V. Rajlich - J. Gallagher
[B67/12] Sicilian: Rauzer (Modern)
29 May 1999, Chicago, IL

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 Nc6

This was new to me, and I was now faced with an interesting practical choice. On the one hand I could play 7.Qd2 and steer the game into normal Richter-Rauzer lines, which I also play as White. Objectively, however, the best move had to be 7.Bxf6, since Black shouldn’t be able to force a transposition into the Rauzer from the 6.Bg5 Najdorf (this would be a nice trick if he could, since not all 6.Bg5 Najdorf players play the Rauzer), particularly with a move I’d never seen.

I think these sorts of decisions need to be made at the board depending on the situation. My opponent is known to be a theoretician, and time on the clock is likely to be very important in an Open Sicilian against a good player, so I went ahead and just played 7.Qd2 nearly instantly. For what it’s worth, Nunn considers 6...Nc6 7.Bxf6 very good for White, and Wells & Osnos in The Complete Richter-Rauzer then give 7...gxf6 8.Nb3 e6 9.Be2 and quote a game without any assessment.

7.Qd2 e6 8.0-0-0 Bd7 9.f4 b5 10.Bxf6 gxf6 11.Kb1

I couldn’t remember the theory here. As it turns out, my opponent had some trouble recalling it himself in the post-mortem, even though he wrote the chapter on the Classical Sicilian recommending this setup for white in his and Nunn’s Beating the Sicilian 3.

11...Qb6 12.f5

I am the first to deviate from the BTS3 recommendation (unintentionally, of course), which goes 12.Nxc6 Bxc6 13.Qe1 with unclear play. My move, however, makes a lot of sense, since White just plunges straight ahead.

Black’s position is now critical. Normal developing moves such as 12...Rc8 seem to give White a very strong initiative, for example 13.fxe6 fxe6 14.Nxc6 Bxc6 15.Qf4 and with ideas of 16.Be2, etc., when it’s hard not to like White.

The obvious continuation is 12...Nxd4 13.Qxd4 Qxd4 14.Rxd4, and at the board I had a hard time imagining something else being played. As it turns out, this ending is in fact critical. Black’s most ambitious, and probably best, move is 14...Ke7, preparing to activate his dark-square bishop with either ...Bf8-h6-f4-e5 or ...Bf8-h6-e3-c5. If he can activate this piece without difficulty, then he will definitely stand better. White, however, can play the active 15.a4, and it seems that he then has enough play for at least equality and perhaps slightly more. 15...bxa4 16.Nxa4 Bh6 17.Be2 is one possibility, when White has ideas of Nb6 plus the pawn on a6 is weak; 15...Rb8 is another possibility, when after 16.axb5 axb5 17.Rb4 White has play against b5.

My opponent rejected this continuation because he thought that in this ending after 15.a4 White would have an edge. At the board I didn’t look into this ending very deeply - it didn’t even occur to me that 14...Ke7 threatens to be good for Black. What also didn’t occur to me is that Black has another plausible continuation, which was chosen by my opponent:

12...b4 13.Nce2 e5

This struck me as very bad at the time. Black’s position is just riddled with holes (especially the one on d5), his kingside pawns seem weak, his dark-square bishop can’t even be activated via h6 due to the weakness on d6, and on top of this White has a very simple and natural plan for exploiting d5: Ng3, Bc4, and Ng3-f1-e3-d5.

However, there are two other considerations which seem to offset all of the above. One is that White’s occupation of d5 takes a lot of time. In the meantime, Black will be able to exploit his main advantage in the position: the fact that his king is a little safer than White’s, as Black has a natural attacking plan (i.e. advancing his queenside pawns), while White’s plan of occupying d5 does not seem to translate into any attacking chances against the black king. My opinion now is that this position is roughly equal.

14.Nf3 a5 15.Ng3 h5 16.Bc4

I played this immediately, and it’s probably best, but black does get a little bit of play against the loose bishop. It wouldn’t shock me if somebody proved that something else was better.

16...Rc8

17.Qd3

This seems best. 17.Nf1 is the obvious move. At the board I rejected it because I concluded that 17...Nd4 seems to then help black after 18.Qd3 Nxf3 19.gxf3 Qc5 since now 20.Ne3 would lose to 20...Bh6. In this variation the trade of knights interferes with White’s plan of getting a knight into d5, and this variation is reason enough to dismiss 17.Nf1.

17.Nf1, however, is in fact even much worse than that. After 17...Nd4 18.Qd3, Black busts out 18...d5 and White is struggling to just stay in the game. This is a tough tactic to see, we didn’t see it even in the post-mortem with the position sitting on the board.

The other possibility was 17.Qd5, forcing 17...Nd8, and only then 18.Qd3. I ended up concluding, I think correctly, that Black’s knight is better on d8 than on c6. On c6 it interferes with his pressure down the c-file and with his ability to cover d5 with ...Bc6, while from d8 it can go to b7 and then c5.

17...a4 18.Nf1 Na5 19.Ne3 Bc6

In these positions Fritz likes White in general and here it suggests the resource 19...Nxc4 20.Nxc4 Qc6, the point being that the obvious 21.Nxd6+ runs into 21...Bxd6 22.Qxd6 b3+ with an edge for Black. However, White can just play 21.Ne3 when he seems to have a slight edge. I prefer my opponent’s more natural move 19...Bc6.

After 19...Bc6, the following position has arisen:

As I think back on this game, it is really here that I start to go wrong. So far, my play has been good, I’ve seen pretty much the right amount of stuff, and I’ve played quickly, which is appropriate since most of White’s moves have been pretty natural. The one semi-think I’ve had so far was in selecting 17.Qd3 over 17.Nf1 and 17.Qd5, and that was a reasonable decision to take some time on. In fact, here I should just continue this policy: I should quickly play the one simple obvious move (20.Rhe1, defending e4, etc.) and see what Black does.

In fact, I spent a bunch of time on two very improbable possibilities (20.Bd5 Bb5 21.Bxf7+ and 20.g4) and after a long think played a third only slightly more moderate (and slightly less incorrect) choice.

20.Bd5 Bb5 21.Bxf7+, which I probably spent about 15 minutes on, as far as I can guess, loses to 21...Kxf7+ 22.Ng5+ (That 22.Qd5+ Kg7 23.Ng5 Be8 - or just about anything actually - loses didn’t take me much time. 22.Ng5+ is the attempt to rescue this line.) 22...Kg7 23.Ne6+ Kh7 and White is just down a piece. 24.Nd5 does not work due to 24...Bxe3 25.Nxf6+ Kh6 and White’s checks run out.

20.g4 loses simply to 20...Nxc4 21.Nxc4 Qc5 (threatening 22...Bb5) 22.Nxd6+ Bxd6 23.Qxd6 Qxd6 24.Rxd6 Bxe4 etc.

When I show these lines now, they look far more pointless than they seemed at the board, and this is actually natural. For example, in the 20.Bd5 Bb5 21.Bxf7+ Kxf7 22.Ng5+ line, 22...Kg7 23.Ne6+ seemed dangerous for White, so I first went down the line 22...fxg5 23.Qe6 Rh6 24.Qxc8 Qxe3 25.Qc7+ Kg8 25.Kxa5, which is just unclear. I guessed wrong about which defense was most likely to work, and this cost me time.

The same thing happened with 20.g4. I was looking at 20...hxg4 21.Nxg4 Bxe4 22.Nxf6+ Kd8 23.Nxe4 Nxc4 first (this also is just unclear).

However, while it’s natural that the mechanics of calculating these lines might take some time, I think the process needs to be intercepted higher up. Ben Finegold likes to tell me how often very strong players miss easy tactical wins like mates in 3, etc., and really how irrelevant it is that they do this, and I think that here is a good example of what that means.

When you play a player in the 2200-2400 (and possibly higher) range, while there may be some so-called positional stuff going on, there are also going to be a lot of tactical errors. So, it makes some sense to just look around at random tactics. After all, they often work - a game is going along, you start looking at some move like 20.Bd5 Bb5 21.Bxf7 or 20.g4, and next thing you know it just wins. At the same time, good positional play really isn’t going to be that important, since tactical chances will exist either way later on; and saving time on the clock will also be less important. It’s never a good idea to get into horrible time trouble, one of these 10 moves in 2 or 3 minutes things, but it’s easy to get away with 10 moves in 10 or 15 minutes, since the endings which arise are not very likely to be complicated and even if they are complicated then there won’t be a great a need to play them accurately, since again plenty of chances will exist to recover.

Against a stronger player, however, these things change. For one thing, the chance of just finding some such random win decreases greatly. Furthermore, the quality of one’s positional play becomes important. A slight disadvantage might last for a while. Finally, and probably most importantly, the time will be very important all the way to the end of the time control. In this game I eventually entered an only slightly worse 3 piece ending in which I had about 15 or 20 minutes for about 15 moves. Normally this would be fine, but I really needed more time to really get to the bottom of the ending and play it right, and as it was I had to cut short my analysis, misplaced my pieces, and wasn’t given a chance to recover. I’d rather take the small chance of missing working out something shocking like 20.Bd5 Bb5 21.Bxf7 or 20.g4 than to be done in by a series of small mistakes again, and I guess when good players play each other they think the same thing, and this is why they occasionally overlook some improbable tactic.

20.Rhg1? Nxc4

My opponent jumps at the chance to force the ending.

21.Nxc4 Qc5 22.Rge1 a3 23.Nxd6+

Forced, as 23...Bb5 was threatened.

23...Bxd6 24.Qxd6 Qxd6 25.Rxd6 Ke7

Black’s compensation is obvious, but when I played 20.Rhg1 it never occurred to me that forcing the ending might be a viable idea for Black. This is what getting to the bottom of a position consists of - checking these endings, taking more than a superficial look at them, and, if need be, buying time for this fundamental search of the position by bypassing looking at some tempting tactics.

26.Rd2?

This is where having an hour on the clock instead of 15 minutes would have been nice. With 26.Rd2 I allow Black to misplace my pieces. Best was 26.Rd3 and if 26...Rc7 then 27.c4. I’ll claim equality here, though if I had to pick a side it would be Black.

26...axb2 27.Kxb2 Rc7 28.Re3 Ra8 29.Rd1

Ugly, but necessary to meet ...Rca7.

29...Rca7 30.Ra1 Ra3

Black’s pieces are all much better than White’s.

31.Ree1

Else 31...Bb5 and 32...Bc4.

31...Rd8 32.Re2 Bb5 33.Rf2?

This obvious move loses material and probably the game. Surprisingly, the correct move was 33.Rd2 and now if 33...Rda8 then 34.Rdd1 Bc4 35.Nd2 and now if 35...Bxa2 then White just shuffles his d1-rook around and Black can’t make progress. So, Black would have to stay away from the a-pawn, and White can keep his e-pawn, though Black is still considerably better.

33...Rda8 34.Nd2 Re3 35.Rb1 Bc6 36.Ka1

Actually, my original idea was 36.Kc1, I missed 36...Rxa2 37.Rxb4 Re1#.

36...Bxe4 37.Nc4?

Necessary was 37.Rxb4 Bxc2 38.Rb7+ Ke8 39.Kb2, though I can’t imagine that Black won’t win this.

37...Bd5

Obviously I missed this, I expected 37...Rc3 when after 38.Rxb4 White is still in the game after either 38...Rxc2 39.Rxc2 Bxc2 40.Rb7+ or 38...Bxc2 39.Nb6.

(Time), 0-1.

I just sat there for the last two minutes or so trying to discover something.