I pointed out how Mark Hebden had made a living out of it on the weekend circuit, and how it used to be a favourite with Garry Kasparov.
But that was quite a few years ago, or as the authoritative source Wikipedia puts it: "In the early 2000s the opening's popularity suffered after Vladimir Kramnik began scoring excellently against it; so much so that Kasparov gave up the opening after several critical losses to Kramnik."
However, that may be about to change after Magnus Carlsen used the defence to beat Dutch grandmaster Max Warmerdam in Germany's Bundesliga.
Notes in italics are by international master Malcolm Pein in Chess magazine.
Warmerdam (Solingen 2646) - Carlsen (St Pauli 2831)
English/Modern/King's Indian
1.c4 e5 2.g3 g6
At this point ChessBase classifies the opening as "A20: English Opening: 1...e5."
3.d4!?
This is much less popular than 3.Bg2, but scores 15 percentage points better in the 2025 Mega database, and is preferred by Stockfish17 and Dragon1.
3...d6!?
A provocative reply.
4.Nf3!?
It must have been very tempting to proceed 4.dxe5 dxe5 5.Qxd8+, which is how Yasser Seirawan beat Anatoly Karpov at Haninge (Sweden) 1990, although the engines reckon Black's position is fine.
4...Nd7 5.Nc3 Bg7
ChessBase now calls the game "A41: 1.d4 d6: Tartakower System and Modern Defence."
5.Bg2 Ngf6 7.0-0 0-0
 |
| ChessBase finally settles on "E67: King's Indian: Fianchetto: 6...Nbd7" |
*****
*****
*****
*****
8.Qc2 Re8 9.Rd1!? e4!?
Ambitious. 9...c6 and 9...Qe7 are standard, but not 9...exd4 10.Nxd4 Nb6 11.c5!
Most popular in Mega25 is 9...c6, followed by 9...exd4 and 9...e4!? The engines like 9...Qe7!?
10.Ne1
10.Ng5 e3 11.Bxe3?! Rxe3! 12.fxe3 Ng4 gives full compensation.
10...Nb6 11.c5?
The computer [unspecified] prefers 11.Bg5 when 11...Nxc4?! 12.Nxe4 is good for White, as is 11...Bf5 12.Nxe4 Kh8 13.f3 h6 14.Bxf6 Bxf6 15.Qb3. Instead, 11...e3!? 12.fxe3 Nxc4 13.Nd5! also looks better for White, but he has to reckon with 13...Nxd5!? 14.Bxd8 Ndxe3 15.Qc3 Nxd1 16.Rxd1 Ne3 17.Rd2 Nxg2, which looks pretty messy.
11...Nbd5 12.cxd6
12.Nxe4 Nxe4 13.Bxe4 Nb4 14.Qb1 d5! is terrible for White after 15.Bf3 Bf5 or 15.Bd3 Bxd4.
12...Qxd6 13.Bg5
Not 13.Nxe4 Nxe4 14.Bxe4 Nb4 15.Qb1 Qe7 or 13.e3? Bg4.
Walter Browne (2540) - Yuri Balashov (2600), Buenos Aires 1980, went 13.Qb3 Bg4 14.Nc2!? Nxc3 15.bxc3 Bxe2 16.Re1 Qa6!?, with at least a slight edge for Black, according to the engines (0-1, 39 moves).
13...e3!?
The wedge is driven into White's position with the aid of a supporting cast.
14.f3 Nxc3 15.bxc3
15.Qxc3 Nd5 16.Qb3 h6 is even worse.
15...Nd5 16.c4 h6!? 17.cxd5 hxg5 18.Rac1 Bf5!? 19.Qb3
If 19.Qxc7 [then] 19...Qxd5.
19...Rab8
The engines prefer the immediate 19...Rad8!?
20.Qc4?!
The engines much prefer 20.Nc2.
20...Rbd8!? 21.f4!?
Exposing the king, but otherwise White might lose both d pawns for no compensation.
21...gxf4 22.Qxc7!? fxg3
22...Qa6 23.Nd3 fxg3 24.hxg3 Rc8 was even better.
23.Qxd6 Rxd6 24.hxg3
 |
| How would you assess this queenless middlegame? |
*****
*****
*****
*****
Black has the bishop-pair, the better pawn-structure and, with his next move, confirms Black's possession of the initiative, adding up to a winning advantage, according to the engines.
24...Ra6 25.Bf3 Bf8!?
As Pein points out, there is nothing wrong with grabbing the a2 pawn. He suggests Carlsen may have been worried that 25...Rxa2 could be successfully answered by 26.d6?, but that makes White's position worse, especially after 26...Rd2. However 26.g4 Bd7 27.Rc7 is more promising for White, although the engines reckon Black is winning after 27...Bb5!?
26.g4
Pein's suggestion of 26.Rc7 is preferred by the engines.
26...Be4 27.Rc7!? Rxa2 28.Kf1
If 28.Rxb7 [then] 28...Bxf3 29.Nxf3 Rxe2 and wins.
28...Rd2 29.Ra1 Rxd4 30.Rxa7 Bxd5 31.Bxd5 Rxd5 32.Kg2 Rd2 33.Kf3 Rd1 0-1
So will the King's Indian Defence undergo a revival, at high levels as well as in club play?
Time will tell, and after all it may be significant that Carlsen only committed to a KID setup after White committed to a kingside fianchetto.
You wonder how Magnus would have played it if his opponent had used the long standing normal idea of 8 h3 and 9 Re1.
ReplyDeleteWould he have kept it closed waiting for White to play d5 or played one of the variations after exd4?
I think it was assumed in the 1950s and 1960s that it was better not to allow Black to get .. d4 in. That's after some people allowed it and got into trouble.
RdC
Presumably he must have had something in mind as his opening repertoire seems to be quite deep as well as very broad. I'd love to know how much of the opening was prep.
Delete