Page1 I page2

Randy, upon reflection: Bent Larsen once made a perceptive comment that a player must always examine every position with a fresh perspective, without regard to the moves played previously. Here black is still under the impression that taking the pawn is inadvisable, but, as Yermo shows, this isn't necessarily the case.

24.e5 Nbd7

Randy: Black is going to prepare play on the b-file so gets the knight off the file. He also prevents an intrusive Bc5.

GM Yermolinsky: Black has a slight advantage.

Randy: Black needs counterplay. This hits the weak f-pawn and envisions play with, perhaps, ...a5-a4.  The absence of the bishop from e3 makes this possible.

Randy: Understandable, but I think that Mike just missed black's idea. White wants to free his rook on d1 for offensive action, and 27.Be3 means that the rook must stay on the d-file because of the threat of ...d5-d4.  27.Bg3 turns the bishop into a big pawn, but now …

FANTASTIC!

GM Yermolinsky: Fantastic concept!

Randy: Another fascinating move that I hadn't considered, yet Mike played it fairly quickly. At this point I had about 20 minutes and Mike 15 to get to move 45, and I didn't spend too much time thinking about my response. Objectively, I felt that I had a good position and didn't want to subject myself to a lightning attack. It's not entirely clear that black cannot take the exchange, but the attack is easier to play. After 30...Bxd4 31.Bxd4 e5 white has several ways to pursue the attack.

After 32.Qh8+ Kf7 33.Rh7+ black may be able to defend with 33...Ke8 34.Bc5 Qd8 when the threat to the g5 pawn is hard for white to meet, and winning the queen by 35.Re7+ Qxe7 36.Bxe7 Kxe7 looks better for black. A better try for white seems to be the immediate 32.Bc5, the point being that 32...Nxc5? 33.Qh8+ Kf7 34.Rh7+ wins the queen under much better circumstances for white. After 32.Bc5 Qd8 33.Qh8+ Kf7 34.Rh7+! Nxh7 (34...Ke6 35.Qg8 mate) 35.Qxh7+ Ke6 36.Qxg6+ Nf6 37.Bd3! e5 38.Bf1! threatens both Bd4 and Bh3.  Still, there may be a better defense for black in there somewhere.

GM Yermolinsky: 30.Bd4 Bxd4+ 31.Rxd4 Qe5.

30...Qe5 31.Bf1 Qg7 32.Rd3 Be5 33.Rh1 Bd6 34.R3h3

BLACK MUST MAKE A DIFFICULT DECISION

34...Rb8

Randy: Black decides to take a tempo to get his rook on a square covered by his minor pieces to prevent tactical ideas based on winning the rook at the tail end of a checking combination. Still, there were some other ways for black to pursue the position. Black could play, for example, 34...a4, intending to meet 35.bxa4 with 35...Nc5.

GM Yermolinsky: 34...e5 35.Qh8+ Qxh8 36.Rxh8+ Kf7 37.Rah7+ Nxh7 38.Rxa8 Nxg5 39.Rxa5 e4 with a clear advantage for black.

Randy, upon reflection: John Nunn talks about lazy moves, and this is, as Yermo's analysis shows, an example of one. While it's understandable that black wants to remove those troublesome tactics based on the undefended rook, it is not necessary at this particular point in the game.  The tactics don't work for white, and black should be seeking ways to develop his play, not respond to threats that aren't there yet.

Randy: White needs to soften up the strong black pawn center to get his light-squared bishop involved. This concept is something that black will soon fail to grasp, to his detriment.

35...Nc5 36.Be3 a4

Yermolinsky: Maybe you shouldn't have given him a passer (36...Ne4).

Randy, upon reflection: Actually, I recall being moderately surprised by this move. I think that, for whatever reason, I had a mental block and didn't think that white could play 38.b4.

37.cxd5 cxd5 38.b4 Ne4 39.b5

A CRITICAL POSITION

Randy: This is the critical position. White has succeeded in creating play somewhere other than the h-file, which black has effectively defused. Indeed, both white bishops are now engaged and he has a passed b-pawn. Black, however, has also helped himself. His knight on e4 is very strong, his king is relatively safe (note how well the knight on f8 holds white at bay on the h-file), and with the white big guns on the h-file, it was time for black to turn his attention to a real problem for white, the a-pawn.

Randy: With about 3 minutes to get to the time control, black falters. A much better idea is 39...Qa1!, which threatens both ...Rxb5 and ...Qxa2. Black would intend to play ...Be5 next, followed by taking the a-pawn, when his a4-pawn would be very hard to stop. This may force white into 40.Qh8+ Qxh8 41.Rxh8+ Kf7, but the trade of queens would cut down on white's attacking chances, and black would seem to have the better chances in that endgame.

Black's move weakens the central pawn bulwark that had been so effective in holding together the light squares. By recognizing their vulnerability, white quickly takes control.

GM Yermolinsky: You give this move a question mark, but I think you made some mistakes later on. The move e6-e5 is a bit loosening, but how are you going to win this game if not with the pawns?

Randy, upon reflection: Grandmaster Yermolinsky did not have my notes when he looked at the game, so I don't know if he considered my suggestion of 39...Qa1 or not. After I sent a follow-up query about this position, Yermolinsky was good enough to have Boris Men, another member of the Yermo Chess Academy and former participant in the U.S. Championship, take a look at the game.  This was his response:

Boris Men: About 39...Qa1.  Alex asked me to look at your game too. I didn't follow his analysis. In my opinion, this position is winning for black. You simply must keep track of the h8-square and a1-h8 diagonal. Say, for example, 39...Qa1 40.Kg1 Be5 (not 40...Rxb5? 41.Qh8+) 41.b6 Qxa2 42.Bf4 Bd4+ 43.Be3 Qb2. I would not go 39...e5 in order to keep control over this artery a1-h8. And of course, your pawns d5, e6, f5 along with your Ne4 is a guarantee of safety. I would not consider altering the structure.

Randy: This is an instructive exchange. To my way of thinking, the point is that, in chess, there are many positions where there will be a legitimate disagreement among strong players about the best plan. I think that both the move played and the suggested 39...Qa1 are good for black in this position – a sure sign that things are going well for the second player.

Randy: Hitting black's Achilles heel.

41...Qf7 42.Bf2

BLACK TO MOVE AND WIN

42...Nxf2

GM Yermolinsky: I understand that you wanted to simplify, but there was a better move – 42...Nxg5.  You must have missed that one. After 43.Rd3 Rxb5 44.Qxa4 Qb7 black has a winning advantage.

Randy, upon reflection: Yermo is pointing out another common flaw in the thinking of the lower-rated player – that exchanges should always be welcomed. As noted, black should, with active minor pieces, not shrink away from the material gain. Over the next couple of moves black throws away a very good game.

GM Yermolinsky: What was the time control? You play leaves an impression of a time scramble.

Randy, upon reflection: Yes, indeed, both players had less than a minute to get to move 45, and black suffers from the old "duffer sees a check, duffer gives a check" syndrome.  Still, black totally falls apart at the end, and the clock shouldn't be the only excuse.

44.Kg2 Qf7

GM Yermolinsky: 44...Qc5 and black is still better!

BLACK IS STILL IN CHARGE!

Randy, upon reflection: This is amazing to me. I had, during the game and while annotating it, already given black up for lost. The point, of course, is that black's pawns are very threatening. If white would play 45.Qxa4 e4! looks very promising for black. I became fixated by the number of pawns, when actually the advanced nature of the black pawns is much more important than whether he has one or two pawns for the exchange.  This is, of course, the type of thing that separates the great players from the not so great. 

45.Bc4! dxc4 46.Qxd6

Randy: Things have changed dramatically, and white's attack is now overwhelming.   After 46...Rxb5 47.Rh8+ Kg7 48.Qxf8+! Qxf8 49.R1h7 mate.

Yermolinsky: 46...Rxb5 47.Rh8+ Kg7 48.Qxf8+ Qxf8 49.R1h7 mate.

46...Qb7+? 47.Kf1, 1-0.

Randy: The moral of the story: seize your opportunities, but remember that pawns can't move backward.

Randy, upon reflection: Grandmaster Yermolinsky makes a telling point that black's mistakes didn't necessarily come from the pawn advances but from not seizing other tactical opportunities that presented themselves along the way.

In many ways, this is a remarkable insight for me. I have always felt that I let this game get away with loosening play. In reality, it was from a lack of analytic precision, and maybe a bit of chess laziness sprinkled with over-respect for my opponent. This has given me some things to work on before my next meeting with an internationally titled opponent.

Page1 I page2