World Open (Philadelphia) 1994
Modern Defence
1.e4 g6 2.d4 Bg7 3.Nc3 d6 4.Be3
I am rather surprised to find this is by far the most-popular choice in ChessBase's 2020 Mega database.
4...a6 5.a4
But here the most-popular move is 5.Qd2, when the idea is often to follow-up with Bh6 as soon as Black develops his king's knight.
5...Nc6
Not particularly popular in Mega20, but it turns out to be Stockfish10's pick.
6.Qd2 e5 7.dxe5!?
Not in Mega20, but it is Stockfish10's choice. 7.d5, which is liked by Komodo10, was played in four of the eight games to reach this point.
7...Nxe5 8.Bg5!?
Komodo10's choice. Stockfish10 likes 8.h3.
8...Bf6?!
This seems strange, and I am not surprised the engines much-prefer 8...Qd7 or 8...f6.
9.Bxf6 Nxf6 10.f4 Nc6 11.Nf3 Bg4 12.Be2
The engines want White to go for it with 12.0-0-0.
12...Qe7 13.0-0
Now the engines prefer this to long castling.
13...0-0-0
The engines reckon White is better, but it is a difficult position to judge. I think opposite-side castling, unless one side is clearly much better, has the practical effect of favouring the stronger player, who is more likely to make correct positional decisions about offence and defence.
14.Nd5!?
An interesting move that forces Black's reply
14...Nxd5
Not 14...Qe6?? 15.Ng5.
15.exd5 Nb8 16.Rfe1 Qf6 17.b4 Rde8
 |
| How should White proceed? |
*****
*****
*****
*****
18.b5?!
This is Komodo10's second choice, but it allows Black to keep more lines closed on the queenside than would be the case if White played a5 before b5, although the immediate 18.a5!? runs into 18...Re4 with an unclear position.
The engines give 18.h3 Bxf3 19.Bxf3 Nd7 20.a5 with a small edge for White.
18...a5 19.b6!?
The engines dislike this at first, but come to quite like it if given enough time.
19...cxb6 20.Bb5 Rxe1+ 21.Rxe1 Bxf3 22.gxf3 Nd7 23.Bxd7+!?
This seems correct - White has a lot of pawns on the same-coloured squares as the bishop, and Black's knight would be unchallengeable on c5.
23...Kxd7 24.Qe3 Qd8 25.Rb1 Re8 26.Qxb6 Qxb6 27.Rxb6
White has got back the pawn back sacrificed on move 19, and exchanged off into a rook-and-pawn ending, but the pawn-structure means Black is much better.
27...Kc7 28.Rb5 Re2!?
Better may be 28...b6, eg 29.Kf2 (to stop the black rook invading) Kb7 30.Rb2 Rc8, when White must lose a pawn.
29.Rxa5 Rxc2
The engines prefer 29...Kb8, keeping the white rook away from a8 where it can support an advance of the a pawn as well as swoop along the back rank.
30.Ra8 Rc4 31.f5!?
This seems better than allowing ...Rxf4.
31...Rf4 32.fxg6 hxg6 33.Kf2 Kb6 34.Ke3
This is OK, but according to the engines White is fully equal after 34.a5+ Kc5 35.Kg3, eg 35...Rb4 36.Rf8 Kxd5 37.Rxf7.
34...g5
The engines reckon Black is slightly better after 34...Rb4 35.a5+ Kc5 36.Rf8 f5.
35.a5+ Kc5 36.a6?!
Almost certainly better is the engines' 36.Rb8 Rb4 37.Rf8 Kxd5 38.Rxf7.
36...bxa6 37.Rxa6 Rf5 38.Ra5+ Kc4 39.Ra4+ Kxd5 40.h4 Re5+ 41.Kd3 gxh4 42.Rxh4 Re1 43.Rh5+ Ke6 44.Rh6+ f6 45.Kd2 Ra1
Petursson said afterwards he did not know if this end were winning for Black, but he did know how to play it.
46.Ke3 Ra3+ 47.Kf4 Ra4+ 48.Ke3
The engines' somewhat counter-intuitive 48.Kg3! is better. If then, as in the game, 48...Ke5, White can play 49.Rh5+, meeting 49...f5 with 50.Rh8, when 50...f4+ 51.Kg4 seems completely equal. Black should probably instead play 48...d5, still with a slight edge.
48...Ke5 49.Rh8
The point is that now 49.Rh5+ f5 50.Rh8 f4+ drives the white king to the second rank, either immediately or after 51.Kd3 Ra3+.
49...Ra3+ 50.Ke2 f5 51.Re8+?
This drives the black king to where it wants to go, but White's game seems beyond saving anyway.
51...Kf4 52.Rd8 Re3+ 53.Kd2 Re6 54.Rf8 Re5 55.Rd8 d5 56.Kd3 Kxf3
Black has won a second pawn (0-1, 65 moves).
LESSONS FROM THIS GAME
Simple moves can allow White to keep a slight edge against the Modern Defence. But the defence is naturally unbalancing and may easily allow Black to go for lines which, while objectively slightly inferior, require a lot of positonal understanding.
I probably mishandled White's queenside play, which should have been focused on opening lines. Instead I allowed Black to keep lines closed.
Black reached a rook-and-pawn ending in which he was much better. But the tricky nature of such endings is shown by the fact that the grandmaster allowed his opponent equalising chances, although these were not taken due to not fully understanding the nuances.
No comments:
Post a Comment