Aside from 22. g4, suggested by Dionyseus, I didn't find an improvement on Rybka's line. What I did find was that black has no option on his 22nd except the text, and the continuation is really not bad for white: 25...gxf5 26. gxf5 Nh5 27. Rc3 Rae8 28. Nf4 Nxf4 29. Rxf4 Rc8 with a promising-looking position. I have not yet found a winning continuation, but the position evaluates well in the computer.

So, to conclude, if you don't like 22.g4 or find something wrong with it, 22. Nce2 may just get the job done.

19. Bg5 Nh4 20. Ne4 d5 21. Bxe7 Qxe7 22. Nc2 Qd8

(22...Qh4 23. d4 Nf6 24. g3 Qh5 25. fxg6 Qxg6 0.97)

23. d4 Ng7 24. fxg6 fxg6 25. Nb4 Nf5 26. Rf3 Rc8 27. Qf2 Qd6 28. Nd3 Re4! 29. Nxe4 dxe4 30. Rxf5 gxf5 31. Nf4 Bb5 0.45

I dragged the line out to include the charming 28...Re4. Although black has better pawns, his king is exposed and this is by no means the endgame.

We should strive to create such a position rather than enter an endgame. We will get waxed in an endgame as our computers become virtually useless then.

In the Rybka line with 22. Nce2, I prefer 25. Rad1 to Ra3.

score 0.57 = 19.Bg5 Nh5 20.Bxe7 Qxe7 21.Rae1 Qh4 22.Ne4 Bxe4 23.dxe4 Rac8 24.fxg6 hxg6 25.Qxd6 Rxe4

score 0.58 = Bg5 Nh5 20.Nd4 f6 21.Be3 d5 22.Rf3 Bb4 23.Qf2 g5 24.Rh3 Ng7 25.Kh1 Rac8

<ughaibu> says: Returning to one of Dioynseus' "simple" defenses, 19.Bg5 Ng4 20.d4 Bb5 "21.f6, black is up a pawn. 21...Bxf1 22.Rxf1 Bf8 23.h3 h6 24.hxg4 hxg5 25.Qxg5 Qe6", after 26.Nd5 how is back meant to defend this? 26....Rac8 27.Rf3 etc.

Dionyseus resonds to <ughaibu> : <21.h3 also hits the eye.>

19.Bg5 Ng4 20.d4 Bb5 21.h3 and black gets a simple draw, for example 21...Bxf1 22.Rxf1 f6 23.hxg4 fxg5 24.f6 Bxf6 25.Rxf6 Qxg4 26.d5 Qh4 27.Re6 g4, white has nothing, draw.

<Rookfile> says:

As I explained earlier, I'm more inclined to trade dark squared bishops if white can keep everything else on, and maybe refrain from fxg6 at all.

There's a lot of engine analysis to wade through. For now, I will post what I consider as the leading for and against 19. Bg5 Kibitzer viewpoints.

<thorsson> says:
If you like Bg5 you need to find improvements in this line: 19.Bg5 Nh5 20.Nd4 d5 21.Bxe7 Qxe7 22.Nce2 Qe3+ when White has a small advantage, but not as big as the 19.Bb6 lines.

<Rookfile> says:
As I explained earlier, I'm more inclined to trade dark squared bishops if white can keep everything else on, and maybe refrain from fxg6 at all.

19. Bg5 Nh5 20. Nd4 d5 21. Bxe7 Qxe7 22. Nce2 Qe3+ 23. Qxe3 Rxe3 24. g4 Ng7 25. Rad1 Rae8 26. Ng3 0.67/16

19. Bg5 Nh5 20.Nd4 d5 21.Bxe7 Qxe7 22.Nce2 Qe3+ 23.Qxe3 Rxe3 24.g4 Ng7 25.Ra3

(23-ply) Sliding forward and extending the old analysis

1. (0.27): 22.Nce2 Qe3+ 23.Qxe3 Rxe3 24.g4 Ng7 25.Ra3 Rc8 26.Kf2 Rh3 27.Kg2 Re3 28.Rb3 gxf5

2. = (0.13): 22.Rf2 Qe3 23.Qxe3 Rxe3 24.Rd1 Rae8 25.g4 Ng7 26.Nc2 Rh3 27.fxg6 hxg6 28.Kg2 Rh6

(27-ply) Again sliding forward and extending the analysis

1. (0.35): 25...Ne8 26.Nf4 g5 27.Nh5 Rc8 28.Kf2 Rh3 29.Kg2 Re3 30.Rc1 Bd7 31.Rxc8 Bxc8 32.Rc3

2. (0.35): 25...gxf5 26.gxf5 Nh5 27.Rb3 Rae8 28.Kf2 Kh8 29.Rg1 h6 30.Rg4 Nf6 31.Rg3 R3e7 32.Nf4

Your fluid writing impressed me. A bit of jealousy here, too. The light sprinkle of whimsy and wordplay was nice.

Come visit and post on my funny farm, too.

By contrast, 19. Bg5 is not so easy for black to refute. Black could try to answer with ... Qd8, but then that really opens it up for 20. Rae1 with the extremely nasty threat of sacrificing the exchange and getting rid of black's best defender with 21. Rxe7! So 19. ... Qd8 isn't satisfactory, imo.

That only leaves the main Rybka line with ... Nh5. I don't want to go into all the lines, but a few things stand out. First, there is nothing that forces us to play Bxe7. Let black play ... Bxg5, allowing our queen to come into the endgame immediately. Instead, we take the extra tempo to centralize the Nb3 now that we have the maximum pressure on the dark squares. Second, in several of the lines after exchanges there is the possibility of keeping the black Nh5 out of the game with g3.

I'm back for a
quick scan by shredder10 18ply no sliding:

2. ± (0.72): 27...d5 28.Nd6 Re6 29.Qxf7+ Qxf7 30.Nxf7 Nc5 31.d4 Nb3 32.Ng5 Re2 33.Rf6 Nxd4 34.Rxg6+ Kh8 35.Nf3 Nxf3+ 36.gxf3 c5 37.Rg2

About ..Rf3 in Monad trap, my Rybka 1.2.n (**) burst out with this suggestion quite early (at 17 or 18 ply).

Could later evidence be due to 2.1.o changes???
According to CEGt extensive testing (http://www.husvankempen.de/nunn/), it seems that on single process machines the old Rybka 1.2.f performs better than 2.1o (something was lost in the upgrade?) I did similar analysis, on smaller samples using paired test comparisons(*), and got the same results. May I suggest that somebody, who bought Rybka, could download 1.2.f too, in order to avoid possible anomalous choices by 2.1.o?

(* I would like to have a weaker but faster result than Elo rating as CEGT: if we are interested in ranking engines, a powerful test is : eng A better or equal eng B, pairing test results on the same positions after "last book move": paired tests allow smaller test sample, with same significance)

(** Rybka 1.2.n is a private Raijlic version, based on 1.2.f . After short paired tests, I like it more than 1.2.f, but this is feeling, no statistical evidence behind)

Elvis Loses Top-Earning Dead Celeb Crown

Rock 'n' roll legend Elvis Presley ceded his crown to Nirvana lead singer Kurt Cobain on Forbes.com's list as the top-earning dead celebrity. The list, published on Tuesday, said grunge rocker Cobain earned $50 million between October 2005 and October 2006. Presley wound up in the No. 2 slot with $42 million, down from last year's $45 million.

Forbes.com bases its dollar amounts on licensing deals for using the deceased celebrities' work or image in advertising or elsewhere.

This was Cobain's first time on the list in its six years of publication. Presley has ruled the roost since its inception, said Forbes.com staff writer Lacey Rose.

Cobain's coup was due to his widow, actress and singer Courtney Love , who sold a 25-percent stake in the Seattle grunge group's song catalog to New York music publishing company PrimeWave.

Ranked after Presley is "Peanuts" cartoon strip creator Charles Schulz at $35 million. Rounding out the top five were Beatle John Lennon at $24 million and groundbreaking physicist Albert Einstein at $20 million, whose estate profited from such licensing deals as the popular "Baby Einstein" educational videos.

White has just moved 20.Nd4 and this workshop will analyse and discuss variations arising from a possible response by Black of 20...Ne5:

click for larger view