Tuesday, January 1, 2013

Non-Random Chess960 Trial Game 7: SP749

The PGN database for these trials is here.

In this run through of SP749, the engines and CCRL database are indicating that white has a significant start advantage. Playing as white, I tried an early development of the queen. The idea did not work out all that well. Yes the queen can come out early in this SP because there are many squares that cannot be attacked. However, she is not all that effective unless well supported by other pieces. Here is a key moment:

SP749 trial: key moment

This is a key moment because white has pushed back black's knight to the back row, and I assumed that yet another move where black does not advance forward and develop, would have to be crippling. However, black's move 7....Qe8 simultaneously defending d7, attacking e5 and indirectly controlling the e-file was amazingly effective. White's queen is now without a target and she blocks the g1 bishop. White could play Nf3 but that would block the d1 bishop.

In this trial, I never solved the problem about what to do with the d1 bishop. Bg4 is obvious, but more difficult to achieve than I thought, since black has counter-play; the very thing I was trying to stop by the hyper rushed queen attack. Note that black's equivalent bishop on d8 can get to g5 but in black's case, there is no knight that impedes it's path there.

Good luck to the future generations of Chess960 players playing this SP as white or black. Is there a forced win for white? I think black could well be OK in this SP for at least a draw, but it is going to require memorization of the basic concepts, otherwise in practical play black will run out of time on the clock.

30 comments:

  1. Same move as last trial except inverse.

    1.e4

    ...same reasons as last trial

    ReplyDelete
  2. On top of everything else, we had to euthanize the family dog this week. It's always a sad event...

    1.e4 e5

    ...I see no reason to deviate from the plan of the previous game. - Mark

    ReplyDelete
  3. My thoughts go out to you. It's really hard to loose a family member. We euthanized our cat of 19 years and that was really tough. The heeling takes time.

    On another topic, I discovered in my other studies that the win rate for SP482 on CCRL is 81% after 1.c4!

    Add that one to our list of SPs to try!

    1.e4 e5 2.f4

    ...It seems that if white does not continue with the main line, all initiative dissolves away.

    ReplyDelete
  4. SP482 1.c4, noted. How did you discover that?...

    1.e4 e5 2.f4 Ned6

    ...Continuing with the plan from SP408. The different castling situation seems to have little impact on the plan. - Mark

    ReplyDelete
  5. I've been systematically working through 1-960(or 0-959) and am up to SP482. In Fritz-12 I create a 960 opening book and import the CCRL database into it. The stats are automatically created and displayed under the 'opening book' tab. This cannot be done in Chessbase-11 because the opening book software is broken for Chess960. Not sure about the latest version of those software programs.

    1.e4 e5 2.f4 Ned6 3.d3

    ...gives up a tempo but achieves a number of things as sufficient compensation:
    1) Qf4 is ok
    2) d3/c4/c5 attacking knights on b6/d6
    3) develops the center
    4) black will have trouble developing the knights

    The downside is that I intend to play an aggressive pawn formation which could expose my king.

    ReplyDelete
  6. Sounds like an ambitious project on the CCRL data. If you post any results on your blog, I'll be sure to mention them on mine...

    2.f4 Ned6 3.d3 exf4

    ...White's last move was very classy and Black's response looks forced. The next few moves will be critical for Black's survival.

    Things look pretty bad in Australia these days, at least according to the news reports I'm seeing. I trust you're out of harm's way. - Mark

    ReplyDelete
  7. Being a ~2000 elo player, and a chess960 lover, I can and want to play a game with you.

    ReplyDelete
  8. Yes that would be great. I've updated the main post to explain what we are doing in this trial project. If you are still interested, we can pick a fresh start position and try it out.

    Cheers, Harry

    ReplyDelete
  9. Thanks Mark, we have an exit plan in case of bush fire that's the main thing.

    2.f4 Ned6 3.d3 exf4 4.Qxf4

    ...develops the queen and any attacks on her will be somewhat expensive and committal. Best wishes to black here!

    ReplyDelete
  10. I'm curious. In what part of Australia do you live?...

    3.d3 exf4 4.Qxf4 Ne7

    ...Still scrambling to stay in the game. - Mark

    ReplyDelete
  11. On the peri-urban fringes of Brisbane Mark. In Australia we have a culture of sprawling suburbs around huge cities and these suburbs push into the outlying bush areas that are full of highly flammable eucalypt forest - a recipe for bushfire.

    With the guest appearance of another Chess960 fan, I was prompted to do a bit of site maintenance. Here is a dropbox link to the database of our non-random chess960 trial games, that includes our comments after the move.
    http://chess960jungle.blogspot.com.au/2012/07/fischer-bronstein-chess960-research.html

    Sorry for taking so long to move Mark. Most of the time it's not because I'm studying the moves all that deeply, but that I suffer from brain fog in the immense heat of the Australian summer.

    3.d3 exf4 4.Qxf4 Ne7 5.e5

    ...With my queen commited so early there is a great risk that white's attack will fizzle away because of chronic under-development. Never-the-less in this case white will continue to attack, opening lines up and impeding black's own development, allowing for a rapid catch up in development with initiative. - Harry

    ReplyDelete
  12. Take as much time as you need. It's not like there's a time control. I saw a news report last week that said temperatures were up to 45° in some parts of Australia. In Belgium people complain when it hits 35°...

    4.Qxf4 Ne7 5.e5 Ng6

    ...Black has to play this sooner or later. Better to see where the Queen goes before calculating too deeply. - Mark

    ReplyDelete
  13. Well the next move is pretty much a follow on from my last move so I can play it quite quickly.

    4.Qxf4 Ne7 5.e5 Ng6 6.Qd4

    ...Taking aim at the d7 square to produce a weakness. In some SP's like this one, there is no need to withdraw the queen because there are an abnormal amount of poorly defended squares where the queen may sit - Harry

    ReplyDelete
  14. I take your point except I wouldn't call them 'poorly defended squares'. I would say something like 'squares that can't be attacked'...

    5.e5 Ng6 6.Qd4 Nc8

    ...Forced. The Knight retreats to the square recently vacated by the other Knight. This never happens so soon in SP518. - Mark

    ReplyDelete
  15. Yes that is a clearer description thanks.

    Achh I am making all these ugly queen moves!

    The point of this hackneyed play is that in these trials I'm trying to see if there is a way that white can safely declare a won game in less than 20 moves. Even if it turns out that in the future a million players analyse this SP and find out that white has a forced win in 35 moves (for example), it still does not hurt Chess960, because it's only one line among 959 other SP's. What hurts Chess960, is if we find winning lines in under 20 moves.

    Do you agree?

    5.e5 Ng6 6.Qd4 Nc8 7.Nb3

    ...Makes a serious effort to begin opening development proper, but also puts immediate pressure on d7. - Harry

    ReplyDelete
  16. Yes, a forced win for White in any SP would be a serious problem for chess960...

    6.Qd4 Nc8 7.Nb3 Qe8

    ...Counterattacks and pins the e-pawn. - Mark

    ReplyDelete
  17. Well, I had a blind spot in my initial analysis for the strength of your last move.

    6.Qd4 Nc8 7.Nb3 Qe8 8.Qe4

    ...Although extremely reluctant to allow queens to trade after wasting so many tempi on moving her, I honestly can find nothing better than moving her one more time. - Harry

    ReplyDelete
  18. Despite the Queen moves, White isn't lagging in development. Black has spent several moves moving pieces to the back rank...

    7.Nb3 Qe8 8.Qe4 c6

    ...Stops mate threats on b7 and prepares ...Bd5, hitting the Queen again. - Mark

    ReplyDelete
  19. Another delay in playing a move, this time because 230,000 homes are without power after a massive storm hit. Our power came back on after only 24hrs but some will be without for three days. I hear that Belgium got a big dumping of snow which would have made life difficult.

    ...I dunno but Hikaru's opening's at the Tata Steel tournament looks like a great player that is frustrated by traditional chess to me. I certainly wouldn't blame him for being frustrated when we have Chess960 to enjoy.

    7.Nb3 Qe8 8.Qe4 c6 9.Nc5

    ...One last attempt to try and push back the same knight to the back rank....yes twice! I've been confused about what I should do with the d1 bishop all game - Harry

    Personally, I would be happy to finish this SP up at move 10 since I haven't been able to find anything concrete after my creative little queen foray, proving that indeed the queen can come out early in Chess960...but why?

    ReplyDelete
  20. Yes, Queensland weather has been in the news here again, with clips of catastrophic flooding. Re Belgium, I grew up in New England. By their standards, the amount of snow we received during the last few weeks made for a mild winter...

    8.Qe4 c6 9.Nc5 b6

    ...I could be mistaken, but the initiative seems to be passing to Black. Let's continue playing until we decide which SP to tackle next. - Mark

    ReplyDelete
  21. Well something is mistaken! I thought you played 8..f6 not 8...c6 my mistake! I detected this because your last move results in 10.Qb7#! I must have had a power failure in my head as well as my house!

    We'll have to wind back to move 8:
    7.Nb3 Qe8 8.Qe4 c6 (not ...f6) 9.Nf3

    ...The knight develops to it's natural square and only temporarily blocks the light bishop I hope. Unfortunately white is now only aiming for a slightly better position without queens. - Harry

    ReplyDelete
  22. Yes, it's the same sort of error that used to happen when chess was played via postcard. It's a pity we don't have a server where we can specifiy the SP, but that's life...

    8.Qe4 c6 9.Nf3 f6

    ...The choice was between 9...d6 and the move played. My move attacks the extended center, develops the Bishop, and prepares to attack the Queen. Not bad for a single move. - Mark

    ReplyDelete
  23. Well, at least you can download the exact PGN I am using which updates on dropbox, the minute I've played my move (see the main page for this blog).

    In parallel with this, I'm playing through SP484 at the moment. It features every major piece undefended including the king! There's only a handful of such SP's but so far they seem nice and balanced.

    8.Qe4 c6 9.Nf3 f6 10.Nc5

    ...In keeping with play so far. White should keep the position tactical, otherwise what was the point of developing the queen early? - Harry

    ReplyDelete
  24. How about tackling SP868 QBBRKRNN for the next game? I mentioned it in December as a comment on the SP408 game...

    9.Nf3 f6 10.Nc5 b6

    ...This move looks familiar. Oh, yes, I already played it on the aborted 9th move! - Mark

    ReplyDelete
  25. SP868 = QBBRKRNN What a horror position looks like fun!! QBB (Queen BEES) swarming two innocent undefended pawns on the opposite wing sounds like a horror movie! Not only that, but the two military knights in the corner can go to work on the king almost instantly. This one is going to be diabolical for black. Please, I beg you, play black for me? You are the stronger player and I doubt I have the stomach to bear this SP as black.

    On another topic, I've decided to try and freshen up my studies of Chess960 by going through the entire database of GM level Chess960 games from Mainz-through to the KvQ 2011 tournament.

    So the very first true professional game played in Chess960 history (Leko v Adams Mainz 2001) that I studied last night was SP748 only one off this SP. The sublimely ridiculous thing about it is that almost the identical opening was played e4/e5/Nd3/f5/exf5 but it was black that played the two pawns out not white. Get that!!

    9.Nf3 f6 10.Nc5 b6 11.exf6

    ...White has lost the initiative, but black's d7 and g7 pawns are weak and so black cannot yet claim to have taken it. I'm keen to try and prevent ...Bxf6 which would take the long dark diagonal from me - Harry

    ReplyDelete
  26. I don't mind playing Black. The objective of the current set of games is to verify that Black always has resources. If I play White and get an overwhelming position, I still need to tackle Black. Might as well just tackle it immediately...

    10.Nc5 b6 11.exf6 Bd5

    ...The current game looks fairly balanced and any result is still possible. - Mark

    ReplyDelete
  27. Have really enjoyed watching you find these resources available to black. I can find some of them as well, but my confidence and understanding of how they work is less clear.

    Bye bye to my queen...and bye bye to the initiative unless black makes a mistake that is.

    12.Qxe8 if Rxe8 then 13.Nxd7+ if Kc7 then 14.Bd4 ...

    ...I just think that the 3PvN exchange looks tidy enough that white can consolidate again after the over-excited opening plan. It's quite interesting to see the fate of white's 6th rank pawn and it's effect on the position over time. One major problem that has existed since the beginning, is what to do about the light bishop! - Harry

    ReplyDelete
  28. All good points, although theory says the minor piece is more valuable than the three Pawns in the middlegame. It depends how you categorize the current position; I would say it's a Queenless middlegame, but some might call it an endgame...

    11.exf6 Bd5 12.Qxe8 Rxe8 13.Nxd7+ Kc7 14.Bd4 Kxd7; if 15.fxg7 Nd6

    ...Rather than the suspense of discovering the fate of the g7-Pawn, I would prefer starting a new game. Any takers? - Mark

    ReplyDelete
  29. Agree, the over excited queen attack didn't work. I needed to back her up with a slower, more coordinated combined attack.

    I've set up the post for the trial of SP868. I'll play my first move over there. Thanks for going with me on this opening experiment!

    http://chess960jungle.blogspot.com.au/2013/02/non-random-chess-960-trial-game-8-sp868.html

    ReplyDelete