1.e4 d5 I 'knew' my opponent was going to play this; a couple of years ago I played the same opponent with the same colours in the same tournament and lost the game, and I was pretty sure my opponent would play the same opening as last time. 2.exd5 Nf6 3.d4 Bg4 4.Be2 Bxe2 5.Qxe2 Qxd5 6.Nf3 the best move no doubt; last time we played I played 6.Qf3?!, not seeing that it was possible to play the simple 6.Nf3 (defending the d4 and g2 pawns) and I ended up lagging in development with a bad endgame, which I lost as I mentioned above 6...e6 7.0–0 Not necessarily the best option, as in positions where black castles queenside in the scandinavian are often easier for black to attack, whereas it can be comparatively tricky for white to attack the black queenside fast enough [F likes this first 7.Nc3 Bb4 8.0–0] 7...Nc6 8.Nc3 Qf5 pretty good so far from both sides, and F evaluates it as equal 9.Nb5 [F prefers 9.a3 0–0–0 10.Be3] 9...0–0–0 10.Ne5 A good move I think. 10...Nxe5 [10...Nxd4 was also playable, but not quite as good according to F 11.Nxd4 Rxd4 12.Nxf7 Rg8 13.Ng5 e.g. 13...Rd6 14.c3 Be7 15.Re1 F evaluates about equal, but I prefer the position for white slightly as the isolated black pawn may be difficult to prove a strength instead of a weakness for black] 11.Qxe5 Threatening mate on c7 forcing the trade of queens, I thought at first I was winning a pawn after ..Qxe5 dxe5 N moves Nxa7+ but obvsiously the knight can move to g4 keeping material equal [Fritz like at first: 11.Nxa7+ but changes it's mind, prefering 11.dxe5 11...Kb8 12.dxe5 Ng4 13.h3 (13.Nb5 Bc5 looks strong for black 14.Nc3 h5 15.h3 Bd4) 13...h5 14.Nb5 Bc5 doesn't look very pleasant for white; 11.dxe5 Fritz's first move 11...Ne4 12.Nxa7+ Kb8 13.Be3] 11...Qxe5 [11...Qxc2 F thinks it would have been better to do this 12.Bf4 (12.Nxa7+? Kb8 13.Qa5 (13.Nb5?? Rd5) 13...Rxd4 and the position is very poor for white) 12...Rd5 13.Qxc7+ Qxc7 14.Nxc7 Rxd4 15.Rac1 (15.Be3 Rd7 16.Nb5 b6 white has basically no compensation for the pawn) 15...Kd8 F –0.4; white has some work to do to justify being a pawn down; he does have better development for the time being] 12.dxe5 [the cheeky move 12.Nxa7+ was also good / maybe better 12...Kb8 13.dxe5 Kxa7 (13...Ng4 14.Nb5 Nxe5 15.Bf4 f6 16.Bxe5 fxe5 an odd and interesting endgame position) 14.Be3+ intermezzo 14...b6 15.exf6 gxf6 another interesting endgame position; black has a bit more central control and has half open files for the rooks, whereas white has less pawn islands] 12...Ng4 [F suggests the move 12...a6 13.Na7+ (13.exf6 axb5 14.fxg7 Bxg7 15.c3 looks a little nicer for black perhaps due to the diagonal pressure on a1 b2 c3; 13.Nc3 Nd5 maybe a little better for black) 13...Kb8 14.exf6 gxf6 15.Be3 c5 16.b4 Kxa7 17.bxc5 (17.Bxc5+? Bxc5 18.bxc5 Rd5 white should lose the pawn and be a pawn down) 17...Kb8 18.Rad1 Fritz only gives black an edge. May it feels wish active play white can make up for the hideously weakened pawns] 13.Nxa7+ Kb8 14.Nb5 Nxe5 15.Bf4! a nice move to be able to play, with the X-ray effect on the c7 square and X-ray pin on the king, and forcing black into a narrow range of options 15...Bd6 Fritz seems to be torn between this and 15..f6 and can't quite make its mind up [15...f6 16.Bxe5 fxe5 this would have been an interesting endgame, black having weak isolated doubled pawns but a superior minor piece; it doesn't really look very nice for black in practice though; it could constitute a great deal of study though] 16.Nxd6 cxd6 17.Bxe5 I was happy to trade and start a double rook ending with a pawn structure imbalance. I felt my 3 to 1 majority on the queenside would offer me good winning changes, although objectively this may be a bit of an exaggerated evaluation. [F prefers to keep the bishop, e.g. 17.Rfd1 f6 18.a4 g5 19.Bd2 Nc6 the engine evaluates this more or less equal with an edge for white, and it is a very interesting endgame to analyse further, (however I am trying not to get too side tracked and to stick to the game for now, and some of these endgames can be analysed separately at a later date; it is not too relevant for the game though as they are all roughly equal in evaluation)] 17...dxe5 F actually evaluates this about equal, but perhaps it is dumb with this sort of endgame situation and cannot evaluate the long term strength of the 3 to 1 majority, so I'd probably need a very strong club player or master level player to look at the position and tell me their opinion of it 18.Rfd1 Contesting the open file and clearing the path for the king to come to the middle [F suggests 18.Rad1 but this is almost amusing; I can't see any benefit to using this rook and the king is still hemmed in] 18...Rd5 good move, logically preparing to double rooks 19.Kf1 the natural move [it never crossed my mind in the game, but just to show 19.Rxd5 would be a horrible move, letting black spread his pawn chain across and reducing the strength of the white pawn majority on the queenside 19...exd5] 19...Kc7 [19...Rhd8 is more or less the same 20.Ke2] 20.Ke2 Kc6 naturally enough bringing the king towards the centre; there are also several playable alternatives [20...Ra8 e.g. 21.a4 Ra6 22.b3; 20...Rhd8; 20...e4 21.c4 Rh5 (21...Rg5 22.g3 Rh5 23.h4 Ra5 the f2 pawn could be a potential annoyance for white) 22.h3 Rg5 (22...Ra8 23.Rd4 Ra4 24.b3) 23.g3 Ra5] 21.Rd3 I was happy to offer a trade this way and prepare to double rooks [21.c4 may have been better 21...Rd4 I think this was the move I was worried about when looking at it in the game, but F suggests taking for white 22.Rxd4 exd4 23.Kd3 e5 24.a4 Ra8 (24...Rd8 25.b4 looks better for white but is non trivial) 25.b4] 21...e4 [if 21...Rxd3 I was planning 22.Kxd3 (the engine suggests 22.cxd3 but this looks wrong to me) 22...Rd8+ 23.Ke2] 22.Rc3+ [22.Rxd5 exd5 I'm not too keen on this which F suggests, (although the evaluation is about equal for this and 22.Rc3+)] 22...Kb6 I think the king seems a little cut off by the rook on the c file, but it's very hard to tell whether that has any relevance or not in the scheme of the whole position [F suggests 22...Kd6 ] 23.Rd1 the computer still gives a dead equal evaluation; because this endgame is such a long term situation the engine would probably have to look forward a ridiculous number of plys before reaching conclusions 23...f5? I feel this is slightly dubious as the position is trick for black after Rxd5 exd5 Ke3; F for the first time gives a definite slight advantage to white. Probably the main problem is black already had a potential disadvantage with the 3 to 1 minority on the queenside and this move is committal and resolves the position somewhat, and maybe black should play more flexibly and less committally. I recall being pleased after this was played, and looking at the next couple of moves with the engine, it appears this 23rd move by back was the turning point when the game becomes a quite dire situation and very difficult defence. [23...Rhd8 was probaby better, where Fritz can come up with nothing better than: 24.Rxd5 exd5 which looks pretty equal] 24.Rxd5 exd5 25.Ke3 white is threatening Kd4 Rc5 Rxd5 (the main point being the black king is cut off from helping defend the d5 pawn and the black rook cannot defend it on its own). 25...g5 It is more or less impossible for black to avoid losing a pawn I think, so he needs counterplay and correctly plans to advance the pawns in the centre and cause threats of queening [Fritz seems to prefer Ra8. I believe I was looking at this in the game and concerned about it, but: 25...Ra8 e.g. 26.a3 Rd8 (if 26...Ra4 then 27.b4 so I realised at least I hadn't blundered (27.Rc8 g5 (if 27...d4+ then 28.Kd2 f4 this looks a bit alarming 29.Rf8 g5 30.Rf6+ Kc7 31.Rf7+ Kc6 32.f3 e3+ 33.Kd3 Ra5 a pretty crazy position where white is still likely to win a pawn but black has dangerou counterplay (similar to the game)) 28.Kd2 f4 29.Re8 Rd4+ 30.Kc1 h5 31.Re5 another insane position) 27...Ra8 28.Kd4 g6 29.Rh3 Rc8 e.g. 30.c3 Kc6 31.Rxh7 Kd6 32.Rg7 (32.Rh3?? f4 white is mated! 33.c4 Rxc4#) 32...Rc4+ 33.Ke3 Rxc3+ 34.Kf4 d4 35.Rxg6+ Kc7 36.Kxf5 d3 37.Rg7+ Kd6 38.Kxe4 d2 39.Rg6+ Ke7 40.Rg7+ Kd6 and Fritz draws this position) 27.Kd4 f4 28.g3 fxg3 29.hxg3 h5 30.f3 g5 31.fxe4 dxe4+ 32.Kxe4 h4 33.gxh4 gxh4 34.Rd3 Re8+ 35.Kf3 white should probably win but there are probably improvements for both sides; this kind of endgame analysis is extremely difficult because every time you try a candidate move there is a massive web of rapidly branching variations 35...Rg8 36.Rd4 Rg3+ 37.Kf2 Rh3 38.Rc4 Ka6 39.Rc8 b5 40.Ra8+ Kb7 41.Rg8 Kc7 42.Rg7+ white should have winning chances but it seems like it could take forever analysing such a rook ending] 26.Kd4 Rd8?! a mistake according to Fritz which gives white over a pawn advantage now. Black cannot prevent the loss of a pawn now and this appears to be the second main turning point where the game first went from difficult to defend (..23.f5) to clearly losing [26...Re8 the best defence according to F 27.Kxd5 Rd8+ 28.Ke5 Rd2 29.Kxf5 Rxf2+ 30.Kxg5 (30.Kxe4 Rxg2 31.h3) 30...Rxg2+ 31.Kf4 Rxh2 32.Kxe4 It still looks unpleasant for black but would have been potentially very difficult for white to win against good defence] 27.Rh3! Because 27.Rc5 followed by Rxd5 was now futile I looked for other moves, and saw black's position had a second weakness [I was aware that 27.Rc5 was no longer on after 27...f4 28.h3 for example.(28.Rxd5?? loses 28...Rxd5+ 29.Kxd5 e3 and black queens) ] 27...f4 A good move under the circumstances forcing white to calculate very carefully 28.Rxh7 [I considered checking before taking the h pawn 28.Rh6+ but didn't see any good reason to 28...Ka5 for example; remember the phrase from Fischer "Patzer sees a check patzer plays a check"; this move isn't necessarily bad but pointless unless I have a clear reason for playing it] 28...Re8 another accurate move from black 29.Rh3?! this appears to be a mistake judging by the engine [29.Rf7 was the right move according to F, presumably as it prepares a potential Rf1 Re1 blocking the e pawn from promoting once the position opens up] 29...Kc6?! letting white off the hook, and it is back to about +1 to white [29...g4 this would have been unpleasant. It is clear looking through some example lines that this would have cause white real problems and made it potentially very hard to win 30.Rh6+ (30.Rc3 f3 31.gxf3 (31.g3 Rh8 an example line 32.Kxd5 Rxh2 33.Rb3+ Kc7 34.Rc3+ Kb6 35.Kxe4 Rxf2 36.Kf4 Rd2 37.Kxg4 f2 38.Rf3 Rxc2 39.Kh3 Rxb2 40.a3 Kc5 41.Kg2 Ra2 42.Rb3 Kd5 43.Rxb7 Rxa3 44.Kxf2 Ke6 looks like it should be drawn) 31...exf3 32.Re3 Rh8 33.Rb3+ Kc7 34.Rb4 Rxh2 35.Ke3 Rg2 36.Rd4 Kc6 and the computer says this is about equal) 30...Kb5 31.Rf6 f3 32.gxf3 exf3 e.g.: 33.Rg6 Re2 34.Rxg4 Rxf2 35.a4+ Ka5 36.Kc3 Rxh2 37.Rd4 b5 38.Rxd5 Kxa4 39.b3+ Ka5 40.Rf5 f2 41.Rf7 b4+ 42.Kd3 Kb5 43.Ke3 f1N+ 44.Rxf1 Rxc2 45.Rf5+ Rc5 46.Rxc5+ Kxc5 47.Kd3 Kb5 48.Kc2 Ka5 49.Kb1 Ka6 draw 50.Ka1 Ka5 51.Ka2 Ka6 52.Ka1 Ka5 53.Ka2] 30.Rc3+ this appears to be a mistake, as Fritz reduces white's advantage to less than half a pawn after thinking for a while [F also suggests 30.a4 ; 30.Rh6+; F likes 30.g4 b5 (30...fxg3 31.hxg3 Rf8 32.Rh6+ Kb5 33.Rh2 white has a clear advantage by the look of it) 31.b3 Re6 32.a3 Kd6 33.c3 now F puts the evaluation about equal; I think on all of these moves the engine needs a long time to think as the evaluation can change from a win for one side to a draw or vice versa all of a sudden] 30...Kd6 F evaluates this more or less equal, slightly better for white 31.h3 I felt obliged to play this as I felt after g4 the h2 pawn would be too weak but F doesn't seem concerned with this [31.a4 g4 32.Rb3 f3 33.gxf3 exf3 34.Rxb7 Re2 35.Rb6+ Kc7 36.Rb4 Rxc2 37.Ke3 d4+ 38.Rxd4 Re2+ 39.Kf4 Rxf2 40.Rc4+ Kd6 41.Kxg4 Rxh2 42.Kxf3 Rxb2 technically a draw according to F and certainly looks like it] 31...e3 F agrees with this and gives equal evaluation 32.fxe3 Re4+ I knew this was coming but felt it was harmless. I believe around this point in the game I was thinking to myself that the main thing was not to lose and to at least draw the game, and a win would just be a bonus. 33.Kd3 Rxe3+?! I was surprised my opponent played this and though I was not winning fairly clearly [I was expecting 33...fxe3 which would have made it difficult for white to win 34.Ke2 then e.g. 34...d4 35.Rc8 Rf4 36.Kd3 Re4 37.Ke2] 34.Kd2 d4? this is technically a blunder, although I didn't think at the time I could trade rooks and judged it to be a likely draw, and maybe even a potential loss in some cases. It is interesting how often a pawn move turns out to be a critical mistake in endgames. Obviously you can't be afraid of making pawn moves but they seem the most hazardous kind of move, no doubt because of their committal nature. Having looked forward a few moves it appears that black's error was actually incorrectly allowing a lost pawn ending. It is well known that you have to be extremely careful about allowing or trading into a pawn ending because of how cut and dry they are. [34...Re8 was the correct move according to F 35.Rb3] 35.Rb3 White is still doing very well after this (F gives over a pawn to white) but trading rooks was apparently the best. I think that as I was in doubt I was sensible not to trade into the pawn ending, but it must have been possible to work out it was winning [35.Rxe3! e.g. 35...dxe3+ 36.Kd3 Ke5 37.c4 it is soon becoming clear the advantage white had from the beginning - 3 vs 1 pawn majority on the queenside - is now going to win the game 37...Kd6 38.g3 now it is over: if black takes then the white king will mop up the e and g pawns, and if black does something else allowing gxf4 gxf4, the passed pawns that will be on both sides of the board will be impossible for the black king to stop 38...Ke5 39.a4] 35...Kc6?! [35...Re7 apparently black has to retreat the rook] 36.Ra3?! F now only gives slightly more than half a pawn to white [36.Rxe3 as before, this is winning; 36.Rd3 would probably have been better if I was always to be intent not to trade rooks in this kind of position] 36...Kc5?! Again allowing the winning RxR [at first F suggests 36...Rxa3 but looking deeper I think it seems unhappy about it 37.bxa3 e.g. 37...b5 38.Ke2 Kd5 39.Kd1 Kc6 40.Ke1 Kd6 41.Kf2 Kd5 42.Kf3 Kc5 43.Ke4 Kc4 44.Kf5 Kc3 45.Ke5 Kd2 46.Kxd4 Ke2 47.c4 Kf2 48.cxb5 Kxg2 49.b6 f3 50.b7 f2 51.b8Q f1Q it is unclear whether white has a realistic chance of winning this against good defence 52.Qc8 Qf4+ 53.Kd5 Qd2+ 54.Ke4 Qf4+ 55.Kd3 Qg3+ 56.Kc4 Qf4+ 57.Kb5 Qe5+ 58.Kb4 Qe1+ 59.Kb3 Qb1+ 60.Ka4 Qd1+ 61.Ka5 Qe1+ 62.Kb6 Qe3+ 63.Kc7 Qc3+ 64.Kd8 Qh8+ 65.Kc7 Qc3+ 66.Kb7 Qb2+ 67.Kc6 Qc3+ 68.Kb7 Qb2+ 69.Kc6 Qc3+ 70.Kb7 Qb2+ In this case black can keep checking until the 50 move limit or a 3 fold repetition] 37.Ra5+ I oddly thought that my opponent had blundered and allowed R+ K moves Rxg5 for free, but overlooked the obvious b5, however it turned out to be a blessing in disguise as it force the position to resolve a little, and clearly I was never going to see that RxR was a winning pawn ending for white and had made my mind up it was drawn or a dodgy looking pawn ending, and the following sequence turns out fine for white 37...b5 38.a4! white has this very pleasant move. I was anxious still about Rg3 and Rxg2+ then threatening to push the f pawn, but this turns out to be harmless 38...Rg3 the logical test for white 39.Rxb5+ Kc6 40.Rf5 there is no fight left in the position technically, although I still took my time and made sure to calculate carefully [F prefers 40.b3 giving a slightly better evaluation than the game move, but as a human the best computer move is not necessarily the best human move, and it's better to simplify and take the sting out of the position as long as it doesn't give up a winning advantage (or doesn't make the win more difficult) 40...Rxg2+ 41.Kd3] 40...d3 this is not good, but black's position is already hopeless [40...Rxg2+ was the most logical test to white 41.Kd3 e.g. 41...Kd6 42.a5 Ke6 43.Rf8 Rg1 44.Kxd4 Rd1+ 45.Ke4 Ra1 46.b4 Rb1 47.Rb8 white is still easily winnig but has to be a bit careful] 41.cxd3 Rxg2+ 42.Kc3 having spent most of the time at the board calculating if my king could get in front of black's threatened passed pawns and prever a loss, this was another such calculation, and I saw that I would have time to go via d4 to e3 and stop the black f pawn with my king and rook. The fact that black's king is cut off by my rook and therefore cannot prevent Kd4 is just salt in the wound for him. 42...Rg3 43.Kd4 Kb6 there are no good moves for black 44.b4 f3 45.Ke3 from about here on my opponent, (who had only about 5 minutes left on the clock compared to about 15 minutes), blitzed out his moves; I believe this was in the hope that I would also blitz out my moves and increase the probability of a blunder. I am too experienced for that however, and calmly took my time (but not too much time) on each move, double and triple checking to make sure there were no unexpected tricks from black. 45...Rxh3 46.Rxf3 Rh4 47.d4 g4 48.Rg3 Rh2 49.Rxg4 Rh3+ by about now it was slightly farcical, as I clearly wasn't rushing my moves and was up on the clock, but I suppose there is always the chance of a horrid stalemate if white is really stupid. I remember thinking that drawing such a position would be substantially worse than the average loss. At a couple of points in the remaining moves a smiled at some of the Kibitzers watching, but I wasn't really happy until the game was officially won. 50.Ke4 Rb3 51.Rg6+ Kb7 52.b5 Rb4 53.Ra6 Rc4 54.Kd5 Rc3 55.Rc6 Ra3 56.Ra6 Rc3 57.a5 Rc2 58.Rc6 Rh2 59.a6+ Ka7 60.Rc7+ Kb6 61.Rb7+ Ka5 62.a7 Rh5+ 63.Kc4 Rh8 64.Rb8 I had no doubt that black couldn't seriously drum up any threats since about move 49, but it still disturbed me a little since I expected black to resign. Since not playing for a couple of years this was the first win against a strong club player, and also after losing a lot of games against strong opponents already this year I was a bit low in confidence in my playing, so it was still a relief to finally win the game. Many of the endgame positions in this game could be analysed in great depth, and it's quite difficult to decide which lines to look at and how deeply. No doubt for this reason it is important to always try and think in terms of plans for both sides and use that as a guide, as it's impossible to calculate exhaustively all the variations to any depth in many of these positions. It seems white always had a much easier game because of the structural advantage, and the onus was on black to create active counterplay whereas white's advantage was permanent and therefore wouldn't have to worry about acting especially quickly or dynamically but playing methodically. One of the most interesting oversights in this game is the fact the RxR pawn ending was winning for white, and both players overlooked this about 3 times. Overall black put up a respectable defence for most of the game in a very difficult endgame position for him. On a side note it was nice to get revenge against an opponent who beat me two years previously in exactly the same opening with the same colours and in a game where he beat me in the endgame. 1–0
No comments:
Post a Comment