Réti
1.Nf3 d5 2.c4 e6 3.b3!?
Although 3.g3 is more associated with Richard Réti, the text seems to have been first played by Réti in a game in 1923, the same year as he first played 3.g3.
3...Nf6 4.Bb2 Be7 5.g3 0-0 6.Bg2 c5 7.e3 Nc6 8.0-0
 |
| Now both sides have castled, how would you assess the position? |
*****
*****
*****
*****
The position occurs 4,488 times in ChessBase's 2025 Mega database, with White scoring a resounding 61%. However, in those games that were rated, which were the vast majority, whites turned in a 2282 performance, marginally down on their average rating of 2284.
8...d4!? 9.exd4 cxd4 10.Re1 Re8 11.Na3!?
Fabiano Caruana is among grandmasters who have played this, but it is not liked by Stockfish17 or Dragon1. They suggest 11.d3 or 11.a3!?
11...Bc5 12.d3 e5 13.Nc2 Bg4
The engines prefer 13...Qd6 or 13...a5. The latter occurred in Caruana (2786) - Theodore Slade (2027), Chess.com Blitz 2023, which continued 14.Ng5 Bf5 15.a3 Qc7 16.Bc1 h6 17.Ne4 Nxe4 18.Bxe4 Bxe4 19.dxe4 f5!?, when Black had the upper hand, according to the engines (but 1-0, 38 moves).
14.h3 Bh5 15.g4!?
This thrust is nearly always controversial, but here it is engine-approved.
15...Bg6 16.Nh4 Qb6 17.Nxg6 hxg6
 |
| White has won the bishop-pair, but Black's central pawn-duo means the white dark-square bishop has very little scope |
*****
*****
*****
*****
18.Qf3!?
The engines strongly dislike this, suggesting 18.g5 Nh5 19.Bc1!?
18...a5 19.Rab1 Rad8 20.g5 Nh5 21.Bc1 Bb4!? 22.Nxb4 Qxb4 23.Re2 Qd6 24.Re4
White is slightly better after 24.a3, according to the engines.
24...Nb4!?
 |
| How should White proceed? |
*****
*****
*****
*****
25.a3?
Stockfish17 gives 25.Qe2, while Dragon1 marginally prefers 25.Ra1.
25...Na2
But not 25...Nc2?? as the knight is lost after 26.Qe2 Nxa3 27.Ra1.
26.b4!?
Possibly slightly better is 26.Re1, but 26...Nxc1 27.Rbxc1 Qxa3 wins a pawn, at least temporarily, and more importantly gains the f4 square for the king's knight, and anyway the engines' 26...e4!? may be even stronger. Probably worse is 26.Bb2?! Nc3 27.Bxc3 dxc3.
26...axb4 27.axb4 Nc3 28.c5?!
The engines much prefer 28.Rb2 or 28.Rb3.
28...Qc6
Even stronger seems to be 28...Qa6.
29.Ra1 Nxe4 30.Qxe4 Qxe4?
Black is still winning after 30...Qb5 or 30...Ra8, according to the engines.
31.Bxe4 Ra8 32.Rb1
Not 32.Rxa8?? Rxa8 33.Bxb7 Ra1 34.c6 Rxc1+ 35.Kh2 Nf4 36.b5 Nd5.
32...Ra7 33.b5!?
The engines at first prefer 33.Kf1, but after being shown the text it comes to be Dragon1's top choice, although Stockfish17 sticks with 33.Kf1.
33...Rc8
Stockfish17 reckons 33...Rea8!? is winning, but Dragon1 claims it leaves the game completely equal. Both engines reckon 33...Rb8 gives Black a slight edge, while they agree the text leads to equality.
34.b6 Ra5 35.Bxb7 Rcxc5 36.Be4 Rcb5 37.Rxb5 Rxb5 38.b7
 |
| This position lends weight to Siegbert Tarrasch's opinion that two bishops are as strong as a rook and a knight |
*****
*****
*****
*****
38...Nf4!
This is the only move that stops White gaining a substantial advantage, eg 38...Rb1 is easily met by 39.Kh2, when Black cannot prevent the dark-square bishop arriving at d6 via a3, while 38...f5? is even worse as 39.Bd5+ forces the black king to the h file, since 39...Kf8?? 40.Ba3+ loses the rook to 40...Ke8 41.Bc6+ etc.
39.Bxf4
If 39.Ba3, then 39...Nxh3+ 40.Kg2 Nf4+ 41.Kf3 Rb3!? (41...Ne6 may also hold) 42.Bd6 f5 43.b8=Q+ (43.Bc6?? Rxd3#) Rxb8 44.Bxb8 fxe4+ 45.Kxe4 Nh3 draws. The engines point out that after 43.gxf6 gxf6, White still has nothing better than b8=Q+. And if 40.Kh2, then 40...Nxf2 41.Bc6 (forced) Rb1 42.Bd6 Nxd3 43.b8=Q+ Rxb8 44.Bxb8 is also a draw, according to the engines.
39...exf4 40.h4 f5 41.gxf6 gxf6 42.Kg2 g5
 |
| How should White proceed? |
*****
*****
*****
*****
43.h5?
Simplest is 43.hxg5 fxg5 44.Kf3, with complete equality, according to the engines, but other moves also draw, including 43.Kh3 and 43.Bc6!?
43...Kg7 44.Kf3
 |
| Black to play and win |
*****
*****
*****
*****
*****
44...Kh6
Winning is 44...f5! After 45.Bxf5 Rxb7 the rook is free to leave the b file, and the position becomes like any normal exchange-up ending, as will be seen later in the game. White can try 45.Bc6, but the rook simply moves along the b file, and then the black king will pick up the h pawn.
 |
| White to play and draw |
*****
*****
*****
*****
45.Kg4
Drawing is 45.Bc6! Indeed Black has to be careful not to lose, eg 45...Rb4 46.Ke4 Kxh5? 47.Kd5 is a simple win for White, one line running 47...g4 48.Kc5 Rb2 49.Bb5 etc. However, 46...Rb1! seems to hold, and the draw is simpler for Black after 45...Rb1 or 45...Rb2.
 |
| Black to play and win |
*****
*****
*****
*****
45...f5+
This is a straightforward win now that White has no alternative to capturing the pawn.
46.Bxf5 Rxb7 47.Bg6 Rb2 48.Kf3 Kg7 49.Be4 Rb5 50.Kg4 Kh6 51.Bg6 Rb1 52.Kf3 Ke1 53.Bf7 Kg7 54.Bd5 Kf6 55.Be4 Rh1 56.Bg6 Rh2 57.Be8 Kf5 58.Bg6+ Ke5 59.Bf7
Or 59.Kg4 Rxf2 60.Kxg5 Rxg2+ 61.Kh4 f3 62.Kh3 Kf4 63.Be4 Rg3+ 64.Kh2 Rg5 etc.
59...Rh4
It has taken longer than it should, but at last the black kingside pawns are ready to advance.
60.Bg6 g4+ 61.Kg2 Rh3 62.Kg1 f3 63.Be8 Kf4 64.Bf7 g3 65.fxg3+ Kxg3 66.Bg6 f2+ 0-1
No comments:
Post a Comment