This is the Ragozin Variation of the Queen’s Gambit Declined, an ambitious choice selected by Fischer on occasion. It has made a comeback in recent years.

5.Qb3

Black has to be prepared to meet six possible, strong, White fifth moves. In order of popularity these are 5.Bg5, 5.cxd5, 5.e3, 5.Qa4+, 5.Qb3, and 5.g3.

5…a5

Both sides are deviating from main lines quickly. Here, 5…c5 is played two thirds of the time because it protects the Bishop and applies pressure to a central square (d4), but 5…Nc6 and 5…a5 have its specialist devotees.

6.a3

White has preferred 6.Bg5 in the past, but Etiene Bacrot won a nice game last year by putting the question to the Bishop. Has Shankland seen that game and included that line in his preparation? Moving 6.a3 now makes sense since the move inevitably arises later in the 6.Bg5 line. Perhaps g5 is not going to prove the optimal square for White’s Bishop.

6…Be7 7.Bf4

Sure enough. As Bacrot played before him, Shankland opts for the f4 square for his Bishop.

7…O-O

This position has to be nerve-racking for Black. He has only two pieces developed compared to White’s four. Nevertheless, White is not currently putting direct pressure on the position. This leaves Black a range of options, including 7…c6 (as Bacrot’s much lower rated opponent played and lost with), 7…a4 (kicking the Queen), 7…Nh5 (querying the Bishop), or even 7…Nc6 (to try to catch up in development). But Black plays it cool by simply castling, resulting in a position that has not been reached before.

8.e3 b6 9.Rc1

This move came as a surprise to me. Any time in Queen’s Gambit Declineds when Black plays …b6 to shore up the long light diagonal, cxd5 is White’s reflex reaction to cement a Black pawn on d5 and limit the scope of Black’s light-squared Bishop. That reaction still looks good to me even though I now see that White had another plan in mind instead.

9…Bb7 10.cxd5 Nxd5

The principled response. Since White is allowing Black to keep the diagonal open, far be it from Black to shut it down by placing a pawn on d5.

11.Nxd5 Bxd5 12.Bc4

White’s alternative was 12.Qc2, but 12…Bd6 seems adequate for equality. With the text, White tries to keep an edge.

12…Bxf3

Black gives up the “minor exchange”. In retrospect, I imagine he wishes he had tried 12…Bxc4 13.Qxc4 Bd6 instead.

13.gxf3 Bd6 14.Bg3

White has an advantage here. Besides having slightly more space, his position is more flexible. The exchange …Bxg3 will never be to Black’ advantage as long as the Rook can bear down the h-file. However, there may come a time when Bxd6 works to White’s advantage.

14…Nd7 15.Bb5

The threat is Bc6, forever squelching any possible Black counterplay. Without …c5 as a possibility, Black has no possible pawn breaks.

15…Nf6 16.Ke2 Nh5 17.Rc6 Nxg3+ 18.hxg3 e5

Black has worked hard (by removing the Bishop on g3) to manufacture this central pawn break.

19.dxe5 Bxe5 20.Bd3!

I imagine White might have entered this line planning to play 20.f4 now when his light squared dominance would eventually score the point, but there is a neat, hard-to-see tactic afoot. Will Black recognize it and counter it? Does it cost White anything to try it and see? The answer to the second is “no”; therefore…

20…g6?

Black missed the tactic, or misevaluated the consequences. His only possible hope to salvage the position was with 20…h6 instead, but things are then positionally grim after White presses on with 21.f4.

21.Bxg6 hxg6 22.Rxg6+ Bg7 23.Rh4!

This is the move whose strength Black probably failed to see.

23…a4 24.Qa2

Changes nothing.

24…Qd7 25.Rhg4

The point. Black recovers the piece two pawns ahead. Game over.

25…Qb5+

25…Rad8 is only another way to lose. 26.Rxg7+ Kh8 27.Qb1! +-

26.Ke1 Rad8 27.Rxg7+ Kh8 28.R7g5 Rd1+ 29.Kxd1 Qf1+ 30.Kd2 Rd8+ 31.Rd5 Qxf2+ 32.Kd3 Qf1+ 33.Ke4 f5+

After 34.Kf4 fxg4 (or 34…Rxd5 35.Rh4+ followed by Qxd5) 35.Rxd8+ White mates within a few more moves.