| Sep-20-09 | | WhiteRook48: Marshall manages a draw despite being at a disadvantage |
|
| Jan-18-12 | | King Death: This is the kind of endgame I'd expect Capablanca to convert and it's surprising that he didn't bring it in for the whole point. |
|
| Aug-12-15 | | Howard: Yes, one would think the passed h-pawn would probably have brought the win home. |
|
| May-27-16 | | edubueno: Capa obtained an advantage from the early opening. But Marshall, who played very well in this Tournemant, defended the position and achieved a draw. |
|
| Dec-08-20 | | zydeco: Alekhine says the winning approach was 40...Kg7 and if 41.Nd7 Ng8 followed by ...Rd4. |
|
Dec-08-20
 | | beatgiant: <zydeco>
40...Kg7 41. Nd7 Ng8 42. Re6 Rd4 43. Ne5, and White looks much more active, so a Black win is far from clear (to me at least). |
|
| Dec-08-20 | | zydeco: <beatgiant> After 42.Re6 Kf7 43.R moves Rd4 44.Ne5+ Kf6 and white is going backwards. Alekhine's larger point is that black should keep the king close to the kingside pawns. With 40...Kf8?! Capablanca gives white his opportunity for counterplay. |
|
Dec-08-20
| | beatgiant: <zydeco>
On 40...Kg7 41. Nd7 Ng8 42. Re6 Kf7, White has <43. Rc6> threatening both the kingside (Ne5+) and the queenside (Rc7) and seems to me White will win back the pawn, at least. |
|
| Dec-09-20 | | RandomVisitor: Stockfish thinks that by move 39 white is within a half pawn in a difficult ending that is completely drawn after 39...Rxh4. Black might try 31...g5! breaking open the game and allowing moves like Ng6 and Nf4 or Ne5: 
click for larger viewStockfish_20120516_x64_modern:
<57/83 9:23:00 -3.74 32.Kf1 Ng6> 33.Re8+ Kg7 34.Nc6 Kf6 35.a3 Rc7 36.Nb8 h5 37.Na6 Rc6 38.Ra8 Nf4 39.Rxa7 Rd6 40.Ke1 Nd3+ 41.Kd2 Nxf2+ 42.Kc2 Nd3 56/86 8:51:02 -4.13 32.f4 g4 33.f3 Ng6 34.Re8+ Kg7 35.Nc6 Rd6 36.Ne5 Nxf4 37.fxg4 fxg4 38.Nxg4 Rg6 39.Re7+ Kg8 40.Kf2 Rxg4 41.Rxa7 Nd3+ 42.Ke3 Nc1 56/81 9:23:00 -4.47 32.Re1 Ng6 33.Nc6 Kg7 34.Kf1 Kf6 35.Re8 h5 36.Ke1 Rc7 37.Nb8 Ne5 38.Ke2 g4 39.fxg4 hxg4 40.Na6 Rh7 41.Rf8+ Kg5 42.Rg8+ Kf4 |
|
| Dec-09-20 | | zydeco: <beatgiant> Looks like you're right. beatgiant 1 Alekhine 0 |
|
| Dec-09-20 | | RandomVisitor: A final look after the suggested winning move 31...g5: nothing seems to hold for white
click for larger viewStockfish_20120516_x64_modern:
<59/81 22:26:18 -4.36 32.Re1 Ng6> 33.Nc6 Kg7 34.Kf1 Kf6 35.Re8 h5 36.b4 cxb4 37.Nxb4 Nf4 38.Ke1 Rc7 39.Rf8+ Ke5 40.Re8+ Kd4 41.Nd5 Rxc4 42.Ne3 Ra4 59/84 22:26:18 -4.40 32.Kf1 Ng6 33.Re8+ Kg7 34.Nc6 Kf6 35.a3 Rc7 36.Nb8 h5 37.Na6 Rc6 38.Ra8 Nf4 39.Rxa7 Rd6 40.Ke1 Nd3+ 41.Ke2 Nc1+ 42.Ke1 Nxb3 59/90 22:26:18 -4.94 32.f4 g4 33.f3 Ng6 34.Ne6 gxf3 35.Re3 h6 36.Kf2 Ne7 37.h4 Nc6 38.Rxf3 Kf7 39.Rg3 Rd2+ 40.Ke1 Rxa2 41.Nc7 Rh2 42.Re3 Rxh4 58/89 16:06:58 -4.79 32.h3 Ng6 33.Ne6 Kf7 34.Nxg5+ Kf6 35.h4 Nf4 36.Re8 h6 37.Rf8+ Kg7 38.Rxf5 hxg5 39.Kh2 Rd3 40.Rxg5+ Kf6 41.Kg3 Ne2+ 42.Kg2 Rd2 |
|
Dec-09-20
| | beatgiant: <RandomVisitor>
A typical counter against ...g5 for White would be to play <h4> to break up the pawn structure. This will lead to -2.00 type engine evals (Black is 2 pawns up), but the resulting position may be technically difficult to win (extra doubled rim pawns). I'll post analysis of that idea if time permits. |
|
Dec-10-20
| | beatgiant: For example, 31...g5 32. h4 gxh4 33. a4 Ng6
34. Nc6 Kg7 35. a5 Nf4 36. Re1
click for larger viewYour engine eval will probably be around -2.00, but I claim it is not technically easy for Black to win. The extra pawns are doubled rim pawns, and White has good activity. |
|
| Dec-10-20 | | RandomVisitor: <beatgiant>Let's say your line is best play. Stockfish - given enough time and loaded with tablebases which can detect drawn endings - can see if an ending is going nowhere or if the side going for the draw has to give ground. Stockfish after a few dozen minutes sees that the black player can continuously make multiple threats with an active king, and likely white will have to give ground:
click for larger viewStockfish_20120516_x64_modern:
NNUE evaluation using nn-62ef826d1a6d.nnue enabled <45/72 34:56 -7.61 36...bxa5 37.Rf1 Kf6> 38.Nxa5 Rg7+ 39.Kh1 h3 40.Rg1 Rd7 41.Rf1 Rd3 42.Nb7 Ne6 43.Rb1 Rd7 44.Nxc5 Nxc5 45.Kh2 a5 46.b4 Rb7 47.b5 Ke5 48.Rd1 Rd7 49.Rc1 Kd4 50.Kxh3 a4 51.f4 Rg7 52.b6 Rb7 53.Kh4 Rxb6 54.Kg5 Ra6 55.Ra1 Ra5 56.Ra3 Ra7 57.Kxf5 Ra5 58.Kg4 h5+ 59.Kxh5 Ne6+ 60.Kg6 Nxf4+ 61.Kg7 |
|
Dec-10-20
| | beatgiant: <RandomVisitor>
From the diagram above, in the <36...bxa5> line, I think White would try to trade into a rook ending like 36...bxa5 <37. Ne7> Kf6 38. Nd5+. The result looks to me like some real technical difficulties for Black, with all those scattered pawns. Many pawn exchanges will follow, and in the end Black's extra pawns are rook's pawns. But before I do lots of work trying to demonstrate that, I'm curious to see what your computer analysis will say. By the way, 36...bxa5 was a major surprise for me, allowing the complete breakup of Black's pawn structure. |
|
| Dec-11-20 | | RandomVisitor: <beatgiant>I have setup your line and will post when I return from work. |
|
| Dec-11-20 | | RandomVisitor: <beatgiant><From the diagram above, in the <36...bxa5> line, I think White would try to trade into a rook ending like 36...bxa5 <37. Ne7> Kf6 38. Nd5+. The result looks to me like some real technical difficulties for Black> I have followed up wth 38...Nxd5 39.cxd5 Rxd5
click for larger viewStockfish_20120516_x64_modern:
NNUE evaluation using nn-62ef826d1a6d.nnue enabled 54/60 5:17:26 -11.54 40.Ra1 Ke5 41.Kg2 h6 42.Ra4 Rd3 43.Rxa5 Kd4 44.Rb5 a5 45.Rb8 Rc3 46.Rb6 h3+ 47.Kg3 h2 48.Kxh2 Rxf3 49.Kg2 Rc3 50.Rd6+ Ke4 51.Re6+ Kd3 52.Re5 f4 53.Rd5+ Kc2 54.Rd8 Rxb3 55.Rc8 Rb5 56.Kf3 a4 57.Ke4 Kb3 58.Rd8 c4 59.Rd1 c3 60.Rb1+ Kc4 61.Rc1 a3 62.f3 Rb4 63.Ra1 Kb3+ 64.Kd5 a2 65.Kc5 Rb7 66.Kc6 Rb8 |
|
Dec-11-20
| | beatgiant: <RandomVisitor>
There are a number of mysterious moves in the computer line, but the overall conclusion seems right, Black is probably winning. The winning plan is to play for a passed c-pawn. I'm not sure if that counts as technically difficult, although I do know I would not resign such a position against a fellow member of my local chess club. |
|
| Dec-11-20 | | RandomVisitor: <beatgiant>< I do know I would not resign such a position against a fellow member of my local chess club.>Correct, the position needs to be played out. The 31...g5 32.h4 gxh4 line is a curious ending where black can possess a number of individually weak pawns yet build lasting threats. |
|
| Dec-17-20 | | RandomVisitor: After 31...g5 32.h4 gxh4 white cannot take advantage of black's fractured pawns, black meanwhite can maximize piece activity and dominate the board:
click for larger viewStockfish_20120516_x64_modern:
<59/80 36:29:47 -6.62 33.Re1> Ng6 34.Nc6 Kg7 35.b4 Kf6 36.a4 cxb4 37.Nxb4 Ne5 38.Rc1 Nxf3+ 39.Kg2 Nd2 40.Rd1 Ke5 41.Kh3 Rd6 42.Re1+ Kd4 43.Nd5 Kxc4 59/100 36:29:47 -7.31 33.Kh2 Ng6 34.Nc6 Kf7 35.b4 Kf6 36.Rb2 Kg5 37.bxc5 bxc5 38.a4 Kf4 39.Rb5 Rc7 40.Nd8 Kxf3 41.Ne6 Rc6 42.Nxc5 Kxf2 43.Nd3+ Ke3 58/76 36:29:47 -7.23 33.a3 Ng6 34.Nc6 Kg7 35.b4 Kf6 36.Kh2 Kg5 37.Rb2 Rc7 38.Nd8 Kf6 39.bxc5 Rxc5 40.a4 Rxc4 41.a5 bxa5 42.Rb7 Rc2 43.Rf7+ Ke5 |
|
Kibitzing
icon.