaves the edge pawn undefended and without a simple developmental defense?.....There are none!

    BR****** This case is easy to prove because playing b6 protects the edge pawn and opens the long diagonal for a bishop. This is always good:

    Proof 1A:
    -> It simply takes too long to exploit the a6 hole with a knight, and the best outcome would be to capture a rook. IT'S NOT WORTH IT.

    3) NR****** This case was proved in the post on SP362: The only way to immediately attack the edge pawn is via a diagonal attack:

    Proof 1B:
    -> But if the queen takes the edge pawn diagonally, the defense of Nb6/Ra8 traps her which is always BAD.
    -> If a bishop attacks the a7 pawn, Nb6 defends it and develops. I agree that this is the weakest argument, but I highly doubt that biffing the knight via the pawn advances a4-5 or c4-5 actually work. BY THAT TIME BLACK WILL HAVE DEVELOPED SUFFICIENTLY TO COUNTER THE TEMPO EXPENDED.

    Proof of 2:
    What combination of back rank pieces allow for a simple defense of the edge pawn, but that can be rebuffed in some way or another?

    *N****** This case is reasonably simple to prove because Nb8-Nc6 is a useful move that protects the edge pawn immediately:

    Proof 2A:
    -> Note that a mate via Nxa7# is not possible because if a king were on c8, then a rook must be on a8 which then defends a7.

    {but what about the case that a rook or queen then get drawn onto a7 and captured? See proof 3}

    Proof 2B:
    A knight on c6 can be biffed by either the b-pawn or d-pawn advance,  but by the time the advance has taken place, the biff can be defended:
    -> A pawn biff takes too long

    Proof 2C:
    The case that is a bit more complicated to prove is when a bishop is in the opposite corner and captures the c6 knight so that the edge pawn can be taken:
    -> But think about what starting positions cause this situation....
    If a bishop is at h1, then either a queen or bishop must be on g1. If that is the case then either a rook or knight must be on f1. If that is the case then either a king must be on e1-d1-c1. A knight must have been either on b1 or d1 and a rook cannot be on a1.
    {These cases need a bit more thought over coming days! Help Mark!}

    Proof 2D:
    The final case is where a knight on a1 or c1 tries to exchange off a knight on c6 via a5 so that the edge pawn can be diagonally captured:
    {Still to do....Help Mark!}

    Proof of 3:
    Still in the works!

    Pheeeeoh! That was not as simple to prove as I thought!

    There might be an even simpler proof. One simpler explanation is that the edge pawn is not worth playing 3-4 moves in order to capture!

    Whatever. I think the edge pawn is as safe as it was in standard chess SP518. C'mon try to disprove me! I After all ANY discussion about Chess960 is better than none. Even if you want to take these ideas and write a book it is not my concern. All I wish for is that Chess960 becomes more popular.

    The next case would be to prove the severity of threats on the b-c-d-pawns. That will interesting! I'm slowly getting better at formulating proofs that require some forward and some reverse engineering to justify! Already in my head it is clear that there is no significant threats on these pawns that cannot be countered with initiative, but the proofs will have to wait for now.

    Saturday, March 26, 2011

    Chess960: SP362

    Well what do you say about Chess960?! It just keeps on surprising! Still we have not found any seriously flawed starting position yet (where white's advantage is so great that it destroys a sense of competition). This starting position is an an example of a really interesting starting position that has knights in the corners and rooks that hug the king. Please Chess960 simplifiers, don't take this position out of your reduced set!

    Over at Mark has highlighted SP362 as one of those SP's that contains what you can say is a "prickly pawn". A prickly pawn is a pawn that causes an immediate pricking sensation for white if he grabs it, followed by slow bleeding :-). The amazing thing is that after a shallow 16 ply (8 moves) analysis in Rybka4-960 of all 18 possible first moves for white, at best white's evaluation is in the range {+0.07...-0.25}!

    SP362 - The Prickly Pawn
    Firstly here is the variation that Mark played in a real live game scenario. It contains a nice tactic that Rybka4-960 finds:
    1. f4 d5!     Black gives up the a7 pawn
    2. Qxa7 Nb6   Black get's compensation...
    3. Qa3 Ra8 
    4. Qb3 e6 
    5. a3 Bd6     White's move could well be required
    6. Bf2? Nc4!  White's knight outpost is stable!
    7. e3? Rxa3!! A nice rook sacrifice!
    8. bxa3 Bxa3+ 
    9. Rb2 Bxb2+ 
    10.Kb1 Bf6 
    11.Bxc4 dxc4 
    12.Qxc4 Bc6   Black has a winning position

    Ok let's get to the crunch question, why is the a7 pawn a "prickly pawn" for white. The simplest explanation is that because of black's structure Nb6/Ra8, white's queen must retreat after the capture. However no matter where the queen goes, she either comes under attack down the track or obstructs white's development from there. For example one queen sequence is:

    SP362 - The Flight of the Knight
    The red squares show how white's queen must retreat and subsequently block further development and also waste tempo. That amount of waste is a serious penalty for the material gain of the a7 pawn. So the conclusion is that black doesn't have to protect the prickly pawn! 

    With that question answered, what possible replies does black have to 1.f4? If you notice, even 1...Ng6 is possible! Yes believe it or not....

    Get this! Black has a stable outpost for a knight on c4 that is continuous across many variations! When you think about it, the c4/c5 outposts are there from the start for a number of reasons:
    1. b3/b6 to stop Nc4/c5 is bad
    2. d3/d6 to stop Nc4/c5 blocks the f1/f8 bishop
    3. A knight on c4/c5 can go to a3/a6 and stay there
    4. The enemy knights cannot dislodge a knight on c4/c5
    5. Exchanging the f1/f8 bishop for a knight on c4/c5 is dubious
    6. c4/c5 can quickly be protected with Qg1/g8 and Bf2/f7
    7. c4/c5 is a quickly accessible location for the a1/a8 knight
    Here is an extreme example:
    1. f4   Ng6!?   What the??
    2. f5   Ne5!?   What the??
    3. d4   Nc4!?   The knight is stable on c4!
    4. e4   d5!     Rybka4-960 is happy for black!

    SP362 - The amazing knight outpost swindle!
    What do you say about that! Chess960 is amazing. Ridiculously funny at times and quite tragic at others! I'm sorry but I have to take a break this is making me laugh too much. You would think that white is doing ok but Rybka4-960 likes the position for black even at 13ply with a score variation of {1. Bxc4 -0.30...10. Bd3 -0.63}. I've seen this phenomenon before in 960. The amazing swift and majestic knight! Here is the water torture continuation:
    5. exd5 b5!    Black prepares a rook lift down the track
    6. b3?  
    White can't shake off that knight without troubles because the knight becomes stable on a3. Essentially you can ask the question what is black's knight going to do on a3, but the answer is not what it is going to there, but what is white's knight going to do on a1!

    Thursday, March 24, 2011

    A Second Super GM Chess960 advocate shows em all!

    Congratulations Levon Aronian for winning Amber! That means this year there has been two victories for super GM's that have played Chess960 and advocated it at one stage or another. Aronian at Amber 2011 and Nakamura at Tata Steel 2011. Great stuff.

    Here is an example of one of their Chess960 encounters (scroll to bottom of page):

    Aronian quote from the Misha interviews care of
    MS: You are one of the best Chess-960 players in the world...

    LA: It is difficult to measure, as there aren’t many tournaments nor many people playing it. However, I like Chess-960. There are fresh positions, not so overloaded with tactics in the opening. I enjoy it.

    MS: What specific skills are required to succeed at it?

    LA: The most important is that one has to seriously think about the very first moves. It is also important to sense positions, to “intuit,” as they say in Moscow. A well-developed imagination is crucial as well.

    MS: How do you decide on the first moves?

    LA: Gabrielian told me that the best opening approach is similar to normal chess – the first move must free one of the bishops. But I am not so sure. First I just look at the initial position, deciding where to bring my pieces and how to seize the center. Of course, controlling the center is paramount. However, it is not always possible, and then you have to think about restricting the opponent’s play.

    MS: Did you master its complicated method of castling quickly? (blog comment: complicated?? C'mon castling in Chess960 should be understandable by anyone who is keen enough to play chess at all)

    LA: Usually I postpone it for as long as I can! In the beginning I try to
    develop my pieces harmoniously. (blog comment: Aronian delays castling not because he doesn't understand the rules, but that he likes to play as actively as possible)

    MS: Is there already an established elite in Chess-960?

    LA: No. There are too few people playing. You know, sometimes I compare regular chess and Chess-960 with two forms of jazz, swing and bop. Swing is casual, everybody listens to it. Bop is different: you don’t know what you play in the beginning. You just play and see what happens. A free-style jazz. Naturally, bop doesn’t have as many supporters, and mastering bop is more difficult. One must have so many melodies in one’s head to play bop. This is similar to Chess-960 – one needs to learn many more positional patterns to

    Tuesday, March 22, 2011

    Chess960 and Agricultural Practices

    What do we want to be? Nomadic chess960 wanderers or settle down to domesticated subsistence? Yeah it's a bit bizarre but hey don't blame me blame my mind!........

    It's amazing how right at this minute I am sure that some chess players on the planet are deciding what they think is a useful subset of the total set of 960 chess starting positions. The debate is a good thing but boy does it have so much in common with land clearance practices in agriculture.

    Think about it. Chess960 is a vast jungle at maximum possible diversity. Could humans live in such a jungle? They would need to clear a bit of land for themselves perhaps? But what land and how much? What diversity is expendable and which is not?

    In this current age we are familiar with Chess SP518. Will we clear the Chess960 jungle to make it look more like SP518? If we do that, it might be much like what the English did when they came to Australia. They turned Australia into an English landscape and then came to realise later on that they had destroyed so much diversity. Through the force of habit they tried to transform an unfamiliar landscape into a landscape they were used to and could understand.

    Will we clear just a bit of the Chess960 jungle? Perhaps just the starting positions that we think are simply unnecessary and then we leave the rest. What justification would we use? Who will decide that? Kramnik? Anand? Kasparov? Karpov? FIDE? What authority will decide that? Do we really need to cut back the Chess960 Jungle at all? When such authority speaks for the broad chess playing public (we the patzers!), is it not possible that the patzers will accept Chess960 easier than authority can perhaps? That would be a very interesting possibility.

    Here is an example. Perhaps this "weed" in the Chess960 jungle needs to be chopped?:

    SP131 and it's variants - Bishops in the corner

    The bishops are in the corner and it seems that on first glance they will most likely be exchanged off the board. In my own games from this position that does seem to happen more often than not. But is that undesirable? Personally I think bishops in the corners is a very beautiful subset of the Chess960 positions that emphasizes the structural framework of pawns in a more skeletal form. The framework is vital to get right because the knights need outposts and they need paths to those outposts that they can traverse through the pawn structure.

    Should we cull those positions Vladimir Kramnik? You the great master what do you think truthfully and without your tainted SP518 glasses on? If not to remove the bishops in the corners then what, the knights in the corners? Only the rooks in the corners?

    As far as I'm concerned, it's how we perceive the diversity and how to work with the existing diversity. It's not about reducing the diversity down to suit ourselves. I guess it's the same debate about whether we want to be nomadic wanderers living off what we discover through curiosity and accumulated knowledge, or simply settle down to a cleared patch of domesticated subsistence that is more certain and predictable.

    The curiosity is that the seasonal wanderers that were indigenous to my country came to great insights about diversity itself and have outlived most if not all other civilizations. They accepted diversity and worked with it. But unlike real life, us nomadic chess wanderers that move through the jungle of 960 do not have to worry about diseases and predators. The only thing we have to worry about is ignorance!!

    Thursday, March 17, 2011

    Chess960: The sublime and the ridiculous

    Check this ridiculous tactical situation from one of my Chess960 games the other day and independently verified by Rybka-4 960. Black is fighting to survive and the lesson of this variation is don't give up!:

    SP130: Black to move and retain the only fighting chance!
    Black is down a pawn, his king is being harassed and his pieces are badly coordinated and trapped. A move like ...g6 doesn't work because after Qh3 white get's a nasty combination of attacks including the a1-h8 long diagonal. What does black do?
    1. ....... Qxf4!! Black's only chance
    2. Qh3 Qxe5!! Black must accept an exchange sacrifice
    3. Qxc8 Bc6!! The only move. Black tries to trap the white queen with ...Bd7
    4. Qh3 d4!!     Clears the long light diagonal and forces white to respond
    5. Nd1 Qa5!!  Black's queen attacks white's weakest point d2
    6. Bxc4....       White takes material and begins to target the f7 square

    Intermediate diagram: Black to play with the only fighting chance:
    1. .......b5!! Black encourages an attack on the f7 square
    2. Bb3?! Qxd2
    3. O-O O-O A forced mutual castle showing how vulnerable both sides are!
    4. Rb1 dxe3
    5. Nxe3 Bc5
    6. Rfe1 Nf6 and black's diagonal bishop attack compensates for being an exchange down
    There are many many other incredible tactical motifs in this position including mating lines on white's king along the a7-g1 diagonal. The position is definitely a relic from the odd starting position of SP130. There is some bizarre stuff going on in Chess960 and occasionally the position explodes into dozens of complex tactical lines. This is one of those positions. Although I did not find these lines on my own, it's really enjoyable just to ponder them in retrospect.

    Monday, March 14, 2011

    Small Town Chess960

    Hitrel at messaged me the other day. He joined the AdvancedChess960 group there a few months ago and we got to talking a bit. He lives in a small town and wants to start a Chess960 club in his town! A great idea which really touched me because I have wanted to do the same thing many times. He asked how he could fund such a club. At this stage I'm clueless for ideas. The only idea I could come up with for Hitrel  is to make some spare money via blogging online if he can think of a topic that creates enough interest! :-)

    I guess the main costs these days for starting a chess club is the hall hire and insurance. Good luck to him. Whatever. Chess960 is a great chess club activity because it encourages open minded fun and lively discussions for all ages. I should suggest to him that he get in contact with Adam at his blog:

    Saturday, March 12, 2011

    Chess without the 960

    Well, I thought I might as well start up a list of reasons why it could be that Chess960 seems to be hovering around the five percent mark of all Chess activities on the internet at best. I'm guessing five percent based on what I have seen personally on the chess servers and general internet news.

    Top Reasons why I think Chess960 is at 5% of total Chess activity with no sign of change:
    1. Chess960 does not quickly improve a players "Chess rating". There is an irony with Chess960. Chess960 does improve your chess play because you are thinking both strategically and tactically from move one of the game and you acquire a lot of general purpose and new skills. I have seen the improvement in my own standard SP518 chess play where I'm more willing to play actively and more flexibly. However the standard chess opening can be pre-arranged by memory tasks. Therefore in rated games of standard chess where time is everything, memory tends to trump general skills because skill needs that little bit of extra time to execute. Thus the general "over the board" skills that Chess960 fosters does quickly and easily translate into an improved rating when under time pressure. If it does not translate into better ratings, then people won't play Chess960. This is because ratings improvement is by far the biggest motivating factor why people play Chess online. It is much easier to memorize specific opening lines from a book or CD which tend to give quick and instant results.
      I think the exception is at the elite level where players use Chess960 to freshen their imagination so that when they analyse volumes of standard chess opening theory, they might be more open creatively to novelties. I personally think that explains some of the reason why Nakamura has actually continued to improve as a chess player because he has personally experienced the creativity that is inherent in Chess960.
    2. Players are so used to seeing the same Chess opening configuration. When they see other configurations that appear in Chess960, it is an affront to their familiarity. When affronted, we humans either react with aversion or curiosity. For some reason when Chess players are affronted, they tend to show much more aversion than curiousity. This phenomenon of aversion is prevalent at all ratings categories from GM's to Patzers. The only time curiosity trumps aversion, is when a player is willing to sacrifice some chess ratings points to explore the vast universe of Chess960 just for the sake of it. But the guess is that 95% of chess players are strongly motivated by ratings, and thus aversion against Chess960 is much stronger than curiosity towards Chess960. Once there is a strong aversion, all sorts of bogus reasons begin to circulate in the minds of the Chess community as to why Chess960 is bad. But the only true reason is that aversion has won over true curiosity. There is always a bit of "superficial" curiosity. But what appears to be curiousity is mainly an activity of confirming one's own preconceived ideas.
    3. Players are unused to seeing some of the Chess960 piece configurations that produce new tactical and strategic ideas. For example Chess960 contains very interesting configurations of pieces such as knights that are only two squares apart. This produces specific tactical motifs that must be understood and are rarely seen in standard Chess. Because we are unfamiliar with these motifs, we get bitten when we play Chess960 and then shy away from it and make up excuses why we do not like it.
    4. The Chess960 marketing fire has not been lit yet. Without a competition base, Chess960 celebrities do not form. If Chess960 celebrity is absent and if it true that Chess960 does not improve a players chess ratings quickly and easily, then a players Identity does not attach to Chess960. If that identity does not attach to Chess960, then marketing has no fuel to light a fire of enthusiasm. If marketing does not emerge, then players perception do not change. In other words, a tiny campfire was lit when the Germans started Mainz and Chess960 competition a decade ago, but the campfire went out.

    Other possible reasons I have overlooked:
    1. I am misguided about how habits change. I think that Chess960 should be taken up a lot quicker than it is, but sociological scale habit might take much longer to change than the pace I imagine from my tiny individual perspective. Likewise, I am missing other possible phenomenon simply because of my tiny perspective on what is a big process. Other possible big scale processes I am not seeing are the competition of other forms of entertainment, or that the very young are not being exposed to Chess as much as as previous generations and thus we are left with an older generation of standard chess players who are not as willing to adapt to Chess960.
    2. I am misguided about how perception works. In other words, there is some cultural/inherited reason within particular societies that causes Chess960 to be perceived as a more or less worthwhile activity. It's interesting to contemplate why Chess960 is more popular in Germany for example.
    3. That Chess960 contains some kind of flaw in it that I am overlooking. But what flaw? Standard chess is included in Chess960 and I have not seen one bad starting position. Yes some are more "drawish" than others and some are simpler than other positions. But I haven't come across a truly flawed starting position yet (one in which one player's chance of winning is so high that it destroys the experience of competition).

    Sunday, March 6, 2011

    Chess960: SP120

    Here's another one of those starting positions that needs a bit of eye training to find the inherent structural weak spot. It's not hard to find but is good instructional material for playing the Chess960 opening. If no-one ever reads this blog, at least it will help to remind myself!
    SP120 - Find the main weak spots
    1. Look for the undefended squares (e2, e7, a1, a8)
    2. Find quick paths to the undefended squares - first the knights then the bishops (because one of the rules of good opening play even in standard chess is to consider playing the knights before the bishops)
    SP120 - What the eye must see - The KNIGHTS
    The key focus is on the e2, e7 square because if a knight get's there it's game over. But it's not just the square itself, but the intermediate squares that attackers will use to get to it! So here are the main features:
    1. The opening focuses on the knights quick attack paths
    2. The intermediate squares for white are f5 and d5. The structure of the starting position suggests that those squares may become outposts. Why? Because black is unlikely to play pawn e6 (blocking and undeveloping) and unlikely to play g6 (weakening/blocking/undeveloping)
    3. Knowing that f5+d5+e7 are weak points for black, both players proceed with their plans. One to attack the weakpoints, the other to neutralise them.
    4. From those weak points it quickly becomes clear what moves are bad. For the black player, a particularly bad move is ...c5. This is because it only strengthens d5 for white. This is because black playing ...e6 is not one of his preferred plays (it might be necessary but that is life for being black and down a tempo!)
    Ok let's move onto the bishops:
    What the eye must see - THE BISHOPS
    1. In terms of the bishops, the key focus is on the b4 (b5) square which fires again at the e7(e2) square.
    This post does not try to suggest a good opening idea, just to highlight how the Chess960 eye must be trained. Once a player has played a few thousand Chess960 openings, these structural elements are visualized in fractions of a second. If kids were alive today that had grown up playing only Chess960, their ability to quickly pick up these structural elements in any of the 960 positions would astound mortal players of standard chess.

    Saturday, March 5, 2011

    Annex Chess Club and Lateral Thinking

    Here's a nice read about Chess960 at the Annex Chess Club:

    And some good old natter on Chess960 from them as well:
    "I should have said Catholicism is slightly more radical an organization than Chess. It changes about once every 100 years, most Chess breakthroughs are on at least a 200 year timeframe or worse. All fun aside, and no harm is meant to the fine fathers of the Church, conservatism is often good. I prefer evolution to revolution in Chess, since the principles of the game are very sound. 960 is a natural extension of the game, not a replacement. Having 959 -other- starting positions makes line memorization irrelevant. It does also put the onus back on the human player to make good opening choices, that are usually based on very esoteric and subtle considerations i.e. if I do this, I get saddled with this blocked piece in a pawn endgame. Obvious to us, but not so obvious to Fritz with a 100 move computational horizon."

    From my personal perspective, it's not so much that memorization in the opening phase is irrelevant, it's just the form of memorization that we are used to is irrelevant. Instead of memorising vast trees of specific lines, we now remember vast sets of specific characteristics and themes. 

    It's just a conceptual distinction but it can be thought of like this:
    Memorization in the opening phase of Standard Chess:
    Mainly vertical but some horizontal (vertical thinking/memorization)

    Memorization in the opening phase of Chess960:
    Mainly horizontal with some vertical (lateral thinking/memorization)

    Ultimately most players want to win and see their rating rise so they do not care whether they think vertically or laterally (it's just a concept anyway). Players are motivated by the desire to improve their rating. The strength of a Chess rating means that there is little motivation to study something that does not directly impact on that goal. It is much better to simply learn some sharp and specific lines that work most of the time. It is not immediately obvious how to do that in Chess960 (although there are ways). That is why Chess960 is slow on the uptake.


    A once in a life time midgame 2R+3P v 2B+N+4P

    Not sure if Chess960 was the cause of this odd midgame situation or not (see the end of this post)! Some scenarios are so unusual that it's best to record them down somewhere because it's possible you will never face them again in your life! Here is one such midgame I got into.

    Let me say that this position is so uncomfortable for me as black because I have difficulty visualizing how two opposite coloured bishops can compete against the two rooks. Doing a bit of analysis on the computer about this situation, many lines seem to revolve around an incredible swarm of pieces on white's king where black must use every piece including his king to force white's king out into the open! The first standout feature is that ...Kg5 is a little safe house for black's king that begins such a swarm. The swarm is necessary not to affect a win, but to cause a draw. All black's minors must cooperate with each other and not only threaten to attack, but shield the king from checks by the rooks. It's pretty tricky stuff.

    Black to play and find the only move that retains initiative

    Black must do something about the backward a7 pawn otherwise white will get a single passed pawn and can use both rooks in cooperation to queen the pawn. But what should black do?

    38... Nb7? 39. Rd7 Bc6 40.Rff7! Bxd7 41. Rxd7 Nd6 42. Rxa7 and white's queening threat put's black in some hot water

    Ok so here is the amazing line!
    38. ... Nb5! 
    39. Rd7 Nc3 from this outpost, black has many ways to retain initiative
    40. Rxa7?? Ne2 mate!

    Alright so I ask the big question. Is it possible that such an unusual midgame was more likely to emerge because of it's particular starting position? Here is the starting position that this midgame emerged from:

    SP119: A Suffragan and Archbishop combined with Chivalry Knights
    The feature of this SP is that in Naming the Bishop Pairs one of the bishops (The "Suffragan Bishop") is quite likely to remain on the long diagonal for a long time, while the other bishop (The "Classic Archbishop") does not seem to come under any attacks from the knights in it's most likely direction of travel. (Visualise the Archbishop and where it might go and compare it with where the knights will try to strike at it). According to Naming the Knight Pairs we see that in this SP we have the "Chivalry Knights" starting on light squares (for black). I think that is a clue! The Chivalry Knights play out and control a 7x5 area but struggle initially to attack their opposite color. This SP might encourage both bishops to remain on the board for longer. 

    In Chess960 my biggest question fundamentally revolves around this:

    When does a bishop become as valuable as a rook!?

    This SP might help to answer that question.

    Thursday, March 3, 2011

    A chess960 special tactic?

    Sometimes I do wonder whether out of all the possible tactical ideas in the opening phase that are possible in chess, they're not all seen in the standard chess SP518 opening. Could this be an example? I found it when I was playing on the other day (a great site for live Chess960). Notice that the opening phase is actually not yet complete arriving at move 14:

    SP854 (RKBNRBNQ):
    [Event "ChessCube Game"]
    [Site ""]
    [Variant "chess 960"]
    [SetUp "1"]
    [FEN "rkbnrbnq/pppppppp/8/8/8/8/PPPPPPPP/RKBNRBNQ w EAea - 0 1"]
    1. e4 Nc6
    2. Nc3 g6 
    3. g3 d6 
    4. f4 Nf6 
    5. Bb5 Bd7 
    6. d3 a6 
    7. Bxc6 Bxc6 
    8. Nd5 e6
    9. Nxf6 Qxf6
    10. Ne2 Bg7 
    11. c3 Qe7 
    12. Qf3 f5 
    13. h3  fxe4 
    14. dxe4 ...
    Black to move: Find the standout winning move:

    Ok here is where the real twist in the plot is. The hint is that the tactic is about the e4 square. But how does black break through? At the time I played d5?! which is not the answer. The answer is that here we see a tactic which is a kind of an x-ray tactic but not on a piece, but on a square (e4).
    14. ... e5!      counter intuitively the only clearly best way to attack e4!
    White is actually is in a kind of a zugswang! Lets assume that white decides to be very cautious and just keep developing some how and get his king out of the way of that terrible bishop fork attack of Bxe4+ (forking the queen and king)
    15. Bd2 exf4 and black has opened up the e-file very quickly and the rest is fairly obvious:
    16. O-O-O fxg3 17.Nd4 Bxe4 18. Qxg3 Qd7 - +
    Ok so white decides that it is better to do something proactive. 
    But this produces the ever present mating line!
    15. f5... white wants to solve the problem of the light bishop forking attack 
    16. Qxf5 Rf8      white has solved the problem but now white has an even bigger problem on the f-file!
    17. Qxh7?? Bh6!! 18. Qxe7 Bxe4#

    Fun eh! A duel bishop mate. A special tactic because of a unique Chess960 SP.

    Wednesday, March 2, 2011

    K+2P v K+3P

    I'm not the best on endgame theory and I found myself on the loosing end of this endgame. I do know about "the opposition" theory etc. However here is an endgame position that has me totally stumped me. Why is it that white is guaranteed to loose in this position? (no chance of a draw).

    White to move

    The reason I ask is that there should be at least a chance of a draw shouldn't there? After all white's king is in the area of the queening squares f/g/h. From my initial look at it, black is guaranteed to get at least one passed pawn, but why fundamentally cannot white somehow escape with a draw? I'll try to work out the reason thanks to the trusty endgame book I've been using for years "A Guide To Chess Endings" (Euwe DW, Hooper D).

    All is not lost for Chess960....I hope

    After the disappointed that Mainz has been cancelled, I thought it was over for Chess960. But a quick search using Google blogger search, reveals these new links:

    A new chess960 server yeah!:

    Some evidence of chess960 club play in the USA!:

    A chess competition by the amazing Armenians:

    The Annex Chess Club in Canada:

    And of course Mark Weeks great Chess960 blog that is still going hopefully for a lot longer:

    Keep it up people! Chess960 is wonderful. There is no need to reduce the 960 start positions down to a smaller subset. Just play a couple of thousand games of Chess960 as it is and you will begin to see the patterns in the start concept a bit like a scientist! You are an explorer, curious, learning, inquisitive, enjoying the experience of loosing as much as winning.

    Don't dumb-down Chess960! We need a sense of inquiry in Chess like we had a couple of hundred years ago. Don't simplify it down to Chess60, Chess48, Chess10 chess whatever. Leave it the way it is and enjoy Bobby Fischer's great gift to chess! There are hundreds of books that could be written into the concepts and patterns of Chess960 and so much to learn and enjoy.

    Chess960: SP118

    Hi there
    Playing the Chess960 opening produces some really interesting stuff a bit like walking through a jungle and picking up some ten legged critter you have never seen before. Here is one of those oddities!

    SP118 contains in it some immediate features that can produce some odd situations. The features are:
    1. The central rook is undefended
    2. The enemy can quickly deploy a piece to attack the undefended rook
    3. If this feature is not seen, it can quickly disturb a harmonious development
    Chess 960-Position 118
    1. e4?! d5!?

    2. exd5 Rxd5
    3. Bc4.... White thinks that they have gained a tempo attacking the rook but:
    3. ....Rd8!? Check this out:

    In this situation Rd8!? has not wasted a tempo for black! On surface glance it appears that white is more developed. After all, white has the move, one of his minor pieces is developed and black has neither. The amazing thing is white has the tempo, but black has all the initiative. Check out all these primary lines that show what I mean:

    {Deep Rybka 4 960 w32:}
    4. d4 c5 black threatens d4
    5. Nb3 Nb6 and black forces white's bishop to retreat to e2 wasting a tempo
    6. Be2 cxd4
    7. Rxd4?! RxR
    8. NxR e5 with initiative
    9. Nb3 Bf5 and black has just totally dictated despite that nothing serious has emerged

    {Deep Rybka 4 960 w32:}
    4. Nf3 Nb6
    5. Be2 c5
    6. Nb3 e5 and black has all the initiative and is better developed

    [{Deep Rybka 4 960 w32:}
    4. Nb3 Nb6
    5. Be2 c5
    6. d4 cxd4
    7. O-O e5 and white has a counter attack chance but is down a pawn

    4. O-O Nb6
    5. Be2 e5
    6. d3 c5
    7. Nf3 Bf5
    8. Nb3 Nf6 and black is better because they have the initiative

    4. c3 c5
    5. Nf3 Nb6
    6. Be2 e5
    7. d4 exd4
    8. cxd4 c4 and black's c4 pawn is stable and disrupts white in a few ways

    The moral is that in the chess960 opening, it is wise to study whether a rook is undefended, and if so, can it be quickly attacked? Just another simple check to add to the list. These observations become a habit and soon you won't have to think much once they are seen. And in Chess960, don't assume that moving a rook twice in the opening is always bad!