Computer analysis and review of this game: http://www.chess.com/votechess/game.html?id=16343

4... Nf6. As always, Rybka likes this early knight development. +.42@15

5... O-O. Rybka's choice early, before falling just slightly behind Nc6. +.30@15

6... c5. Again Rybka likes this early, then goes to Nc6, then back to c5, then Nfd7... clearly we have some flexibility in our setup here. ~+.25@14

7... Na6. Rybka likes e6, and gives a main line of e6 O-O exd5 exd5 giving black some space but still favoring white slightly (+.28@16). e6 dxe6 Bxe6 also favors white but now we can play Nc6 with nice control over the d4 and e5 squares, giving us some compensation. Or e6 dxe6 fxe6 aiming for a stonewall setup after e5, Nc6 and Nd4, again slightly favoring white but basically equal (this would be my preference, since the d6 weakness will be shielded nicely). Our move is slightly worse but nothing dramatic, with the difference probably being due to our being a bit more cramped here then in other lines.

8... Nc7. Rybka likes 8... Ng4, relocating the knight to e5. We never even considered this move, which scores +.30@16. We continue to slip ever-so-slightly, dropping to +.43@14 with our choice.

9... a6. Rybka likes b6 and Bb7, and is content to keep the position closed and await further developments. The computer's main line includes Nd7 Rc8 Qd7 etc, just hanging out and playing defense, scoring +.43@13. our move is more active and scores just as well though, so probably the best line for humans to play :).

10... b5. Correct. +.26@13

11... Rb8. White's 11.Bd2 is too passive, and Rybka now scores 11... Nd7 as completely equal. Our move is also good, scoring a mere +0.06@13 for white. Black has now equalized the game! Our planned line also includes Nd7 with play on the queenside, so we have the right ideas here.

12... e5? Rybka still wants to open up the bishop with potential relocations to e5 by playing Nd7, scoring the game as completely equal (0.00@15). Our move scores much worse, +.38@15.

13... exf4. Correct. White blunders back with 13.f4? a mistake allowing black to reopen the long diagonal for the bishop. Black takes a small lead, -0.14@14.

14... Nh5. Rybka likes this move at a depth of 9, but feels Re8 is a bit stronger after looking deep into the position. We are hovering around equality.

15... b4?! Rybka builds up on the e-file, going for Re8, Qe7, Nf6 and scoring the game as equal. After our move it is slight advantage white: +0.28@14.

16... axb4. I knew in my heart that Rxb4 was the right positional move for black here, keeping the center locked up and a grip on the dark squares and giving up on axb4 Na4 Bxa1. Rybka confirms this idea by seeing cxb4 Na4 f5 (not Bxa1) Ra2 fxe4 Bxe4 Rxf1 Qxf1 Qe7 as the main line for a long time, scoring it as small advantage white. After letting the computer think for 90 minutes and getting to a depth of 20 moves, Rybka comes around and declares that 16... Rxb4!! is best with a maneuvering game Bc2 Ne8 Qd2 Qe7 rae1 Nhf6 Nc1 Nd7 Nd3 Rb8 Bf4 Ne5 Bg5 Qc7.

Jimvger made the only vote this move, saving us from timing out and losing! Good job Jim!

17... Bxa1? Rybka decides not to win the exchange, as it critically weakens the dark squares around the black king. Instead, 17... f5 scores as +.25@14. Our move gives white a big edge as feared in the comments, with e5/c5 and a pawn break up the middle coming for white. Rybka scores the game +1.02@14. Ouch!

The moral of the story... in this system it's dangerous to give up our dark square bishop while white still has theirs on the board... even winning the exchange, it can lead to a loss! This piece as a precious kingside defender, and we should give it up only with the utmost care.

18... Qe7. Rybka is all over the board as it searches this position. Qe7 is the best move at a depth of 9, but Ne8 eventually wins out, scoring +0.83@12. But wait! At a depth of 12 moves, Qe7 is best again, scoring +0.93. Rybka sees the move we feared, as it is planning to play c5 very soon.

Hold the phone! Rybka looks deeper and finds a very interesting sacrificial idea for black: 18... Bxh3! gxh3 Qg5+ which scores +0.78@13. Seems this is a very interesting position...

19... f5? Rybka likes Bd7, scoring +0.99@12. The computer is planning Ne8 next and looks to be trying to hold off the coming pawn storm. After a little more thought the computer switches to Rb7, breaking the x-ray pin from the bishop and scoring +1.07@14. Our move scores quite a bit worse, but white must play it well. 20. c5! ignoring our threat and storming with the pawns scores +1.32@13 for white.

20... Bxf5. White blunders with 20.exf5?? and it is suddenly black who is in charge! 20... Bxf5 is correct and scores a whopping -1.23@16.

21... Rxf5. Correct.

22... gxf5. Correct.

23... Re8. White slips again with 23.Kf2? and we pounce by finding the best move again. It is now -2.06@15 and climbing. Black is winning.

24... Nf6? Rybka likes Qh4, a move we never considered. Qh4 Kf1 f4 scoring -2.49@14. our move is still clearly advantage black, but scores only -1.45@15.

25... Ne4. Correct. White slips again in this highly tactical position and our advantage grows to -2.44@13. 25.Bf4 was better for white.

26... Nc3. We were unsure whether Nc3 or Nc5 was best at the time, but Rybka agrees that our chosen move is correct, scoring it -3.07@17 vs -2.57@15. Both moves were very good for black though.

27... Kf7! Black finds the correct move, marching the king toward the middle onto a square that is eventually proven to be quite safe from harm. This move is almost a full pawn better then the natural move Kh8. The rest of the game, where white struggles to find a check on our king, shows why this move is so strong.

28... bxc3. White blunders again with 28.Nxc3 and now it's game over. We find the best move and by now we're giving forced lines to the victory. Rybka scores the game -6.51@17, even though we are up only the exchange in material at this point.

29... Qe2+. Correct. We have analyzed the win already and just march on with our plan from last move.

30... Re3! We have planned this for a while now, but it's still a strong move. Rybka confirms it is the best.

31... Rxb3! We were unsure whether Rxb3 or the defensive Ne8 was best, but some nice team calculation resulted in the best move again. We proved both moves as winning, but found that Rxb3 was crushing. Rybka agrees, scoring Rxb3 -5.17@16 vs -3.11@14 for Ne8. Wow!

32... Qe4. Rybka likes Qd3 with all the same threats (mainly Rb1) but which takes away Nc3 as a reply from white. Nb2 Qd4+ winning the knight plays out similarly to the actual game.

33... Qd4+. Correct.

34... Qxc3. We weren't sure whether Qxc3 or Rxc3 was best, and Rybka ended up siding with Rxc3, planning to answer Qh6 with Ne8. Qxc3 scores almost as well and we showed a full win with this move, so no worries here.

35... Qe1+! White blunders by missing Qh6, and we find the best response, with the strong idea of heading to e4. Rybka scores this -9.85@15...

36... Qe4. Correct.

37... Rxh3+. White blunders again, but it's totally lost anyways.

38... Qe1#. Wrong! Rybka prefers Qb1# :)

Oddly our biggest mistake was winning the exchange by playing BxR on move 17. I think this is a very instructional game for understanding the importance of our fianchettoed bishop.

Also very instructional for me is Rybka's idea of 7...e6 and how we are going to hide the weakness of the backward d pawn to maintain equality if white play dxe6 (fxe6 / e5 / Nc6 / Nd4 stonewall setup). If white doesn't capture dxe6, then exd5 exd5 and we get some much needed space. Being able to find space in this opening is critical to black maintaining equality.

Great analysis once we got the lead. We found a long string of excellent moves and basically smashed them after they blundered with 20.exf5. Well done team!

The Bad Bishops

Super Admins