Saturday, July 21, 2012

Non-Random Chess960 Trial

Here is the database of our computer assisted games so far, including comments:

Mark Weeks our main blogger on Chess960, posted an idea for selecting Chess960 start positions non-randomly by player agreement. Here is his post Fischer-Bronstein Chess960. Here is a proposal to Mark Weeks or indeed any other fan of Chess960:
  1. We play through a number Chess960 openings up to move 10 (was 15) and stop the game then
  2. We play through the games via comments on this blog computer assisted or not, swapping colors each game and I will publish the start position and the moves here on this website.
Interested Mark or anyone? 

The benefit of doing this is we get to see how diverse the Chess960 positions are, and play through any particular start positions that interest us. I am trying to find out what opening ideas black has in some difficult starts that can redress any imbalance there might be. We cherry pick particular SP's to trial as we continue to explore Chess960.

We play the SP like we are playing a real game and we use computers where necessary. We stop the game as soon as it has evolved into a situation where we feel that black's chances are sufficient to conclude that the SP is reasonably well balanced. So games usually don't go for more than twenty moves.

If something goes wrong and black looks loosing, afterwards we will try to understand what went wrong. Since we are not playing to win a game, we freely disclose our plans to our opponent and the broader community as we go. This has the benefit of also being useful as move comments for future reference.

So the more people to help all the better! There are no move time limits and sometimes it takes days before we play the next move.We are building a database of these trial runs as we go and I try to summarize what happened for each trial in my blog, depending on how much time I have.

What I personally have noticed is that playing detailed Chess960 openings as we do here and sharing our ideas, helps improve my play more generally.



  1. I'm in, although 15 moves is too many. Five is probably too few, so how about ten? Comments to your blog would be fine with me. I'll give you White the first game. - Mark

  2. Ok Mark thanks ten moves it is no time limit. Feel free to amend any form of notation you think is good for a game that is actually commencing prior to move 1!

    [Event "Fischer-Bronstein Chess960 Trial"]
    [Site ""]
    [Date "29-07-2012"]
    [Round "1"]
    [White "Chess960 nerd"]
    [Black "Traditionalist"]
    [Result "*"]
    [Variant "Chess 960"]
    [Variation "Chess 960"]
    [FEN "?"]
    [SetUp "1"]

    -1. Bb1 {breaking tradition with the potential for early activity}

  3. I don't like to comment on games on progress, but I'll make an exception here since this is so new. I place Ba1, creating Bishops that rake the Kingside (h-side). - Mark

    ♗ ♗ * * * * * *

  4. Agree on comments but perhaps stick with them in the pre-game phase. You have already surprised me, considering that you would think that black is in trouble here! So my next move is just simple and played without thought:
    -1. Bb1 ... Ba8
    0. Qc1 {aiming to create a diagonal battery if O-O}

  5. One of the principles of chess960 is that there are no *bad* positions. So why not create one that looks difficult? I place the Knights on e1 and f1 to protect the squares attacked by the Bishops. That creates SP384 BBQRNNKR. If you agree, the first move is yours. Don't expect me to play too quickly. One day per move is an average rhythm for me.

    BTW, you (subconsciously?) created a BBQ (aka barbecue) position. Is this some kind of an Australian tradition? - Mark

    ♗ ♗ ♕ ♖ ♘ ♘ ♔ ♖

  6. No expectations on time. I'm starting to think that I am the traditionalist role player here and you are playing the Chess960 nerd role! Interestingly from a traditionalist perspective I guess it is very natural to want to put down the BBQ (barbecue) setup because it "looks" so completely over powered for white?

    -1. Bb1 ... Bb8
    0. Qc1 ... Ne8f8
    1. c4 {torn between b4/c4/d4}

  7. Nerds make the world go 'round.

    1.c4 c5

    There are ten BBQ positions (plus ten twins of those positions, *****QBB, with similar characteristics). If some/all of these positions were bad, someone would surely have spotted this by now. - Mark

  8. Here's the PGN I'm working with:-

    [Event "Non-random Chess960 trial game"]
    [Site ""]
    [Date "2012.07.29"]
    [Round "?"]
    [White "HarryO"]
    [Black "bemweeks"]
    [Result "*"]
    [Annotator "SP384"]
    [Variant "chess960"]
    [SetUp "1"]
    [FEN "bbqrnnkr/pppppppp/8/8/8/8/PPPPPPPP/BBQRNNKR w KQkq - 0 1"]

    1.c4 c5 2.-- *

  9. Wow only 10 barbecue positions with mirrors. I've had a bit of a look at some of the symmetrical lines after 1.c4 c5 and feel that my opening choice was somehow inadequate and so have come up with an idea:

    1.c4 c5 2.b4 *

  10. I hereby call 2.b4 the "barbecue gambit"

  11. I second that. For the family of BBQ***** SPs and twins, the 1.c4 c5 2.b4 (and its equivalent in *****QBB) is the 'Barbecue Gambit'. Since the best way to refute a gambit is to accept it, I play...

    1.c4 c5 2.b4 cxb4

    ...The extreme case in this family must be BBQNNRKR. I wonder what the CCRL says about it. - Mark

  12. I looked up "barbecue chess960" on the web to find the now famous Chess960 quote "korean barbecue!?" mweeks (2011)

    SP000 (BBQNNRKR) is the very first position in the Chess960 series and features the "barbecue extreme opening". I only have Chessbase light and so have limited stats: White 43% Black %41 draw 16% n=46

    Ok my move continuing on the "theme" of the gambit which I insist on displaying whether it is good or bad:
    1.c4 c5 2.b4 cxb4 3.d4

  13. I thought BBQNNRKR looked familiar, but didn't bother to look it up. The moniker 'Extreme Barbecue Opening' ('Xtreme' is even flashier) sounds OK to me. I call all *****RKR positions (or their twins) examples of 'extreme' castling, because the King has to travel so far to castle to the long side...

    2.b4 cxb4 3.d4 Ng6

    ...Develops a piece to its natural square and prepares castling O-O. - Mark

  14. Moniker's are really useful in Chess960 to "chunk" memorize a certain familiarity with the themes, if not the specifics. I think it is wonderful that there are just ten BBQ's and twins because it just proves to me that there is a wealth of memorization opportunities in Chess960 except that memories are more contextual familiarity based memories.

    The next move is actually very interesting, both of us sequentially moving the same knight but to different spots.

    2.b4 cxb4 3.d4 Ng6 4.Ne3 *
    ...develops a piece to control the center and c4 without impeding the other pieces I hope. - Harry

  15. In fact, for any combination of Bishops & Queen, there are ten possible positions for the Knights (5 * 4 / 2). I had many interesting possibilites for the next move...

    3.d4 Ng6 4.Ne3 Nf6

    ...Develops a piece to a good square, prepares O-O-O (although not likely to be played), and delays Queenside options to see White's next move. - Mark

  16. 3.d4 Ng6 4.Ne3 Nf6 5.Qd2

    ...thematically committed to the gambit line but prepares the c-file and threatens to recapture the gambit pawn.

  17. The Queenside Pawn breaks don't seem to work, so I'll stick to the simple...

    4.Ne3 Nf6 5.Qd2 b6

    ...Develops a piece and protects against several threats. - Mark

  18. I can't calculate that far even with a computer but the whole gambit depends on ...a5 being a wasted move.

    4.Ne3 Nf6 5.Qd2 b6 6.d5

    ...Clears the long diagonal, blocks the long diagonal, continues the theme of the gambit and leads to some truly spectacular variations!

  19. I should have said, the gambit depends on the "premise" that ...a5 is a wasted move!

  20. I had a dental emergency today, so there was no time for a move in this very challenging position. I hope to be back in the saddle tomorrow. - Mark

  21. Ouch, both in the mouth and the pocket. Wishing you the best.

  22. I'm afraid the bulk of the physical and financial suffering is yet to come. Back to more pleasant matters...

    5.Qd2 b6 6.d5 a5

    ...You don't get unlimited time to recapture the Pawn and it's a useful waiting move. Most importantly, I couldn't find anything better. - Mark

  23. Now come some fantastic formations in some of the variations that may never see the light of day. This is more like astronomy than chess.

    5.Qd2 b6 6.d5 a5 7.Nf3

    ...develops the knight, prevents 7...Be5 and relies on the theory that black doesn't have a good move any more.

  24. A little ray of sunshine for chess960. There will be a workshop on it at the Mind-sports Olympiad in London:

  25. Black doesn't have a good move any more? What about this?...

    6.d5 a5 7.Nf3 Bf4

    ...Following your link, chess960 is right up there with 'African Quincala'. What we need is a workshop at the London Chess Classic later this year. - Mark

  26. Yes, but now you will be powerless to my secret weapon! Or at least I think you will would be a great irony if you were to castle queen-side to escape it!

    6.d5 a5 7.Nf3 Bf4 8.h4

    ...begins the "Barbecue Flank Attack"

  27. You know, Harry, I'm 'subscribed' to this post, so every time you alter your comment I receive a separate message. Here's what I have for your last move, 8.h4...

    23:30 'The charge of the little guy begins! White deploys his main weapon, the humble little edge pawn soldier with a grand vision of world conquest.'

    23:47 'The barbecue flank attack begins and the house will burn! or will it? :) White deploys his main weapon.'

    06:10 'Yes, but now you will be powerless to my secret weapon! Or at least I think you will would be a great irony if you were to castle queen-side to escape it! 8.h4 begins the "Barbecue Flank Attack"' a few minor corrections in other messages. This is all creative stuff, but have pity on me! I'm drowning in email as it is. - Mark (my move will follow)

  28. Need to stop a little guy? No need to send the cavalry; just send another little guy...

    7.Nf3 Bf4 8.h4 h5

    ...The possibility of the h2-h4 advance was the reason Black hasn't castled. With a clearer situation on the Kingside, the future safety of Black's King can be judged more circumspectly. - Mark

  29. Sorry Mark had no idea Google is sending you emails! I've been stumbling around this UI looking for an edit button for days.

    7.Nf3 Bf4 8.h4 h5 9. Bxg6

    ...smashing the last chance to hide the king and creating a symmetrical second weakness on g6 to match b6.

    If you want to speed it up, here is my next move:
    If 9...fxg6 then 10.Ne5

    NOTE: I will probably change my mind but will not edit it

  30. No big deal with the multiple emails, but I thought you would like to know that your changes are broadcast...

    8.h4 h5 9.Bxg6 fxg6 10.Ne5 Rh6

    ...Black continues to make White work to recapture the Pawn. The Black King is not in any particular danger and ...b5 threatens to smash White's center.

    That's ten moves, Harry, which is the number we agreed on. I wouldn't mind continuing, but the comments mechanism is a bit clumsy for extended play. Do you know of a site that allows chess960 challenges with a specified SP? If so, we can continue there by repeating the moves we played here.

    As for the non-random aspect, it was an excellent experiment from which I learned several things. First, the BBQ family of SPs is crucial to the soundness of chess960. Second, specifying the start squares requires more thinking about the pros & cons of pieces starting on certain squares. In the random version, I take for granted what I'm given and work from there. In the non-random version, I find myself going back to the initial square selection and comparing the alternatives. Third, you're a very creative chess960 player (not that I had any doubt).

    If you would like another game, I'll take White and place my Bishop on d1, a central file. I suggest you create a new post for the second game, because this one is long enough as it is. When ready I would appreciate your making a comment here so I don't have to look every few hours to see your response. - Mark

  31. Hope you haven't been spammed again. I've been pretty busy on my blog tonight. I've created a couple of new posts on the non-random trial, and here is the post for the next trial:

    Thanks again

  32. Noted & subscribed to the other post. - Mark