|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
| Feb-03-16 | | dfcx: Black has a knight vs 3 pawns. I do not see any mates so will try to win material instead. 31...Ng5
A. 32.Rf1 Ne4 33.Q anywhere Ng3+ forks the rook.
B. 32.Rg3 Qxg3
C. 32.Rf2 Ne4 forks rook and queen
D. 32. fxg5 Rxf3
With a full rook ahead black should win easily. |
|
Feb-03-16
 | | Phony Benoni: Well, that's nice. I was so intent on getting in ...Qe4 that I didn't notice the Rf3 was simply hanging! But there are complications: after <32...Ng5 33.fxg5 Rxf3 34.dxc6>: 
click for larger viewBlack is winning, but has to be careful. |
|
| Feb-03-16 | | saturn2: After 31...Ng5 the knight cannot be taken and the the white rook can only go to f2 or f1. in the first case there is another fork Ne4 in the second case there is also 32...Ne4. Now the white queen has to move but black forks rook and king with 33...Ng3+ |
|
| Feb-03-16 | | agb2002: Black has a knight for three pawns.
White threatens 31.d6 and 31.dxc6.
The knight does not look well placed. Therefore, 31... Ng5: A) 32.fxg5 Rxf3
A.1) 33.Rxb7 Rf1+ 34.Kg2 Qh3#.
A.2) 33.dxc6 Qe4
A.2.a) 34.Rg1 Rf2+ 35.Rg2 Qxg2#.
A.2.b) 34.Qd6+ Kg8 wins (35.Kg1 Qxb1+).
A.3) 33.d6 Qe4 as in A.2.
B) 32.dxc6 Nxf3 33.Qd6+ Kg7 - + [R+N vs 4P].
C) 32.Rff1 Ne4
C.1) 33.Qb2 Ng3+ 34.Kg1 Nxf1 wins decisive material.
C.2) 33.Qe2 Ng3+ wins the queen.
C.3) 33.Qg2 Ng3+ 34.Kg1 (34.Qxg3 Qxg3 wins) 34... Nxf1 wins decisive material. D) 32.Rf2 Ne4 wins decisive material.
E) 32.Rg3 Ne4 also wins decisive material. |
|
| Feb-03-16 | | Nick46: I got 31... Ng5 but didn't follow up correctly. |
|
| Feb-03-16 | | Once: Now that is interesting.
In human mode, I couldn't see anything wrong with 32...Rxf3 instead of 32...Qe4. But here's the position that <phony> gave us after 33. dxc6
click for larger viewBlack's position looks to be overwhelming with a clear rook up, but white has chances of queening a pawn or giving a series of unanswerable perpetual checks. From this position, Fritz finds only one move that wins for black. 33...Qe4 sets up a deadly discovered check and mate in 9. Apparently. Other moves for black lose or draw.
So, yes, both 32...Qe4 and 32...Rxf3 win. But to make 32...Rxf3 stick we also have to spot 33...Qe4. Over the board, and with the tock clicking, 32...Qe4 is probably the better choice, although Fritzie doesn't rate it as highly as 32...Rxf3. It is the more pragmatic move and gives black an easier winning plan. |
|
| Feb-03-16 | | morfishine: <31...Ng5> Seemed strongest all things considered |
|
| Feb-03-16 | | saturn2: After 13 Nc5 e4 black got a piece for a pawn. |
|
| Feb-03-16 | | Cheapo by the Dozen: I calculated far enough to be confident that Black would at least achieve a position in which: 1. He had a rook for 2-3 pawns, albeit somewhat scary ones (passed, connected). 2. Have White thoroughly tied up, by pointing up to 3 heavy pieces at h2. I wasn't sure whether he could actually win a rook outright in all lines rather than just the exchange, but I didn't need to know that to decide on a plan. :) |
|
| Feb-03-16 | | cunctatorg: Real tactics here! |
|
| Feb-03-16 | | whiteshark: <Maybe <31...Ng5>? If 32.fxg5 Qe4 hits both rooks, and the way is also opened for the Rf7 to intensify the attack on f3.> This <double attack> has a more <devastating effect> on the opponent's mind than the simple ♖ook grabbing. |
|
| Feb-03-16 | | Pawn Slayer: The line I like best is 31... Ng5 32 fxg5 Rxf3 33 dxc6 Qe4! and all black's threats are neutralised. |
|
| Feb-03-16 | | cocker: Fritz thinks that the obvious 32 ... Rxf3 is much stronger. |
|
| Feb-03-16 | | patzer2: Tough for a Wednesday. After rejecting three other possibilities, I found 31...Ng5!  .Thought maybe I missed something when I looked at the game with Fritz and saw Black played 31...Ng5! 32. fxg5 Qe4! (-10.14 @ 21 depth, Deep Fritz 15). However, I was pleased to see my simpler choice 31...Ng5! 32. fxg5 Rxf3 (#11 @ 21 depth, Deep Fritz 15) is just as good if not better. For a White improvement, instead of 12. c5 =, 12. Ba3! offers a slight advantage when play might continue 12...Bf8 (12... c5 13. b4 f5 14. Nc3 cxb4 15. Bxb4 Nc6 16. Ba3  ) 13. c5 dxc5 14. Bxc5 f5 15. Ned2 Bxc5 16. Rxc5 Nd7 17. Rc2 Nb6 18. Qa1 Qf6 19. Rc5 Nd7 20. Rcc1 . |
|
| Feb-03-16 | | Once: <cocker: Fritz thinks that the obvious 32 ... Rxf3 is much stronger.> Well, yes and no.
32...Rxf3 does seem to be the stronger move, but only if you follow it up precisely. This is the position after 32...Rxf3: 
click for larger viewBlack is a whole rook up and threatens mate in 2 with something like Rf2 (to scare away the white queen and then Qxh2#. From this position, Fritzie looks at grovelling white moves like 33. Qg2 or 33. Kg1. And we could all probably win from there. But <phony> has pointed out that white has 33. dxc6 when things aren't quite so clear:
click for larger viewTo Fritz's "mind", this position is a simple mate in 9 for Black, which is one of the reasons why he evaluated 32...Rxf3 as stronger than 32...Qe4. But imagine sitting over the board with this position in front of you. You need at least two moves to give mate, but white has immediate threats like Qd6+ and Qd8+. You aren't an infallible calculating machine like Fritz. You'd hate to throw the win away as white squeezes out a miracle draw by perpetual. And it turns out that the "obvious" 33...Rf2 would be a huge mistake as it allows white to draw with 34. Qd6+ From this position, the only move to secure the win is 33...Qe4. So yes in computer-land 32...Rxf3 is much stronger. But in human play 32...Qe4 is much easier to play and we are less likely to go wrong. We need to be careful how we interpret what the computers are telling us. A line that the computer decides is best might not be the best over the board. |
|
| Feb-03-16 | | patzer2: <Once> After having recently gone over 70 moderately easy discovered check combinations for tactics classes with my grandsons, I personally find the continuation 32...Rxf3 33. dxc6 Qe4! to be an irresistable, blitz-speed first choice. However, with my precocious seven year old grandson occasionally winning a blitz game against me, I find blitz-speed calculation isn't always helpful. |
|
| Feb-03-16 | | patzer2: In the opening, instead of 6. Nc3, I slightly prefer 6. d4 as in J L Hammer vs Fressinet, 2015. |
|
Feb-03-16
 | | Sally Simpson: Once's wise words.
" A line that the computer decides is best might not be the best over the board." Should be displayed in bright flashing letters every time someone fires up their chess playing engine. |
|
| Feb-03-16 | | CHESSTTCAMPS: Black has a knight for three pawns, with a strong king-side attack. White threatens 32.dxc6, to be followed by queen checks, or cxb7 with possibly dangerous ascending pawns. However, black can trap a rook, keeping checkmate threats in the back pocket. 31... Ng5! leaves no place to run:
A.32.Rf2 Ne4 33.Q moves Nxf2+ wins
B.32.Rg3 Ne4 33.Q moves Nxg3+ wins
C.32.Rf1 Ne4 33.Qb2/c2 Ng3+ 34.Kg1 Rg7! 35.Kf2 (hxg3 Qxg3+) Nxf1+ 36.Kxf1 Qxh2 is winning.
C.1 32.Qe2 Ng3+ wins queen
C.2 32.Qg2 Ng3+ 33.Kg1 Nxf1 34.Kxf1 (Rxf1 Rg7 wins queen) Qxh2 is winning. D. 32.fxg5 Qe4! 33.dxc6 Qxf3+ 34.Kg1 Qxc6 leaves black down 2 pawns for a rook with no counter-play. E. 32.dxc6 Nxf3 33.Qd6+ (otherwise 33... Qxh2+/#) Kg7 34.Kg2 Qg4+ 35.Kf1 Qg1+ 36.Ke2 Rxh2+ 37.Kd3 (Kf3 Qg2#) Qxb1 wins. Time for review... |
|
| Feb-03-16 | | CHESSTTCAMPS: Obviously D. should read *white* down 2 pawns for a rook. |
|
Feb-03-16
 | | LoveThatJoker: <31...Ng5!>
For example,
A. 32. fxg5 Qe4! 33. Rf1 (say) Rxf3
B. 32. Rf2 Ne4!
C. 32. Rf1 Ne4 and, at the very least (i.e., without calculating much), 33... Ng3+ D. 32. Rg3 Qxg3
LTJ |
|
| Feb-03-16 | | cocker: <Once> thank you, you are quite right. I'm afraid I assumed that White couldn't play 32 fxg5 because of 32 ... Rxf3. Not sure I would have found 32 ... Qe4. |
|
| Feb-03-16 | | kevin86: Black sacs the knight to set up a deadly pin of the rook. |
|
| Feb-03-16 | | Pinkerton: Medium? Might as well be crazy to me. No clue |
|
Feb-03-16
 | | gawain: Way hard for a Wednesday! I saw 31...Ng5 32 fxg5 Qe4! but rejected it because I thought White would simply refuse to capture the knight. Eventually I resorted to setting up a board--something I should do more often. After staring at it (on and off) for hours I finally saw that 31...Ng5 works against all 32 R moves. because of the havoc that Black will create with further knight moves capitalizing on attacks against the white rook (and queen) and the pinned h-pawn. Whew! |
|
 |
|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
Kibitzing
sac: 31...Ng5