|
|
|
26. Qxb5
|
(!) Of course. Thinking 'nothing much has really changed, just take on a2,' Black continued... |
 
|
| |
|
|
26... Rxa2
|
(!?) But maybe ought to have taken the time to look at whether 26...Qc7 or 26....Bxe4 offered more. |
2 comments
|
| |
|
|
27. Nf3
|
(!?) Again, an unanticipated move (27.Nc3), but it may be the best! |
|
| |
|
|
27... g4
|
(?) A complete miscalculation, as the GK engine observes. Originally planning on ...Rxb2, I had the uneasy feeling that I'd already let the advantage, if not the initiative withal, slip away from me (that rook move I overlooked two moves back, of course!). This move was a violent attempt to wrest back the initiative that should have got the kicking it deserved. |
|
| |
|
|
28. hxg4
|
(?!) This really ought to get a (?), as 28.Nxe5 would have placed the advantage firmly in White's hands. I had completely overlooked that possibility, assuming, as I (ought to) know better than to do, that White would retake on g4. However, this recapture leaves the game finely balanced: all on for a battle royal for the initiative! |
2 comments
|
| |
|
|
28... Rxb2
|
(!) As you were. Of course, the rook may not be taken (29.Qxb2; Nxf3ch), but White has attacking chances of his own. |
|
| |
|
|
29. Qa6
|
(!) Things are starting to look a bit weird as the pieces considerably outnumber the pawns. This became more pronounced later on... |
|
| |
|
|
29... Nxg4
|
(!?) The alternative, 29...Rb6, might actually have been better, but I wanted to push the issue. This involved some risk, though I felt sure that I could always bale to a draw if things went turnip shaped. |
1 comment
|
| |
|
|
30. Bxg4
|
(!?) As you'd expect. But as it transpired, White missed a fine opportunity, right here. The Queen sacrifices itself for very promising endgame prospects: 30.Nxd6! Rb6 31.Qxb6!! (What the...?) 31...Qxb6 32.Nxf5 Rxf5 33.Bxg4 Rf8 34.Re6 ... and White has a lot of pressure on Black's position. Well, I didn't see it at the time. As it happens, the GK engine also gives White a plus after 30.Qxc4 Nf6 31.Nxf6. So Black was running close to the extreme edge, here! |
1 comment
|
| |
|
|
30... Bxg4 31. Qxc4
|
Resources have become very slender, with few pawns remaining, but a heck of a lot of pieces. By the way, the 2B bs 2N motif was largely meaningless in this position, so Black's next is hardly surprising... |
|
| |
|
|
31... Bxf3 32. gxf3 Rf5
|
Aiming for the g- and h-files... |
1 comment
|
| |
|
|
33. Qc6
|
(?) The GK engine calls this a mistake, preferring 33.Rc1. Possibly the latter frees the immobile f1-Rook... |
|
| |
|
|
33... Rb4
|
(!) Against an opponent with a similar rating or greater than mine, I would most likely have cut straight to the draw with 33...Rg5ch. But I still thought the game offered Black chances of more. The threat is ...Rxe4, with mate then looming on the g- or h-file. I didn't much like ...Rb6, but maybe it ought to have been looked further into, with ...Rb7 as a possible follow-up. |
|
| |
|
|
34. Qxd6
|
(!?) Suddenly things weren't so simple. White's salvation, it was becoming clear, lay in the timely retreat Ne4-g3, giving up the knight for the likely Rook and Bishop vs Rook ending. This called for some very intricate manoeuvring, with no guarantee that Black would get more than the draw even then. I admit that the Ng3 move I had anticipated somewhat down the track, but the GK engine recommends it right here (34.Ng3!), giving Black only a minimal advantage. |
|
| |
|
|
34... Rg5+
|
It was this or 34...Qxd6 at once. I figured that they transposed into the same line, but stand to be corrected on this. |
|
| |
|
|
35. Kh1
|
(?) I notice the GK engine also points to this as a mistake. At the time I had a feeling this might lead to more for Black than Kh2 would have done, as the latter might have allowed the f1-rook to slip out via g1 or h1. At any rate, the real reason why this retreat turns out badly will emerge later... |
1 comment
|
| |
|
|
35... Qxd6 36. Nxd6 Rb6
|
Pinning the knight against the h6-square. How can White save it? How will it get back in time to sacrifice itself at g3? |
|
| |
|
|
37. Re6 Rxd5
|
Renewing the pin - only now it's a double pin, as the knight is pinned against the rook, AND h6. But White can survive this, at least |
|
| |
|
|
38. Re8+
|
The first unpinning... But at the time, it suddenly dawned on me that 38.Re4!? might have been be feasible. The idea was that Black runs out of pawns after, say 38.Re4 Rdxd6 39.Rxf4 Rh6ch 40.Kg2 Rg6ch 41.Rg4 If a pair of rooks come off, Black has just R and B vs R and 2P, and somehow, it was hard to imagine that the two pawns would transform a book draw (R and B vs R and no pawns!) into a loss. To realise any advantage after 38.Re4, Black would have had to proceed much more subtly. |
|
| |
|
|
38... Kh7
|
Forced. |
|
| |
|
(2332)