A Dead Man Explains It All to Fritz
By IM Timothy Taylor
Those of you who have been following this series can skip this introductory paragraph, but for those joining us late, here’s the word one more time.
Until a few months ago, I had never defeated my “difficult opponent” I.M. Andranik Matikozyan. I succeeded in making one draw, and lost eight times! In nearly every game I obtained a better, often absolutely winning position – and then went on to lose.
Today’s game is no exception. I win the opening battle with Black, and very quickly obtain a big advantage (Fritz puts me 1.1 ahead by move seventeen, in other words my positional advantage is worth more than a pawn. This evaluation is absolutely correct, as I have confirmed the computer’s opinion with my own analysis.
Naturally, having obtained this big advantage, I threw it all away and lost!
Now what is the instructional value of this, you may rightly ask? Have we not already seen no less than four previous games (since one previous column contained two Matikozyan debacles) in all of which I obtained a winning position and then lost? Is this not getting a bit repetitive?
The answer is, simply, “No.”
Do you know why? Because this time I found a new and completely different way to lose my advantage! In all the previous games I have shown, I have lost to my difficult opponent due to some kind of specific blunder – but in this game, for novelty’s sake, I lose due to a faulty plan, not due to a faulty move.
I found the analysis of this game quite instructive – to me, and I hope, to you. Because my error was strategic in nature, rather than tactical, I found my junior partner Mr. Fritz to be almost completely useless! I’ll explain why when we get to the critical position of the game, after White’s seventeenth move.
IM Andranik Matikozyan - IM Timothy Taylor
Philidor Defense Reversed (C44)
Los Angeles 2004
1.e4 Nf6
There is a story behind this move, which I play quite rarely. Since I was always losing to Matikozyan at this time, I took a keen interest in his games. So one day, at some typical Southern California Swiss system tournament, I spied my nemesis replying to Alekhine’s Defense with 2.d3, which Alekhine himself described as “A very tame continuation, which does not offer prospects of any opening advantage.”
Clearly the most challenging move is 2.e5, while the developing 2.Nc3 is solid and sound. But 2.d3? What is this? I was familiar with Alekhine’s two big wins against this passive try, vs. Maroczy at New York 1924 and vs. Thomas at Baden Baden 1925. The move has had a bad reputation ever since, and is represented only by a small footnote in ECO.
So why would my difficult but very strong IM opponent play this silly move? Alekhine is absolutely correct: White can hardly get any advantage by blocking in his own king-Bishop! In other words, Matikozyan voluntarily gave up the advantage of the White pieces on move two!
I found this to be very curious, so that night I looked up Matikozyan vs. Alekhine’s Defense in the Chessbase database. What did I find? There were only two games where he faced Alekhine’s Defense: in both of them he played the strongest, 2.e5 – and he lost both games. Here’s what I saw: 2.e5 Nd5 3.d4 d6 4.c4 Nb6 5.exd6 Qxd6 (5...exd6 6.Nc3 Be7 7.Nf3 0–0 8.Be2 Re8 9.0–0 Bg4 10.b3 Bf6 11.h3 Bh5 12.Bb2 Nc6 13.Nh2 Bg6 14.Nf3 a5 15.a3 Bh5 16.Nd2 Bxe2 17.Nxe2 a4 18.b4 d5 19.c5 Nc4 20.Nxc4 dxc4 21.Rc1 Qd5 22.b5 Ne7 23.Nf4 Qd7 24.Nh5 Nd5 25.Qf3 Re6 26.Rc2 Rae8 27.Rxc4 Qxb5 28.Nxf6+ gxf6 29.Rcc1 c6 30.Bc3 Qb3 31.Bd2 Re4 32.Rfe1 Rxe1+ 33.Rxe1 Rxe1+ 34.Bxe1 Qxf3 35.gxf3 Nf4 36.Kh2 Ne2 37.Kg2 Nxd4 38.Bd2 Kg7 39.Bc3 Ne6 40.Bb4 Kg6 41.Kg3 Kf5 42.Kg2 Nf4+, 0–1, Matikozian-Baklan, Duisburg 1992) 5...cxd6 6.Nc3 g6 7.Nf3 Bg7 8.h3 0–0 9.Be2 Nc6 10.0–0 Bf5 11.Be3 d5 12.c5 Nc4 13.Bc1 b6 14.b3 bxc5 15.bxc4 cxd4 16.g4 Be4 17.Qa4 d3 18.Nxe4 dxe4 19.Re1 dxe2 20.Rb1 exf3 21.Qxc6 Bc3,
0–1, Matikozian-Zilberstein, Bratislava 1993.
I was very cheered by this “Achilles heel” that I had found in my difficult opponent’s opening repertoire, and resolved to play Alekhine’s Defense against him at the first opportunity, which indeed came soon at another Los Angeles Open.
My opponent stared extremely suspiciously at my audacious knight; no doubt remembering his two losses when he pushed his e-pawn, he once again played the passive
2.d3?!
My opening surprise had worked!
2...e5
This was Alekhine’s choice against Maroczy – he defeated Thomas with 2…c5 and also good is 2…d5 (2...d5 3.Nd2 e5 4.Ngf3 Nc6 5.g3 Bc5 6.Bg2 dxe4 7.dxe4 0–0 8.0–0 a5 9.a4 Qe7 10.c3 Rd8 11.Nh4 g6 12.h3 b6 13.Re1 Rd3 14.Nhf3 Bxf2+ 15.Kxf2 Qc5+, 0–1, Todorcevic-Kovacevic, Ljubljana/Portoroz 1989) – in general just about anything works against White’s self-blocking move!
3.Nf3
Maroczy tried 3.f4, which Alekhine thought was the best try, but the latter neutralized it with the effective pawn sacrifice 3…Nc6 4.Nf3 d5! and went on to win in twenty-four moves.
3…Nc6
We have reached Philidor’s Defense, where White has an extra move – but this defense is so quiet that the gained tempo gives White no advantage.
When sharp Black openings are reversed (I’m thinking Sicilian/English or Dutch/Bird) there are quite good chances for a plus, but here, with White hampered by the d3 block, there is nothing special, as many GM games can attest, see the following note.
4.c3
The alternatives also give White nothing: 4.g3 d5 5.exd5 Nxd5 6.Bg2 Bc5 7.0–0 0–0 8.Re1 Re8 9.h3 Bf5 10.Nh4 Be6 11.Nd2 Nf4 12.gxf4 Qxh4 13.Ne4 Bxh3 14.Nxc5 Bxg2 15.Kxg2 exf4 16.Ne4 Nd4 17.f3 Re6 18.Nf2 Qg3+ 19.Kf1 Rg6, 0–1, Manik-Oral, Trencin 1995; 4.Be2 d5 5.Nbd2 g6 6.0–0 Bg7 7.c3 a5 8.a4 0–0 9.Re1 Re8 10.Bf1 b6 11.exd5 Nxd5 12.Qb3 Nf4 13.Ne4 Be6 14.Qc2 Nd5 15.g3 h6 16.Bg2 Qd7 17.Ned2 Rad8 18.Nc4 Ndb4 19.cxb4 Nxb4 20.Qc3 Qxd3 21.Na3 e4 22.Qxd3 Nxd3 23.Rxe4 Nxc1 24.Rxc1 Bxb2 25.Rxc7 Rd1+ 26.Re1 Rxe1+ 27.Nxe1 Bxa3 28.Nd3 Rd8 29.Bf1 Bf5 30.Rc3 Bb4 31.Rb3 Rd4 32.f3 Bd7 33.Kf2 Bxa4 34.Rb2 Be7 35.Ke3 Rd6 36.Ne5 Rd1 37.Bc4 Re1+ 38.Re2 Bc5+ 39.Kd3 Rxe2 40.Kxe2 Kg7 41.Nxf7 Be8 42.Ne5 Bd6, 0–1, Todorcevic-Tal, Marseille 1989.
4...a5 5.Nbd2 d5 6.Be2 Bc5 7.0–0 0–0 8.Qc2
The fork trick doesn’t give White anything: 8.Nxe5 Nxe5 9.d4 Nxe4 10.Nxe4 dxe4 11.dxc5 Qh4 with good attacking chances for Black.
8...Re8 9.b3
A typical Hanham variation of the Philidor (reversed): Black has better center control and better development; White is crouching on the first three ranks and can barely get his Bishops out; and even when they get developed, they have no scope.
Black is already better, and I begin to probe for weaknesses.
9…Nh5 10.g3 Bh3 11.Re1 Nf612.Ng5 Bg4 13.Bf1 h6 14.Ngf3 Qd7 15.Bb2 Rad8 16.a3 d4
Otherwise White might get a little play with 17.b4, which now fails to 17...dxc3.
17.c4?!
White evidently did not like 17.cxd4 Bxd4 18.Nxd4 exd4 when both his Bishops are bad, but this gives White more chances to equalize as he may be able to gain some much needed space later with f4.
However, the move 17.c4 was played – and both at the board, at home, and on the Fritz screen, there is complete agreement: Black has a significant advantage (as I mentioned in the introduction, Fritz puts me at about +1.1).
First of all, let’s analyze Black’s advantages.
The most important factor is space: Black has two guys on the fifth rank (pawn and Bishop) while White can’t even get to the fourth rank, let alone the fifth!
A long-term advantage is that both White Bishops are bad: his K-Bishop is almost completely choked by his own pawns, his Q-Bishop is shut off cold by my pawn chain. Meanwhile my Q-Bishop is very good, and my K-Bishop is not so bad, attacking a3, while also having some possibilities of kingside action with Be7-g5.
It follows logically then that since Black has more space and better pieces, there are some chances to attack the White King as well: it’s clear that White’s cramped position makes it hard for him to bring pieces over to the kingside for defense.
That seems to be more than sufficient advantages for right now, solidly justifying Fritz’s +1.1 evaluation. I repeat, during the game, I was well aware that I was clearly better here, and well aware that I needed to find the right plan to exploit my advantage.
This is not a tactical position at all. There are no good takes or checks in the position, nothing is hanging on either side. This is where strategic planning is a must.
I took a long think here, and came up with the following plan: I decided I would retreat my Knight from c6 to b8, and my Bishop from c5 to f8. Then I would play my pawn to c5, eliminating any potential White counterplay with b4. Finally I would open the b-file with …b5, and pressurize his backward pawn on b3. I felt that this pressure would prevent him from activating his Q-Knight, and then I could operate with threats of …a4; while if he forestalled that with a4, I would get the square b4 for my pieces.
The only problem with this plan, as you will see later when we get back to the main game, is that it doesn’t work! You will see that White adroitly defends b3 with his Rooks, frees his Knight, and finally gets that piece to the active c4-square, a square he would never have reached had I not opened it up with my faulty plan!
So what should I have done?
One advantage of doing this series is that it has forced me to go back and seriously analyze these very painful losses. I hadn’t much looked at this game in the two years since it had been played, and now I was very curious: what plan should Black adopt to increase his undoubted advantage?
My first thought was to ask my pal Mr. Fritz. I have become so used to studying with the computer that this seemed a logical first step. Fritz had nothing much to say at normal speed, so I decided to try out the “deep analysis” feature. This led to … nothing! Quite frankly, the computer did not have a clue! Had no strategic insight! Without any tactics to look at, the machine was helpless. All it did was shuffle the pieces around and lose my advantage.
A typical absurdity was this “computer vs. computer” line: 17...Nh7 18.Bg2 Qe6 19.Rab1 Ra8 20.Bc1 Be7 21.Kf1 Nf8 22.Kg1 Nd7 23.Rf1 Nc5 24.Re1 Bh3 25.Kh1 Rf8 26.Bxh3 Qxh3 27.Ng1 Qh5 28.f4 exf4 29.gxf4 f5 30.exf5 Qxf5 31.Ne4 Bd6 32.Re2 g6 33.Bd2 Kg7 34.b4 axb4 35.axb4 Nxe4 36.dxe4 Qh5 37.b5 Ne7 38.Rg2. Here the computer helpfully pointed out that White was equal! Yes, of course White is equal now, you fish and chip! If you do nothing but exchange pieces, you will free White’s game! So after a few similarly “helpful” examples of computer (non) strategic thinking, I gave up on the machine.
But the fact remained, Black was still clearly better after White’s seventeenth move – but I was as confused as the reader who said he always lost track, in Dashiell Hammett’s novel, RED HARVEST, after the chapter called “The Seventeenth Murder.”
I too was clueless!
Let’s go back to Black’s advantages: more space, better pieces, potential kingside attack.
Which did I say was most important? Space! And let’s look a little deeper at that advantage. I mentioned that Black has “two guys” on the fifth rank, a Bishop and a pawn. Which one is more important? Clearly the pawn, as it is a virtually permanent advantage: no White pawn can attack the Black pawn at d4, so it will clamp down on White’s position forever.
And let’s go further: where does one see that kind of pawn structure in a double King pawn opening – yes! We see this in the Ruy Lopez, as White. Consider this main line: 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0–0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0–0 9.h3 Na5 10.Bc2 c5 11.d4 Qc7 12.Nbd2 Nc6 13.d5 and White reaches the identical central pawn structure that I have in the game, with reversed colors. There are 449 examples of this popular line in the Big Database.
So in effect, my Alekhine’s Defense/Reversed Philidor has now morphed into a main line – for White – of the Ruy Lopez! And while I’ve never played this line for White in my life, I have read a thing or two – at this moment my eyes wandered around my chess library, and my gaze zeroed in on PAWN POWER IN CHESS, by Hans Kmoch.
I cheerfully shouted, “Eureka!”
(Actually, I pissoffedly shouted, “Expletive, Expletive, EXPLETIVE, Benoni Jump!” but don’t people want me to write ‘nice’ articles now?)
I opened PAWN POWER (copyrighted in 1959 by the way) and turned to Chapter VIII, “Benoni Formations.” This Ruy Lopez type formation is called by Kmoch the “Spanish Benoni.”
And then came the fateful words I knew were coming: Kmoch writes, “Before starting to discuss Benoni formations one by one, we must deal with a typical tactical twist that may easily occur in any of them. It consists of the sacrifice of a Knight on KB5. We call it the Benoni jump.
“Benoni systems require the fianchetto of the K-Bishop on the part of the defender. Hence …P-KN3, which is also a preparation for …P-KB4, and often induces White to play P-KN4. This, then, is the pawn situation where N-KB5 may occur – provided White is ready to use the K-Knight file for an assault on the King.”
(Pardon the English notation, it’s an old book!)
Anyway, the light was now on in my brain, and everything was illuminated. I knew what I had to do; now I just had to figure out how to do it. Basically, the correct plan is this: Black should keep the center solid, and keep White under pressure on the queenside (a3 is a handy target). Meanwhile, I should play …g5 both for choking purposes (keeps White passive, prevents the potential break f4) and as the first step toward my kingside attack. I should maneuver a Knight to g6 (behind the g pawn) preparing the Benoni jump, which I will further prepare by bringing at least one or two rooks to the g-file. Finally I will sac on f4, crash through, and mate him!
For inspiration, before I tried to find the exact move order, I looked at the several examples Kmoch gave of the Benoni jump. This one was the most resonant:
C Alexander -Pachman
Hilversum 1947
1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0–0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0–0 9.h3 Na5 10.Bc2 c5 11.d4 Qc7 12.Nbd2 Bb7 13.d5 Bc8 14.Nf1 Re8 15.Kh2 g6 16.Ne3 Bf8 17.g4 Bg7 18.Rg1 Kh8 19.Ng5 Rf8 20.h4 Ng8 21.Qe2 Bd7 22.Bd2 Ne7 23.Nf5!! (the Benoni jump strikes!) 23…gxf5 (Black foolishly accepts, he should try Kmoch’s …Ng8 or Fritz’s …Nf5) 24.gxf5 f6 25.Nxh7 (White crashes through) 25…Be8 (25...Kxh7 26.Qh5+ Kg8 27.Rxg7+ Kxg7 28.Rg1+ and mates) 26.Rxg7 Kxg7 27.Nxf8 Kxf8 28.Bh6+ Kf7 29.Qh5+ Ng6 30.fxg6+ Kg8 31.Qf5 Qe7 32.Rg1 Nc4 33.Bc1 Bd7 34.Qf3 Rf8 35.b3 Nb6 36.h5 f5 37.Bg5 fxe4 38.Qe2 Qe8 39.Bxe4 Bf5 40.Bh6 Rf6 41.Qf3 and Black resigned.
Now back to my analysis: with Hans Kmoch (who died in 1973) whispering in my ear, it was not difficult to find the correct moves, and incidentally, completely baffle Mr. Fritz, who found my moves inexplicable until he was mated.
This is how Black should play after 17.c4:
17...Nh7 (heading for g6) 18.Bg2 Nf8 19.Kh1 g5 (the attacking setup is underway, and White is blocked from achieving f4) 20.Ng1 Ng6 21.Rf1 Kh7 22.h3 Be6 23.Ndf3 Rg8 (here come the rooks – White is too cramped to get effective counterplay) 24.Rfb1 Rg7 25.Bc1 Rdg8 26.Qd1 f6 27.Bd2 Qe7 28.b4 (Fritz lashed out with this move, hoping to get some kind of activity) axb4 29.axb4 Bxb4 30.Bxb4 Nxb4 31.c5 Na2 32.Ne2 Nc3 33.Nxc3 dxc3 34.c6 bxc6 35.Rc1. The computer thinks all is in order here, and gives itself an equal evaluation. When asked to consider moves for Black, it gives as its top three 35…Qd7, 35…Rd8, and 35…Rf7, in other words, three “nothing” moves. But everything is set up, and I had a dead guy in my corner: Kmoch said, “Now!” So I played 35…Nf4!! and Fritz blew a chip! 36.gxf4 The computer says White is clearly better! But then 36...gxf4 and every Black piece attacks. At first, Fritz tried to cling to a White advantage, but finally saw the error of its ways: here is where the computer is helpful. With some attacking help from me (I found Black’s moves, the computer found the best defenses) the machine finally worked out that White is busted in all variations (really, there are only two defenses, Bf1 and Ne1).
Here is our joint human/computer analysis:
A. 37.Ne1 Rxg2 38.Nxg2 Bxh3 39.Ne1 (39.Nh4 f5 40.Qh5 [40.Nxf5 Bxf5 with the idea Qh4 mate] 40...Bg4 wins the Queen) 39...Qg7 and mates.
B. 37.Bf1 Qf7 38.Rxc3 (38.Ne1 Rg1+ 39.Kh2 Qg7 40.Qf3 [40.Qh5 Rh1+] 40...Rxf1–+) 38...Qh5 39.d4 Bg4 (39...Bxh3 is premature in view of 40.Ng5+) 40.dxe5 Kh8 41.exf6 (41.Rb1 Bxh3–+) 41...Rd7 42.Qe2 Bxh3 43.Nh4 Bg2+! and wins.
Of course this long variation is not forced, but it clearly shows the ideas of the position.
I have complete confidence that if I had seen the idea of setting up the Benoni jump, White would have been under such intolerable pressure that I would almost certainly have won the game – but I played on the wrong side of the board (the name Hans Kmoch never crossed my mind!) and lost as follows.
17…Nb8 18.Bg2 Bf8 19.Kh1 c5 20.Ng1 g5 21.Nf1 b5
Ultimately this only gives White a square (c4) for his Knight. Better is 21...Nc6 with the idea (still!) of the Benoni jump with …Ne7-g6 and Black is better – unfortunately I consistently stuck with my erroneous plan!
22.cxb5 Qxb5 23.Bh3 Nc6 24.Nd2 Rb8 25.Reb1 Qb7 26.Bc1 Qd7 27.Kg2 Rb7 28.Nc4 Reb8 29.Bd2 Qd8 30.Rb2 Bd7 31.Bxd7 Nxd7 32.f4
White achieves this important advance.
32...gxf4 33.gxf4 a4 34.Rab1 Qh4 35.f5 Be7 36.bxa4 Kh7?
In mutual time trouble I make a completely unsound rook offer.
37.Be1?
Quicker is 37.Rxb7 Rg8+ 38.Bg5! Qxg5+ (38...Bxg5 39.Rxd7+-; 38...Rxg5+ 39.Kh1+-) 39.Kh1+-.
37...Rxb2 38.Nxb2 Qg4+ 39.Kh1 Rg8 40.Bg3 h5 41.Qd1
The time control has been reached.
41...Qg5 42.Nh3 Qh6 43.Nc4
The White Knight reaches this wonderful “gift” square: White is clearly better.
43…Nf6 44.Rb6 h4 45.Bxe5 Ng4 46.Qf1 Rc8 47.Bf4 Qg7 48.Qe2 f6 49.Rb1 Rg8 50.Rg1 Nce5 51.Nxe5 fxe5 52.Bg5 Bxg5 53.Qxg4 Bh6 54.Qxg7+ Rxg7 55.Rxg7+ Kxg7 56.Kg2 Bc1 57.a5 c4 58.a6 c3 59.a7 Be3 60.a8=Q c2 61.Qb7+, 1–0.
There is a simple moral to this debacle: read your chess books! Especially, read your old chess books!
The computer can’t help you with strategic thinking – but a book can. Furthermore, by naming ideas (like “Benoni Jump”) the author helps the reader to deal with these otherwise abstract concepts. A computer might be powerful enough (for example, Hydra) to find a Benoni jump on its own, but even if you had your personal Hydra, it would not explain the ideas to you.
For that you need a person, like the late Mr. Kmoch. His slightly stiff, German accented English will stay in your mind (I wish it had stayed in mine, but it’s back now!).
We’ve all had this problem: I know I’m better, but what the hell do I do? The answer is probably not on your computer screen. The answer may well be in a dusty book.
And now, a quick preview of coming attractions: in next month’s column, the final one of this arc, The Difficult Opponent, I will show my tenth game against Matikozyan: and my first win!!
