|KEG: Post II
Following Rosen's 19...Re8, Sterling wanted to Castle but faced the seemingly daunting threat of 20...BxN leaving the White Bishop hanging on e2. So, Sterling played a move with which he "only" lost a pawn:
What Sterling overlooked is that he could have played 20. 0-0 with impunity since after 20...BxN he had the amazing retort 21. Bc4!
Both of the players--as well as Rosenthal in commenting on the game in the Tournament Book--missed this possibility. As a result, Sterling lost a pawn--though with some compensation.
Of course not 21. NxN+ BxN and the pin on the e-file becomes fatal. So Sterling submits to the loss of a pawn.
An alternative to the text, and maybe a better one here for Sterling, was 21. Bc4, though this might well have transposed into the line actually played.
The position was now:
click for larger view
Here Rosen chose to take the undefended pawn immediately with:
This, however, allowed Sterling to get some compensation for the lost pawn. Better was 22...Be6, eliminating the threat of Bc4. The weak, isolated e-pawn will not run away.
Now Sterling gets attacking chances which he only partially exploits.
23. Bc4 Bg6
Right Rook, wrong file--as Sterling apparently recognized two moves later. His compensation for the lost pawn is now tenuous.
25. Bb6 Qe7
Implicitly conceding that his 24th move was a mistake.
Unduly defensive and giving Sterling more fuel for his attack. Much better was 26...h6 or 26...Rac8 or even 26...Rab8. Rosen's move was especially bad considering his follow-up on the very next move.
Sterling's attacking instincts are correct here, but he first should have gotten his King to safety with 27. Kh1. The attack on Black's e5 Bishop is an illusion--that apparently fooled both players.
Missing his chance. There was no need to defend the Bishop. Instead, Rosen should have played 27...Qb4! with chances of exploiting his pawn plus.
What was right on the last move is wrong here. Sterling should have played 28. Qb3 with good counterplay for the pawn.
Again missing a chance. This Queen move accomplished nothing. By contrast, 28...Qd7 would have made Sterling scramble.
Setting a trap!
Rosen sees it! Had he played the superficially tempting 29...Qg3 Sterling would--as noted by Rosenthal in the Tournament Book--have had excellent chances with 30. Bxf7+ BxB 31. QxB+ Kh8 32. Bg1!
30. Bf2 Qf4
31, Bg1 Qc7
32. Bd5 Rab8
As pointed out by Rosenthal, 32...Reb8 would have led to trouble after 33. Bh2! But to call White's position here winning--as does Rosenthal--is clearly an overstatement.
After Rosen's actual move, the position was:
click for larger view
Sterling here ended the game with a neat little drawing Queen sacrifice:
Rosen, of course, could not play 33...RxQ because of 34. RxR mate. And after 33...QxQ 34. BxQ a5 35. RxR+ RxR the game should be drawn. (But not on Rosenthal's awful 35. Ba6 in this line which would have allowed Rosen excellent winning chances after 35...RxR 36. RxR h6.
In any case, the players here--not unreasonably--agreed to a draw (which meant the game would be replayed with colors reversed two days later--an encounter in which Rosen prevailed).