Chess960: Car crash openings part 1

A Car Crash Opening is any opening where the game is over before the tenth move because the loosing side was asleep at the wheel - enjoy the ride!

Here's a thread on some of the totally trashed and ridiculously bad Chess960 openings that I have played so many times! Embarrassing but I save face by saying that they were 15 minute games where white's timer was ticking down immediately at the start of the game and so I was under some pressure (as white should be!). I'll be using terms from Naming the bishops and Naming the knights because I find it really useful to put a name to the particular characteristics that the minor pieces have in 960.

SP167: How white should NOT play the opening!

1. d4...g5 Ok d4 is not too bad opening up a line for the archbishop on c1 but it puts white on the defensive straight away. Black's g5 has startled me although it shouldn't have. Immediately black is attacking the undefended pawn. Black's g5 is a great move because now white cannot develop the suffragan bishop via g3 without opening up all the light colored squares on the g-side that black's archbishop can quickly and effectively exploit.

2. e3?! ... Qh3!? Now e3 is incredibly timid and a structurally bad move and any plans white had for the center are over. What was I thinking? I thought it would be good to release the queen on f1. I didn't want to block the long diagonal with e4. Black has startled me with the pestering annoyance to capture h2 but it is just a smoke screen. I should have listened to my own advise and simply not worried about h2. If black takes it, they will have wasted a lot of tempo and their own h-pawn will be weak in any case because the h-file opens. Not only that but their queen would be over-committed and exposed. White can castle c-side and so it really is no big deal. The essential issue is that what has happened to white's initiative?

3. g3?! ... Nb6 And now is the coming of the car wreck for white. The central feature is that white has setup a terrible hedgehog pawn structure which creates a huge number of weaknesses that blacks archbishop and d8 knight can exploit VERY quickly. Black can play the central pawns out with initiative and white has nothing. They can't even play the pawn out to e4 because they wasted the time moving it to e3.
4. Be4?! ... d5 White decides that they are going to actively deploy the bishop and that transferring it to d3 looks fine because it doesn't impede the knights in any way and there are lots of interesting lines for at least one bishop. But the stupidity is that it only encourages black to play d5 and causes white to waste a tempo on realigning the suffragan bishop.
5. Bd3 ...Nc6 Essentially the game is already over for white. Aside from all the other problems, now the d3 bishop block's the queen's active release. The point is that there are not and never were going to be any useful targets that white could hit on the f1-a6 light diagonal.

6. Nc3? ... Bh3 White thinks that it is no big deal if black biffs the queen with Bh3 because white can play Qe2. White is oblivious to the arrowed lines of attack that black has, particularly black's c6 knight and black's central pawns.

7. Qe2 ... e5 Black now threatens to put his knight on e5 at some stage.

8. Nb3 ... O-O-O Black is in total control of the center and also the g-side wing.

9. dxe5 ... Nxe5 and now black's knight threatens to remove the light coloured bishop which leaves all the light squares on the g-side undefended, or instead black's e5 knight sits where it is threatening a fork on f3.

Aaaggh. Crash! (Good fun)

Chess960: SP051 and Svidler

Had a look at Mark's blog on Leko vs Svidler at Mainz during the first decade of this century: http://chess960frc.blogspot.com/2011/04/svidler-leko-mainz-2003.html
The game that Mark highlighted is a really nice example of Chess960 opening theory. I actually think that Svidler broke one of the fundamental laws of general opening play:

1. "Try to move only two pawns if possible before starting to develop at least a knight."

In Chess960 the rule is modified to:

1. "Try to make only two optional pawns moves (possibly three?!) before starting to develop at least a knight. The exception is when there are necessary pawn moves that need to be played that also need to be played by the enemy"

Amazingly this is the game that I needed to find that proves me wrong that making more than three pawn moves in the Chess960 opening is bad! Normally it is, but this game is an exception (I think!?). Notice that it contains all the ideas of rule (1) above including the necessary move e3/e6. If you think of the g3/g6 and the b3/b6 pawn moves as being necessary moves (because both players need to exercise them), then this opening actually does not violate any of the rules of standard chess opening theory because only two pawn moves were made that were optional!

Another interesting thought is that if white plays b3 to release a bishop, can you consider the alternative pawn move b4 as necessary or optional? Does that mean that b4 would count as one of the two optional pawn moves in rule (1) of the opening but b3 would not?

SP051: Svidler's alternative line that would have accomplished what he was after

The line that is showing is the alternative that Svidler could have played. It is essentially the same idea as his, except that it does not break rule (1) above! Note this is a pure Suffragan bishops opening (Naming the Bishop Pairs) and thus is only ever seen in 6% of all Chess960 games. In other words like I said in (Insider info on the Bishop Pairs) forget about the Suffragan games. All of us Chess960 players should be concentrating on the Patriarch Bishop varieties because they occur roughly 50% of the time.

On another topic, Mark mentions that Svidler hung up his Chess960 gloves during the 2006 Mainz tournament after loosing to Aronian. What I think happened is that if you take a look at Svidler's ratings during that period, you can see that all through the period where Svidler is actively playing Chess960 as well as Chess, his rating was rising:

http://ratings.fide.com/id.phtml?event=4102142

At the very first noticeable drop in his ratings that roughly coincided with Mainz in 2006 and the loss to Aronian, Svidler probably had to make a choice (or the choice was made for him). He could sense that his rating peak might not sustain, and so what has to go first? Of course Chess960 has to drop so that the super GM can avoid all possible distractions that impede the massive research effort that is required to find novelties in the standard Chess opening systems at the very top of the chess ladder. Svidler's ratings never regained their momentum after quitting Chess960 for a variety of reasons probably including being out-competed in the the vital off-the-board opening pre-arrangement research department.

Nakamura is now in a similar position that Svidler was in 2006! The two should get together and have a chat! Like I've said before, I believe that Nakamura's Chess has been greatly benefited by Chess960, because it has facilitated Nakamura to think up creative novelties in the Chess opening system that he could do on his own without needing big teams and powerful computers. However the burden of being near the top of the Chess ladder is that you now need the teams of assistance and the powerful computers to continue to win.

The situation is nuts, but what can I say? That period of collaboration between Magnus and Kasparov is a classic case. Both of those greats knew that Magnus needed a little something extra to make it to the very top. What was that? It was getting access to Garry K's very private opening database of novelties, that is what it was! After getting access to that, the intense collaboration between them was no longer needed.

It's not the novelties themselves that are the issue! Exactly like Bobby Fischer said on his flight out of Japanese detention, it is that the novelties pre-arrange midgame scenarios that the novelty player understands more deeply. So there is an inevitable flow on effect that filters through the rest of the game after the novelty is executed.

The teams of assistants aren't needed to find the novelties! There are specialised programs that run on computers that find the novelties automatically twenty four hours a day (talk to Topalov). The teams of assistants provide the super GM not with the global context of where the current opening trend in chess is at now, but where it is likely to be (something a computer cannot do). Super GM's must also forget certain opening theories in order to remember the one's that his assistants are telling him will be in vogue and that his computer is finding as well. So you have a continual renewal of old material as well as the occasional computer discovered novelty.

Whatever.

Chess960: Naming the Knight Pairs

There are a number of structural elements to a Chess960 position and a number of different ways of looking at it:

1. The King with respect to the Rooks
2. The Independent Queen
3. How many times each pawn is defended (SP518: 11144111)
4. The Bishop Pairs (Insider info on the Bishops)
5. The Knights Pairs (Insider info on the Knights)

For me personally, the last two regarding the minor pairs are pretty important. They hold up the character of the starting position to me. Check out the insider info links above as well as the respective naming entries in Wikipedia:

Wiki on the Knight

Each name has a historical analog to their Chess960 character (the Knight Orders).

Note that if you want to win at Chess960, better brush up on the theory behind the Military Knights and the Chivalry Knights. Together they account for roughly 50% of all games...

The Knight Pair Configurations and their Naming:

The Military Knights (except on DE)

• Can work together or rapidly split apart
• Can support each other in enemy territory
• Complicated interaction confusing the enemy
• Able king protectors forming a defensive wall
• Need to be pawn protected for max effect
• Have two blind squares directly in front
• Found in 216 SP's (23%)

The Chivalry Knights (1 or 3 squares apart)

• Exist in two forms N.N or N...N
• N.N->Incredibly cooperative attackers
• N.N->Morph in complex and confusing ways
• N.N->Attack focused
• N.N->Flexible but unbalanced defenders
• N...N->Surprise attackers that can defend
• N...N->quickly morphs into N.N
• Have a significant pawn forking weak spot
• Tend to compete with each other for the same color squares
• Found in 320 SP's in both varieties (33%)

The Monastic Knights

• Dual colored ambiguous attackers
• Can morph into military or chivalry roles
• Can deploy into wing attacks
• Can cover the center more in shifted forms
• Can form a strong defensive wall for a king
• Found in 180 SP's (19%)

The Hereditary Knights

• Deploy to traditional positions
• Flexible balanced attackers
• Flexible balanced defenders
• Found in 144 SP's (15% with the DE form)

The Ceremonial Knights (5 or 6 squares apart)

• May become ineffective in the corners
• Might deploy into rapid attack on a wing
• Important home territory defenders
• May come out in ceremony late in the game
• Rarely co-operate in tandem
• Found in 100 SP's (10%)

Chess960: Naming the Bishops Pairs

There are a number of structural elements to a Chess960 position and a number of different ways of looking at it:

1. The King with respect to the Rooks
2. The Independent Queen
3. How many times each pawn is defended (SP518: 11144111)
4. The Bishop Pairs (Insider info on the Bishops)
5. The Knights Pairs (Insider info on the Knights)

For me personally, the last two regarding the minor pairs are pretty important. They hold up the character of the starting position simply because they are the most numerous. Check out the insider info links above as well as the respective naming entries in Wikipedia:

Wiki on the Bishop

Each name has a historical analog to their chess960 character. Do please excuse me if you find the "male oriented" namings of the bishops offensive but I must mention that to be fair, the most powerful piece on the board is a female and so I think the namings balance out :-)

Note that if you want to win at Chess960, better brush up on the theory behind the Patriarch variants below. They account for roughly 50% of all games....

The Bishop Pair Configurations and their Naming:

The Primary Patriarchs (AB-GH files)

• Autocratic dual color long range controllers
• Work in tandem powerfully sweeping the board
• Often remain on their original squares
• Dangerous and intimidating from long range
• Can morph into dynamic king protectors
• Weak against rank and file attack
• Have two undefended squares in front of them
  displaced by one rank
• Found in 120 SP's (13%)

The Patriarch Bishops (BC-FG files)

• Autocratic tandem control of two diagonal
  with more versatility than primaries
• Sometimes remain on their original squares
• Work together but can split
• Can morph their influence onto both wings
• Have two undefended squares in front
  of them displaced by one rank
• Found in 120 SP's (13%)

The Minor Patriarch Bishops (CD-EF files)

• Subtler autocratic control than the other forms
• Can work together but also likely to split
• Able to side skirt the central pawn formation
• Can adopt the roles of the other patriarch's
• Can be deployed ambiguously like the Titulars
• Deployable in creative king protection ideas
• Found in 120 SP's (13%)

The Archbishops

• The standard bishop separation distance
• Have the maximum scope possible
• When shifted on the back rank they become
  independently operating bishops difficult
  to formalize their roles
• Found in 300 SP's (31% with shifted forms)

The Metropolitan Bishops

• Independent operators with mirrored
  possibilities and restrictions
• Cannot work the long diagonals easily
• Versatile but usually one of the duo
  is tricky to coordinate
• Share the same scope simplifying planning
• Shifted forms transform into clearly
  defined roles as patriarchs and archbishops
• Found in 180 SP's (19% with shifted forms)

The Suffragan Bishops

• Cannot hide their intention
• The secret is not how they are deployed
  but when
• Maximum long range effect
• Easily neutralized unless their is a
  willingness to prevent trading
• Found in 60 SP's (6%)

The Titular Bishops

• The philosophical bishops :-)
• Very subtle
• Ambiguous in how they operate
• Tricky to coordinate with the other pieces
• Found in only 60 SP's (6%)

Chess960: Statistics and insider information on bishops

Ok yesterday I thought it was a pretty good finding to have worked out the deal with knights in Chess960. (Insider information on the Knight Pairs).
But today I really blew my simple little mind on these statistics on the bishop pairs. Check out Naming the Bishop Pairs as well.

Bishops together:

1. Bishops 0 squares apart = 420 "The Patriarch Bishops"
2. Bishops 2 squares apart = 300 "The Archbishops"
3. Bishops 4 squares apart = 180 "The Metropolitan Bishops"
4. Bishops 6 squares apart = 060 "The Suffragan Bishops"

What this means is that the Patriarch Bishops are getting onto almost half of all Chess960 games! Not only that, but the Patriarch's are evenly distributed throughout the 960 positions. The traditional arrangement of the Archbishops is roughly one third of all games. So with the staggering Patriarch case, it needs to be broken down further:

 The Patriarch Bishops (BB):

1. Bishops (BB) on AB/GH   = 120 "The Primary Patriarch's"
2. Bishops (BB) on BC/FG   = 120 "The Middle Patriarch's"
3. Bishops (BB) on CD/EF   = 120 "The Minor Patriarch's"
4. Bishops (BB) on DE      = 060 "The Titular Patriarch's"

Why these weird names for the bishop combinations then??? It's quite simple really! Wikipedia defines each of them (Wiki Bishops). If you read the definitions, you will see why I called them what I have. I wanted to combine their Chess960 flavor with the characteristics of their historical Wiki counterparts. Do please excuse me if you find the "male oriented" namings of the Patriarch bishops offensive but I must mention that to be fair, the most powerful piece on the board is a female and so I think the namings balance out :-)

The Pure Forms of the Bishop Pairs
Here are the pure forms, but of some mixing of roles takes place as well:

• The Patriarch's: "Highest Ranking; Autocratic Authority". In terms of Chess960, these bishop pairs dominate the whole feel of the starting position in nearly half of all Chess960 games. The primary and middle forms of the Patriarch's need actively deployed wing pawn formations to operate that must not allow enemy knights to infiltrate and exchange away half of their dual color dominance. As the patriarch bishops move from primary to middle to minor bishops in the various SP's, they become more likely to split apart and thus more difficult to target.
• The Primary Patriarch's can work powerfully together to control both colors along only one skewed diagonal trajectory. They are the patriarch's that wield the most autocratic power along the long dual colored diagonals almost instantly and often do not need to move from their seat of power. The primaries have a dual role of active attack but also active defense. They can morph into being staunch protectors of the king by simply transposing a single square inwards becoming important king shields while retaining their long diagonal authority. 
• The Middle Patriarch's can quickly adopt a similar autocratic authority as the primaries. They can also split apart but are more difficult to justify such splits. Their main weakness is an inability to form a protective shield in creative king defense roles. The sometimes remain seated on their original square, working at a distance like the primaries.
• The Minor Patriarch's are subtler in how they exercise their authority. They can shift to the long diagonal roles of the primary patriarch's but can also morph into solid king protectors. They have more access to both wings of the chess board and are thus more versatile and ambiguous in operation.
• The Archbishops: "Importance in history; Also Metropolitan". In terms of Chess960, these clergy are your traditional distance apart as per Chess. They can morph into various diagonal lines of controls and are very flexible and versatile (less so in their shifted forms).
• The Metropolitans: "In charge of a province". In terms of Chess960, these bishops are far enough apart that they exhibit independent authority over various diagonal combinations but are not as versatile and can be problematic to coordinate within the whole framework of the development phase unless properly thought out. In their shifted forms they morph into complex combinations of the Patriarch's and the Suffragan bishops.
• The Suffragans: "Subordinate to the Metropolitans; In charge of a specific area". In Chess960, these bishops are the subordinate bishops because they are on the long diagonal and control only one specific long diagonal. In their pure form they never get that control alone on the diagonal because they must face their arch rival on the same line. They struggle to be versatile but can be very powerful if not counter-balanced along their color of influence.
• The Titulars: "A bishop without a diocese". Yes the amazing Titulars! If you have ever played Chess960 with two BB bishops in the center of the board, you will know immediately what I'm talking about. The two bishops have no defined role to work in. They have to angle about until finally they find out what their purpose is! The ambiguity of deployment is weird but personally I wish there were more SP's with them. Luckily the minor Patriarch bishops have characteristics in common with these philosophical bishops and so there are quite a few SP's with this level of bishop ambiguity.

Ok so here is a bit of insider information then!:

Because of the phenomenon of the Patriarch Bishops in the Chess960 SP's, if you want to win at Chess960, practice how to deploy your Patriarchs over all the other pairings. In approaching fifty percent of all Chess960 games the tactics around them will dominate and you will have the edge!

Chess960: Statistics and insider information on knights

OK with a way to calculate SP stats via: Chess960: Start position searching, I've done a quick collection of statistics of the quantity of various knight related setups. Also see Naming the Knight Pairs and Insider information on the Bishop Pairs

Knights Together:

1. Knights adjacent except on DE = 216 "The Military Knights"
2. Knights 1 or 3 squares apart  = 320 "The Chivalry Knights"
3. Knights 2 squares apart       = 180 "The Monastic Knights"
4. Knights 4 squares apart       = 144 "The Hereditary Knights"
5. Knights 5 or 6 squares apart  = 100 "The Ceremonial Knights"

These are the statistics I'm most interested in at the moment for particular reasons of being able to judge the "weight" of how certain chess technical skills will count in Chess960. The suspicion is that positions that contain knights next to each other or one square apart that are on the files ABC or FGH tend to focus skills on recognizing knight tactics at a very thorough level. This is because the knights work together in complicated ways that us humans have difficulty appreciating. If you check out the statistics on the different arrangements, you get the striking result that the military knight and chivalry knight combinations count for over half of all Chess960 start positions.

Why these weird names for the knight combinations then??? It's quite simple really! Wikipedia defines each of them (Wiki Knights). If you read the definitions, you will see why I called them what I have. I wanted to combine their Chess960 flavor with the characteristics of their historical Wiki counterparts :

• The Military Order: "Crusaders; Secular-Holy". In terms of Chess960, these knight pairs can work together or rapidly split apart. They may deploy early on one wing or go into attack later in the game. Their interaction is complicated and confuse the enemy. They can also form a wall like defense of the king.  When deployed immediately ahead of a line of pawns, they form an imposing structure that is very like their heavily armored equivalents on the medieval battle field.
• The Chivalry Knights: "Idealized; Romanticized". In terms of Chess960 these rider's have a variety of abilities. In their N.N form they are attack prioritized working together in very cooperative ways because they play out on the same colors. In their N...N form they are still on the same colors and can morph into N.N or remain more flexible both cooperating in attack and working to a lesser degree on general defense. 
• The Monastic Order: "Holy Land Crusaders". In terms of Chess960, these horsemen fight on both colors like the bishops and can morph into Chivalry or Military roles. They can deploy on the wings, assist the center and form a strong defensive wall to protect the king.
• The Hereditary Knights: "Part of the Nobility System". In Chess960, these are the horse-masters that we are most familiar with because of their historical equivalence with Chess. They deploy to traditional squares but in Chess960 they are joined by the knight pair directly in the center on the D and E files. They are balanced attackers and defenders and can cover the wings and the center.
• The Ceremonial Knights: "Symbolic; Purely Honorific; Prestige". In Chess960 these knights might end up being ineffective, deploy ceremonially later in the game but can also be critical home territory defenders. They can rapidly deploy up one wing if necessary. They rarely exist in their pure ceremonial role and infrequently team up together.

Ok so here is a bit of insider information then!:

Because of the phenomenon of the knights in the Chess960 SP's, if you want to win at Chess960, practice your knight skills particularly the start positions with the Chivalry and Military Orders. In roughly fifty percent of all Chess960 games the tactics centered around them will dominate and you will have the edge!

To carry this thought a bit further, here are some predictions:

1. Since we know that Judit Polgar is most fond of the knights in Chess, I'll take a bet that if super GM's played Chess960, Judit would do better at Chess960 than she would at Chess. Now because she is a woman, this will be another reason why the super GM's are not mass defecting to Chess960 (kidding) :-)
2. Another reason why your average chess player is not playing Chess960 is because they have touched it and gotten burnt and then make excuses why they think Chess960 is bad. But the real reason is that most of us average patzer level players are weakest when it comes to visualizing the movement of the knights, particularly in retrograde movements of the knights. We are weak at visualizing the forks that the knights can threaten as well.
3. In another Millennia when mass-consumerism has run it's course and we have stopped rushing around buying things and finally have time to contemplate once more, people will discover that within the entire set of 960 positions, that there are many series of smaller Chess960 positional sets with commonalities, flavors and characteristics. For example the positions from roughly 70-95 are studies in knight tactics. Chess960 will become a huge integrated set of themes and variations like a song with 960 verses.

Chess960: Generating start position statistics

Note: For the moment Google has prevented me from making this Chess960 Google document publicly visible. Contact me if you want a copy.....

Here is a post of one way to calculate how many positions of particular configurations exist in the total set of Chess960 starting positions. I wanted to know how many SP's have bishops in their traditional locations. Here's what I did courtesy of Mark Week's Chess960 Position Generator:

1. Used Mark's SP text list and published it as a Google document:
   Chess960 SP List As Google Doc
2. Incredibly it looks like Google Docs text search facility doesn't allow searching with regular expressions although I'm sure at one point it did.
3. So I downloaded the document into Open Office and did a regular expression search in it.
4. Ctrl-F and make sure "regular expressions" is ticked under more options
5. Typed in [a-z]{2}b..b[a-z]{2}
6. Clicked on "Replace All" to show me how many replacements were made (60)
7. Clicked "undo" with the search box still active and repeated for more searches

So there are 60 positions that have bishops on the c and f files. The question is what would have been the logical way to infer it?