Greatest ever game by a non-master!

My friend IM Anthony Saidy just sent me this old article by another friend of mine IM Jeremy Silman. Hope you enjoy it:

THE MASKED GRANDMASTER REDUX
IM Jeremy Silman

Recently a fan of combinative chess wrote me and asked where he might find a game that I wrote about back in 1974. It isn’t in any database, which is a real shame since the game in question is very memorable. Since this game is highly enjoyable, I will repeat my Chess Life article here (with various major analytical changes and additional information at the end).

It was Oct. 26, 1974 and the Carrol M. Capps Memorial chess tournament in San Francisco was just starting. Aside from the usual list of prizes, the event also had a “most brilliant game” award.

This prize had little meaning for the lower-rated players though. After all, how could non-masters players hope to play a more brilliant game than powerhouses like Walter Browne, Dennis Waterman, and R. Rodriguez?

“Look at my game! Look at my game! I’ve played a brilliancy!” screamed Michael Mills, a class “C” player.

Senior Master Waterman, IM (and U.S. Champion) John Grefe and I stared at him and then we all burst into laughter. “You played a brilliancy? Yeah, sure you did. Okay, let’s see it.”

Mills, seemingly oblivious to the insulting tone of our voices, sat down and began pushing wood:

R.Catig (1500) – M.Mills (1500)

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 g6 5.Nc3 Bg7 6.Be3 Nf6 7.Be2 0-0 8.Qd2?! d5!

So far, we were all silent. Black has played the opening well and has no problems.

9.Nxc6

The continuation 9.exd5 Nxd5 10.Nxd5 Nxd4! favors Black: 11.Nxe7+ (11.c4 keeps black’s edge to a minimum while 11.Bxd4 Qxd5 12.Bxg7 Qxg2! has led to many Black victories) 11…Qxe7 12.Bxd4 Bxd4 13.Qxd4 Bf5 (13…Re8 14.Qe3! does not promise Black as much) 14.Qe3 Qb4+ 15.Qc3 Qxc3+ 16.bxc3 Rfd8 and black’s superior structure promises White a certain measure of endgame pain.

9…bxc6 10.e5 Ng4

A good response. Also possible is 10…Nd7 11.f4 e6 when the following trap has claimed many victims: 12.Na4? Nxe5! 13.fxe5 Qh4+ 14.Bf2 Qxa4.

11.Bxg4 Bxg4 12.h3 Bf5 13.g4?

Desperate to find something to criticize, we all became hysterical. “You fool!” we howled, “Why did you allow him to attack your Bishop with gain of time?”

“Well,” Michael replied coolly, “I was trying to egg him on.”

This was too much for us. We fell on the floor and laughed uncontrollably. Undaunted, Michael continued …

13…Be6

We were too busy making fun of Michael to notice 13…Bxe5! 14.Bh6 (14.gxf5 d4) 14…d4! 15.Bxf8 (or 15.Qe2 Bf6 16.Bxf8 dxc3 17.b4 Qd2+ 18.Qxd2 cxd2+ 19.Kxd2 Kxf8) 15…dxc3 and white’s position is grim.

14.Qd4?

The idea of this move its to post the Queen on c5 — not a bad concept, but it walks into various tactical problems. Far better was 14.f4.

14…f6

A good move, but interesting alternatives existed:

1) 14…Qb8 (with a double attack against b2 and e5) 15.f4 f6 is tempting.

2) 14…Rc8!? is a computer move: 15.Qxa7 (15.Na4 Qc7) 15…Bxe5 16.Bh6 d4 17.Rd1 Ra8 18.Qc5 Ra5 19.Qxc6 Qb8 20.Bxf8 Kxf8 and White is getting wiped out.

15.f4?

Very poor. White should play 15.exf6 Bxf6 16.Qc5 though Black would still retain an obvious advantage. For example 16…Qb6 (16…d4!? 17.0-0-0 dxc3 18.Rxd8 cxb2+ 19.Kb1 Raxd8 looks like it might be interesting, though it’s hardly brave of me to sacrifice someone else’s Queen!) 17.0-0-0 Rfb8.

15…Qc7

The obvious 15…fxe5 16.fxe5 Qc7 nets a free pawn since 17.Bf4 c5 is crushing. However, the text move also leaves White in a bad way, and might even prove stronger than 15…fxe5.

16.exf6 Bxf6 17.Qc5 Bh4+

More straightforward is 17…d4 18.Bxd4 Qxf4 19.Bxf6 Rxf6 with a winning attack. The path Mills chose is far deeper and far more elegant.

18.Ke2

Other moves:

1) 18.Kd1 d4! 19.Nb5 Qa5 20.Qxc6 dxe3 21.Qxe6+ Kg7 22.Nc3 Rxf4 wins for Black.

2) 18.Bf2 Bxf2+ 19.Qxf2 Rxf4 20.Qe3 Bf7 21.0-0-0 e5 has to be winning for Black.

3) 18.Kd2 d4 wins on the spot.

18…Bc8!!

A first-rate move that is even beyond the powers of modern day (2008) computers. Black brings his light-squared Bishop into play and makes way for the advance of his e-pawn, which will rip open the center. Michael’s earlier moves had not made much of an impression on us. But when we saw this move, our pompous smiles began to fade.

19.Nxd5

Very tempting and very greedy, but there really isn’t a fully acceptable defense. Other tries:

1) 19.Kd2 Ba6 (intending both …d4 and …Rxf4) 20.Rad1 (20.Qd4 Bf6!) 20…Rxf4 21.Bxf4?? Qxf4+ 22.Qe3 Qf6 and White has to resign due to threats like …d4 and …Bg5.
Instead of 21.Bxf4??, White should play 21.Nxd5 Rd8 22.Kc1 Rxd5 23.Rxd5 Rc4 24.Qxa7 cxd5 25.Qxa6 Rxc2+ 26.Kb1 Bf6 (26…Rxb2+!? 27.Kxb2 Qe5+ 28.Kc2 Qe4+ 29.Qd3 Qxh1 30.a4 gives White counterplay) 27.Bc1 Rg2! 28.Qb5 Qc2+ 29.Ka1 Kg7 and White’s position is most unpleasant.

One would think that White should go for this, except Black can improve: 19.Kd2 Ba6 20.Rad1 (His only chance is probably 20.Kc1!? though White’s position would then be wretched in every way after 20…e5) 20…e5! 21.fxe5 (21.Nxd5 Qb7 22.Nb4 Be7 is more than White can handle) 21…d4! 22.Bxd4 Bg5+ 23.Ke1 (23.Be3 Be7 and White will suffer massive material losses) 23…Rad8 and Black’s attack should prove decisive. One nice line: 24.Qxa7 Qxe5+! 25.Bxe5 Bh4+ 26.Bg3 Bxg3+ 27.Qf2 Bxf2 mate. Cool!

2) 19.b4 e5 20.fxe5 Qxe5 looks grim.

3) 19.Rag1 Ba6+ 20.Kd1 e5 and all I can say is that I wouldn’t want to be White.

19…Ba6+ 20.c4

Black has many possibilities after 20.Kf3, the simplest of which is 20…Qe5 when White has to jettison a piece by 21.Nxe7+ (21.Qxc6 Rac8 is game over) 21…Qxe7 22.Qxe7 Bxe7.

20…Qb7! 21.Nb4 e5!

White’s holding on as best he can, but Mills (who appears to be channeling Alekhine) won’t stop playing great moves!

22.Nxa6

If 22.fxe5 Be7 23.Qxc6 Qxb4 (much stronger than 23…Bxb4 — why not enjoy an extra piece AND a strong attack?) 24.b3 Bb7 25.Qe6+ Kh8! 26.Rhf1 Bg2 and wins.

22…exf4?!

Black learned that he’s supposed to open lines in this kind of position and he’s making sure he does it! Unfortunately, more direct measures were called for: 22…Qxb2+ 23.Kf3 (23.Bd2 holds out a bit longer, though 23…Rxf4 leaves White in agony) 23…e4+ 24.Kxe4 Rae8+ 25.Kd3 (25.Kf3 Rxf4+! mates) 25…Rd8+ and mate is in the air.

23.Bd4 Rae8+?!

The most natural move in the world, but it turns out to be inaccurate. Instead, 23…f3+ turns out to be quite strong: 24.Kd3 Rad8 25.Nb4 Be7 (the immediate 25…Rxd4+ is also good) 26.Qe5 Rxd4+ 27.Qxd4 Qxb4 28.Qc3 Qc5 29.Kc2 Rf4 30.b4 Qf2+ 31.Kb3 Bf6 32.Raf1 Qe2 33.Qc2 Qe3+ 34.Ka4 f2 gives Black a winning advantage.

24.Kf3??

Kindly allowing a stunning finish. White had to play 24.Kd3 when 24…Qxa6 25.Qxa7 Qxa7 26.Bxa7 f3 27.Bc5 Rd8+ 28.Ke4 Rf7 still leaves White in serious trouble due to the power of the passed f-pawn and the vulnerability of the white King. One sample: 29.Ba3 f2 30.Raf1 Re8+ 31.Kd3 Rd7+ 32.Kc2 Re2+ 33.Kb3 Bf6 and White is helpless.

After this final mistake by White, Mills really does turn into Alekhine!

24…Re3+!!

Still clearing lines like a maniac! At this point all of us were exhibiting signs of shock, jaws hanging to the floor.

25.Kg2

If 25.Bxe3 fxe3+ 26.Ke4 leaves Black with a multitude of winning ideas, with 26…Re8+ 27.Kd3 Rd8+ 28.Ke4 Qxb2 being my personal favorite.

25…f3+ 26.Kf1 Rfe8!!

Black now threatens to mate with …Re1+. Of course, White cannot play 27.Bxe3 due to …Qxb2.

27.Kg1 Bg3!

Tightening the net. The Rook is still immune to capture — an incredible situation. Of course, 27…Re1+ also won easily, but 27…Bg3 is more accurate.

28.Rf1 Re1 29.Bc3 Qxb2!!

Now, since 30.Bxb2 Rxf1+ 31.Kxf1 Re1 is mate, White resigned. 0-1.

“Who was that masked grandmaster?” Larry Christiansen asked when I showed him the game some months later.

How did Mills do it? Was it something he ate? Did he practice celibacy for this event?

Michael was quite willing to give us the answer. “I just followed the advice given in Vukovic’s book, THE ART OF ATTACK.”

And it suddenly all made sense! I had told him to read this classic months before the tournament and he had taken the advice to heart. His use of basic tactical motifs like double attack, the opening of lines, the overworked piece, and building a mating net were all beautifully featured in this game.

As it turned out, he really didn’t calculate that many variations. He just did what Vukovic told him to in such positions and hoped it would work out. Therefore, this game is both a testament to Vukovic’s book as well as a final proof of the validity of the Hundred Monkey Theory. Nevertheless, few players (of any rating) ever create an evergreen such as this, so he can consider himself blessed.

Of course, Mills won the brilliancy prize, and none of us could do anything but applaud him. Truly a fantastic creative effort and, perhaps, the greatest game by a non-master of all time!