ELECTRONIC T-NOTES


CHESSBASE USA'S WEEKLY ON-LINE NEWSLETTER


FOR THE WEEK OF SEPTEMBER 13, 1998


WHAT'S SO "UNCLEAR" ABOUT THIS?

by Steve Lopez

One of the great annoyances in a class-level player's life comes when he or she is reading a chess book, sees something interesting, really starts getting into the game or analysis, and the author stops him short with a cryptic comment such as "The rest is a matter of technique" or "The position is unclear".

"Unclear", my eye! Tell the truth. Tell us that you haven't the foggiest dang clue as to what's going on. At least tell us you didn't feel like doing the analysis. That's what chess is all about: figuring out positions that you haven't seen before, positions that are unclear. Man, whenever I see a comment like that, I'm ready to heave the book or magazine out the nearest window! If I'm really angry my first impulse is to throw it without opening the window first!

For much of my chessplaying career, I've been left hanging like that. It's either figure it out for myself (which is hard to swallow, especially in light of the fact that I, a class-level player, just shelled out $20 for a book so I could learn something and the master-level yutz writing it refuses to tell me what he knows) or else just give up on the whole idea of playing that particular line or understanding that particular position.

This is why I'm so delighted with the "correspondence analysis" feature of Fritz5. I can feed Fritz, Hiarcs, Doctor?, or Junior a position, let it analyze it for a few hours, and receive a full tree of master-level analysis giving me the information I just (supposedly) paid some of my hard-earned ducats to receive, but failed to do so.

I know I promised a look at more opening research techniques for this week's ETN, but I just have to get this particular rant off my chest. Early in the week I was perusing Unorthodox Chess Openings by Eric Schiller (which I find to be a pretty good book). But at one point, Schiller left me hanging. In his section on the Fajarowicz Defense, he gives the following sub-variation:

1.d4 Nf6 2.c4 e5 3.dxe5 Ne4 4.a3 b6 5.Qd5 Nc5 6.Qxa8 Nc6

In his analysis of 4...b6 (the Bonxdorf Variation), Schiller says that Black has an interesting trap involving a Rook sacrifice after 5.Qd5. The words "interesting trap" and "Rook sacrifice" immediately caught my eye as A) an avid Budapest Defense player, and B) a serious lunatic. So I gave this line particular scrutiny. The book says that in this position the White Queen is trapped in the corner. It goes on to say, however, that "further testing is needed".

"Further testing". Alarm bells went off in my head. "This sounds suspiciously like 'the position is unclear,'" thought I. I suppose he meant that the position hasn't been reached enough in master practice to be able to form a conclusive evaluation of it. Then I thought "If this is a good line, why hasn't it been played more? Is it just the result of the natural caution of the master-level professional player?" It's either that or they're relying on Joe Fishboy (i.e. class-level correspondence gambit-freak me) to do the "testing" with my butt and my rating points on the line. Pardon me whilst I experience strong annoyance...

Let's go back for a moment and take a look at the position above. This is a master-level position? This looks like something my pal Bill and I played last night after downing a few frosty ones. I honestly have to wonder about this. Black has given up a Rook and tried to make it look like an accident. Why? His development is not that great and while the White Queen appears to be trapped in the corner there's no way to immediately cop it off and throw it back in the box. Was tossing a Rook overboard really worthwhile to get to this position?

This is a perfect job for the correspondence analysis feature of Fritz. Let's give it to Fritz and have it do the sweat work and we'll see if this is a playable line or not.

Firing up Fritz, I entered the line, jumped to the position after 6.Nc6. and clicked on the correspondence analysis function. In the dialogue box, I used the following settings on my 133MHz Pentium (allowing Fritz 48Mb for its hash tables):

Time: 180 sec.
Branching in 1st move: 3
Branching in 2nd move: 3
Branching in 3rd move: 2
Length of variations: 9
Evaluation window: 80

Let's turn Fritz loose overnight and see the results of its "further testing". I've added some comments of my own to the variations Fritz provided.

Fajarowicz analysis [A51]

1.d4 Nf6 2.c4 e5 3.dxe5 Ne4 4.a3 b6 5.Qd5 Nc5 6.Qxa8 Nc6

7.Bg5 White takes the initiative by immediately attacking the Black Queen. But Black can just reply with ...Qxg5. So what's White up to?

[7.Nf3 Nb3 8.Bg5 f6 9.exf6 gxf6 10.Bxf6 Qxf6 11.Qxc8+ Nd8+- White's best bet here is Ra2. The White Queen can still escape by playing ...Qxc7. Have Fritz run some further analysis on this position - the results may surprise you.]

[7.Nd2 A) 7...Nxe5 8.b4 (8.Ngf3 Nxf3+ 9.Nxf3 Ba6 10.Qxa7 Bxc4 11.Bf4 Ne6+-) 8...Ne6 9.Qxa7 Qf6 10.Bb2 Nd3+ 11.exd3 Qxb2+-; B) 7...a5 8.Ndf3 (8.Ngf3 Be7 9.b3 0-0 10.Rb1 f6 11.exf6 Bxf6+-) 8...Ba6 9.Qxd8+ Nxd8 10.e3 Bb7 11.Bd2 Nde6+- At least Black can still castle, but he's going to get shelled anyway once the Queenside opens up.; C) 7...Ne6 8.Ngf3 (8.e3 Bc5 9.Ne4 0-0 10.Ng3 Qg5 11.e4 Qxe5+- This is so far the best alternative we've seen for Black, but it requires some uninspired play by White to get here. 8.e3 is just a dog.) 8...Bc5 9.Ne4 0-0 10.Nc3 Ned4 11.Nxd4 Bxd4+-]

7...Qxg5 The whole idea is to draw Black's Queen off of the back rank.

[7...Bb7 8.Qxd8+ (8.Bxd8 Bxa8 9.Bxc7 Nb3 10.Ra2 Nc1 11.Nc3 Nxa2+-) 8...Nxd8 9.Nc3 Nde6 10.Bh4 Nb3 11.Rd1 Nbd4+-]

[7...Be7 8.Bxe7 Nxe7 9.b4 Nb3 10.Ra2 Nd4 11.Rd2 c5+-]

8.Qxc8+ Qd8 9.Qxd8+ Kxd8 We see now that the White Queen wasn't trapped after all. Now Black will be unable to castle. 10.Nd2 Nxe5 11.b4 Ne6+-

According to Fritz5, this is the best line for Black. I don't know about you but if I was the player of the Black pieces I'd be flipping the board here. Black has a one-piece development advantage. Unfortunately, both of his developed pieces are the short-ranged Knights. White, meanwhile, has a space advantage on the Queenside and nothing to fear at this point. And Black gave up a Rook for this? Man, you've got to be kidding me! No thank you! Not even after a few frosties!

To be extra certain of the analysis, I'll load the Hiarcs6 engine, give it some extra time (Hiarcs runs a bit more slowly than Fritz), and have it look at the position also. I doubt that it will find anything radically different; if it does, I'll let you know.

If that doesn't happen, I think we can conclusively say that "further testing" is not needed, contrary to what the book says.

The bottom line here is that Fritz is so much more than just a chessplaying program. It's a master-strength analysis aide and can go a long way toward eliminating those annoying cryptic non-committal comments we frequently see in expensive chess books.

Until next week, have fun!

You can e-mail me with your comments and suggestions for Electronic T-Notes.