Cosham A vs Portsmouth A, 03.11.2011
The following was a sharp Sicilian type game in which I got a big advantage early on but managed to throw most of the advantage away in a few inaccurate moves, allowing my opponent strong counterplay, and meanwhile having got into bad time trouble. During my frantic time trouble towards the end, my opponent was playing fast as well - perhaps unwisely, as it is recommended to take your time if your opponent is in bad time trouble and you have lots of time - and I had a crushing attack after an inaccurate move by my opponent, however I failed to find the right continuation in the time scramble and went on to resign before losing on time at some point in the endgame after some frantic moves. This game is a good example of keeping things complicated and creating counterplay in an inferior position (from black's point of view) and also demonstrates how it is treacherous to fall too far behind on the clock even in a technically winning position.
16.Nf4
..31.Qe6
43.Qg5+
1.e4 g6 2.Nf3 I decided on a flexible and sound move. I didn't want to commit to d4 immediately in case of running into unfamiliar territory such as a King's Indian Defence, so I kept it an 'e4' game. 2...Bg7 3.Nc3 I still did not want to commit to any opening so this seems like a strong and logical general move leaving white plenty of options. 3...c5 Sicilian dragon (ish) transposition; an indication it will be a sharp game. 4.g3 I opt for a closed sicilian type approach, although the King's knight is usually developed to e2 so it is going to be sharped than usual. [4.d4 the most common; 4.Bc4 also good] 4...Nc6 5.Bg2 d6 6.d3 [or 6.0–0 ] 6...Bd7 An indication my opponent will use a Qc8 Bd7 setup to trade off my light squared bishop and weaken my king's castled position on the kingside (if I castle there) 7.Be3 Qc8 Fritz gave a tiny edge for black before this but prefers 7.Nf6, fully developing the minor pieces and preparing to castle. In fact the lack of development on the kingside for black is a critical factor in the game. 8.Qd2 [8.0–0 F is happy to castle and allow the bishop trade 8...Bh3 F says this is not even good for black, and nothing to white to worry about after: (F prefers 8...Nf6 ) 9.Bxh3 (9.Bg5 Bxg2 10.Kxg2 Nf6 is another option among several for white) 9...Qxh3 10.Nd5 threatening Nc7+ 10...Qd7 e.g. 11.d4 cxd4 12.Bxd4 Nxd4 13.Nxd4 Nf6 14.Qd3 F thinks this is about equal; I don't think I'd be that happy long term as black's bishop is strong] 8...Rb8 I thought this looked a tad fishy, ignoring the development of the knight 9.h3 I decide to go for the plan h3 g4 then castle to avoid the trade of the bishop on g2. F doesn't like the move too much and preferred castling straight away, (which seems logical since black is a little behind in development. [9.0–0 Nf6 (9...Bh3 10.a3 e.g. 10...Nf6) 10.Bh6] 9...b5 not unsound, but still the lack of developing the King's knight is starting to play with fire a little perhaps 10.g4?! Fritz isn't keen on this. I think it's ok though, and there is a logical reason for it - avoiding the bishop trade. F now gives half a pawn to black. [F prefers 10.0–0 ; 10.a3 the other move F suggests; logically contesting the b4 square] 10...h5?! F flags this move up as well, restoring the evaluation to about equal [10...b4 the critical line according to F 11.Nd1 Nf6 This doesn't look too pretty for white, so I can see why F didn't like 10.g4 : ). Black's setup looks very neat here.] 11.g5?! Another questionable move according to F. Advantage black apparently (half a pawn). [F suggests 11.0–0–0 e.g. 11...Qa6 12.Kb1 Nf6 which would certainly lead to an insane game - usually Fritz's suggestions are a bit 'stodgy', but I suppose the dictates of the position mean 0–0–0 was the strongest option] 11...h4?! Deuce again according to Fritz. [11...b4 12.Nd1 Nd4 would have clearly been unpleasant for white 13.Nh4 e.g. (13.c3? bxc3 14.Nxc3 (14.bxc3 Bxh3 15.Bxh3 (15.cxd4?? even worse 15...Bxg2) 15...Nxf3+ 16.Ke2 Nxd2 17.Bxc8 Nxe4 18.dxe4 Rxc8 white is badly losing) 14...Bxh3) 13...Qa6 14.0–0 e6 this definitely looks a lot better for black] 12.d4 At last a move that the computer agrees with. [F also flirts with 12.0–0–0 b4 13.Ne2 e5 but ultimately prefers 12.d4] 12...b4 13.Ne2 c4 [F prefers 13...Na5 aiming at the c4 square so that e.g. 14.0–0–0 Nc4 is good for black 15.Qe1 a5] 14.c3 [F considers 14.Qd1 Na5 15.a4 b3 16.c3 e6; F also likes 14.Nf4 e6 15.a3 all complicated positions that could consume hours of analysis in their own right] 14...Qa6?! F doesn't like this and gives almost a pawn evaluation to white by now. I remember feeling I had been let off lightly to be able to castle for free now the queen bishop battery has been broken [alternatively 14...e6 this allows white to castle, but black completes development also 15.0–0 Nge7; 14...bxc3 is ok too 15.Qxc3 F likes recapturing with the queen, which to a human looks like a slightly risky move(15.bxc3 e5 16.d5 Na5 17.Qc2 for example 17...Ne7 (17...f5 lots of options) ) ] 15.0–0 f6? A definite mistake. Black still hasn't developed the kingside and this doesn't help. The main tactical problem is the gaping weakness on g6 allowing the following crushing move. [15...bxc3 is good; 15...e6 is also better, but F gives over a pawn to white here still. E.g.: 16.d5 exd5 17.exd5 bxc3 (e.g. 17...Ne5 18.Nxe5 Bxe5 19.cxb4) 18.Nxc3 Nb4 19.Bd4 Rh7 (19...Bxd4 20.Qxd4 Rh7 not nice for black) ] 16.Nf4 F gives white a 2 pawns advantage. 16...Qc8?! F doesn't like this. It's hard to suggest good moves though as the position is now extremely unpleasant for black. [F prefers 16...fxg5 17.Nxg5 bxc3 18.bxc3 Qc8 19.Nxg6 Rh6 attacking the knight 20.Nf4 Nf6 white has a much better position (2 pawns according to F) but it's an extremely sharp and complex position (in my opinion)] 17.Nxg6?! taking the pawn (and attacking the rook) straight away is inaccurate according to fritz. None the less, I felt confident I was clearly winning at this point. [17.d5 bxc3 18.bxc3 Nd8 19.Bxa7 fxg5 20.Nxg5 is it typical of the computer to refuse to take a pawn straight away and going for superior move orders however complicated they look. One thing to keep in mind is computers have zero fear of complications and only care about the numerical evaluation, whereas in human vs human games it is often stupid for the winner to go in for a complicated continuation to try and squeeze a little more advantage out at the risk of blundering when they already have enough advantage to win simply. This does not apply to this game however as I was not able to find any way to win the game simply at all.] 17...Rh5?! Understandable to move the rook, but this square does not look right. [17...Rh7 is not better though according to Fritz, e.g. 18.Qe2 Nd8 19.Nf4 Ne6 20.Nxe6 Bxe6 21.g6 Rh5 22.Kh2 Bd7 23.Rfc1; 17...Bxh3 F says this is the best move. White has to be very careful 18.Bxh3 (18.Nxh8?? Bxg2 19.Kxg2 Qg4+ 20.Kh2 Qxf3 21.Ng6 Qxe4 A super-complicated position that is muddy and definitely wasn't what white had in mind) 18...Qxh3 19.Qe2 Qg4+ (19...Rh7 20.Nf4 Qg4+ 21.Kh2 Nd8 22.Rg1 Qc8) 20.Kh2 Qxe4 21.Nxh8] 18.Kh2 F doesn't like this (but still gives white 2 pawns). My idea was to prepare Rg1 to use the rook on the opening g file (which seems a reasonable human move). [the computer goes for 18.Qe2 threatening a revealed attack on the rook, e.g. 18...bxc3 19.bxc3 fxg5 (19...Na5 20.Nfe5 Rh7 21.Nxd7 (21.gxf6 Nxf6 (21...Bxf6 22.Nxd7 Qxd7 23.Qc2 there are endless possible lines but all hopeless for black 23...e5 24.dxe5 Bxe5 25.f4 Rg7 26.fxe5 Rxg6 27.exd6 Qxd6 28.Rad1 (F gives white about 4 pawns, even though in material he is only one pawn up. The two knights are a joke compared to white's bishops in this position.); 21...dxe5 22.fxg7 Rxg7 23.Qh5) 22.Nxd7 Nxd7 23.Rab1) 21...Qxd7 22.f4) 20.Nxg5 Rh6 21.f4] 18...Rb5?! A mistake according to F. [F prefers 18...fxg5 19.Nxg5 bxc3 20.bxc3 Rh6 21.Nf4] 19.Nf4 My objective at this point is to try and win the game as simply as possible (if possible) but black turns out to have nasty counterplay which although losing in evaluation to the computer can be difficult for a human to deal with [19.Qe2 is stronger according to F, cranking up the pressure with threat of revealed attack on the rook 19...Nd8 and insanely complicated lines, but black clearly has a horrible position 20.e5 fxg5 a) 20...dxe5 21.Nfxe5; b) 20...fxe5 21.Nfxe5 (21.dxe5 bxc3 (21...dxe5 22.Nfxe5 Bxe5+ 23.Nxe5 Rh7 24.Nxd7 Qxd7 25.Qxc4) 22.bxc3 Rh7 23.exd6) ; 21.exd6 Rh6] 19...Rbxg5?! I felt sure this was wrong, and was getting more confident about the game situation with every move [19...Rh8 at least avoids immediate loss of the exchange 20.d5 Ne5 21.Nxe5 fxg5 (21...fxe5 22.Ng6 Rh5 23.Qe2 Rh7 24.f4) 22.Nfg6 Bxe5+ 23.f4 gxf4 24.Bd4 bxc3 25.Qxc3 Rh7 26.Nxe5 dxe5 27.Bxe5 A couple of examples of the many comlicated possibilities] 20.Nxg5 I was choosing between this and Nxh5; I think I went for Nxg5 since after Rx there is the threat of a releaved attack on the rook by the bishop when the f4 knight moves, and after fx the rook on h5 looks a little hopeless 20...Rxg5 21.Rg1 the computer agrees with this move; logically playing the rook on the open g file 21...e5 objectely there are no good moves so this is as good as anything. Whatever complicates the position the most is the best option for black 22.Nd5 a very nice move to be able to play, taking a great square with the knight at the same time as a revealed attack on the rook [F suggests 22.Ne2 bxc3 23.Nxc3 Bh6 24.dxe5 (24.Bxg5 a bad move according to F 24...Bxg5 25.Qd1 Nxd4 26.Nd5 looks good for white but very sharp) 24...dxe5 25.Nb5 Qb8 26.Nd6+ Ke7] 22...bxc3 23.bxc3 Rg6 [23...Rh5 prefers this slight, although it shouldn't make much difference now and the evaluation is about 5 pawns to white. The trick thing is, it seems like white has to open up the position to win, and this means potentially dangerous and sharp counterplay for black and more opportunities to blunder 24.dxe5 Nxe5 25.f4 Nd3 26.Bf1] 24.Bf3 logically threatening Bh5 pinning the rook and exposing the file to trade rooks and play on the g file 24...Rxg1 25.Rxg1 Kf8 26.Bg4?! I naturally wanted to simplify the position as much as possible and win the endgame a rook up, however F changes white's advantage from 6 pawns to 3 pawns here (which is still more than enough to win objectively, but cannot afford to play too many inaccurate moves in such a sharp position) [26.dxe5 it doesn't feel right giving the black knight the strong d5 square like this, but the position needs to be opened to deliver the advantage and make use of the bishops. This was kind of a case of wanting it all, and not being willing to give my opponent any perceived plus to his position, whereas sometimes you have to give back a little to take even more. 26...Nxe5 27.Bh5 full use is being made of all the pieces now and the position is totally crushing, and the nice knight on d5 is irrelevant to the overall disaster of black's position. 27...Qb7 28.Nxf6! Nf3+ (28...Nxf6 29.Qxd6+ Kg8 30.Rxg7+ Kxg7 31.Qxe5 despite both sides having a queen and two minor pieces, Fritz gives white a 20 pawn advantage! All moves lead to mate: 31...Qc8 a) 31...Qxe4 32.Qg5+ Kh7 33.Bf7 theatening Qh6# 33...Ng8 34.Qxg8#; b) 31...Qc6 32.Qe7+ Kh8 33.Bd4 Bf5 34.exf5 (34.Bxf6+ Qxf6 35.Qxf6+) ; 32.Qg5+ Kh7 33.Qg6+ Kh8 34.Qxf6+ Kg8 35.Qf7+ Kh8 36.Bd4#) 29.Bxf3 Nxf6 30.Qxd6+ Kf7 31.Bd4 the position is now a forced mate in all variations, the most obvious threat being Rxg7 then Qxf6 31...Be6 if this then 32.Bxf6 Kxf6 33.Qf4+ Ke7 34.Rxg7+; 26.Bh5 was also stronger than the game move according to F] 26...f5?! apparently giving white about 4 1/2 pawns advantage (instead of 3). From a practical point of view this move is pretty good I think as the position is complicated, which is exactly what black wants [F prefers 26...Bxg4 27.hxg4] 27.exf5?! another inaccurate move according to F [27.Bh5 I can't say I understand exactly why Fritz likes this much more, but it can be hard to see things from the point of view of a calculator] 27...Bxf5 the correct move according to F 28.f4?! A mistake according to F, dropping evaluation to under 2 pawns. Although this isn't a good move techically, it seems there should be several good move options for white here, as the advantage is such that it seems strange it is lost if the exactly right move is not played. [28.dxe5 Nxe5 29.Bh5 F suggests this, Morozevich style putting the bishop on an active diagonal that looks set to weave a mating net and letting the h pawn go 29...Bxh3 if bishop takes then 30.Nf4 Qd7 losing the bishop(if the bishop retreats, e.g. 30...Bd7 31.Qxd6+ forced mate in 7, e.g. 31...Ne7 32.Rxg7 Kxg7 33.Qxe5+ Kh7 34.Qxe7+) ; 28.Bxf5 wouldn't have been any better according to F, still just have 2 pawns for white or less 28...Qxf5; 28.Qd1 maybe a bit better than the game 28...exd4 29.cxd4; 28.Nb4 maybe worse than the game, e.g. 28...Nxb4 29.cxb4 Bxg4 30.Rxg4 Nf6 31.Rg1 Qf5 32.d5; 28.Bg5 maybe a bit better than the game move 28...exd4 29.Qe2 (29.cxd4 Bxd4 30.Ne3) 29...dxc3 (if 29...d3 then 30.Qf3) 30.Qxc4 and so on, with ridiculously complicated positions] 28...exd4 correct according to F; the move that white should not have allowed; now black has some pretty serious counterplay 29.cxd4?! I think I was playing panicky by here as I was already running low on time and things had not gone as swimmingly as planned; this is obviously a stupid move though and I should have traded the bishop instead of making it a bad piece and entombing it. This is another move in the process of rotting a winning position away. F gives white about 1.5 pawns here, (and from a human point of view the counterplay is not easy to deal with). [29.Bxd4 Nxd4 30.cxd4 e.g. 30...Bxg4 31.Rxg4 Nf6 32.Nxf6 Bxf6 black would probably have kept more pieces on to keep the game more complicated, but in any case white does not have a useless bishop like in the game] 29...Nh6 spot on according to F 30.Bxf5?! another mistake according to F [30.Be2 apparently better, not gifting black a strong knight on f5. A difficult move to see though, having to jettison the h3 pawn 30...Bxh3 31.f5! throwing the pawn to activate the dark squared bishop 31...Nxf5 32.Bf2 a super complicated position; according to F white is about 2.5 pawns up] 30...Nxf5 Now F gives white just over a pawn advantage. From a practical point of view I was drastically down on time, (I thought earlier this wouldn't matter with what was a technically winning position). Black now has serious compensation for the exchange and the position is far from trivial for white, so this was a disaster situation as my opponent had well over half an hour if I remember correctly and I was on my last couple of minutes before the time control (which was at 30 moves I think). 31.Bf2?! Now the time control has been reached I am seriously gutted with how the game has gone from a 'clear' win to a very possible loss (especially given that I was so far down on time). This move is a serious mistake according to F and the evaluation is now about equal. [31.Qc2 the bishop needn't be moved as it is defended by the knight and is a poor piece anyway. This threatens the c4 pawn and actively uses the queen.] 31...Qe6 Another spot on move by Peter 32.Nc7 also correct according to F 32...Qd7 [or 32...Qf7 pretty much the same] 33.d5 this is a bit more like it; threatening Ne6+ and after Qx, dxc6 and the white queen is active. Fritz gives white about half a pawn, but again, I was nearing chronic time trouble and my opponent had a lot of time so the situation was pretty dire for me still 33...Ncd4 the correct response according to F [after 33...Qxc7 34.dxc6 Qxc6 35.Qc2 white has very good play and black's counterplay no longer looks very scary (F gives white about 2 pawns advantage)] 34.Ne6+ [34.Bxd4 may have been better 34...Nxd4 (34...Qxc7 35.Bxg7+ Nxg7 36.Qc3 Qf7 37.Qxc4 easily winning for white; 34...Bxd4? 35.Ne6+ Ke7 36.Nxd4 winning for white) 35.Rxg7! (35.Ne6+? Qxe6! 36.dxe6 Nf3+ 37.Kh1 Nxd2 38.e7+ Kxe7 (38...Kf7? 39.Rxg7+) ) 35...Nf3+ 36.Kg2 Qxg7+ 37.Kxf3] 34...Nxe6 Not too bad, and black still has drawing chances (technically - actualy due to the clock he has serious winning chances) but there is a clever tactic that is even better [34...Qxe6! 35.Bxd4 (35.dxe6? Nf3+ 36.Kh1 Nxd2 and black should win) 35...Nxd4 36.dxe6 Nf3+ 37.Kh1 Nxd2 38.e7+ Kxe7 39.Rxg7+ Kf6 40.Rg8 Fritz says this is a draw, but I think black has a harder position unless he is very confident at endgame technique] 35.dxe6 Qxe6 36.Re1 F agrees - this activates the rook nicely 36...c3? a mistake according to F, and should lose a pawn, (evaluation from under a pawn to white to about 2.5 pawns) 37.Qc2? I return the favour and once again evaluation is not far from equal (about half a pawn to white) [37.Qxc3! Bxc3 (or 37...Qd7 38.Qf3 white has a much better position a clear exchange up) 38.Rxe6 now white should be cruising to a win] 37...Qd5 A good move according to F, although it also looks as some other queen moves [37...Qd7; 37...Qc8] 38.Qe2 Fritz agrees with white's move too; it is natural enough to threaten mate on e8 and to put the queen on a much more active position; for now the threat of the c3 pawn is not significant compared to piece activity for white 38...Qc6? a serious mistake according from F (evaluation from half a pawn to white to about 4 pawns to white) [38...Qf7 39.Qd3 Kg8 40.Rb1 although better for white, F still only gives white half a pawn] 39.Qc2? Returning the favour and missing the opportunity by repeating moves [39.Qe6! would have been pretty crushing (F: +4) threatening the knight and dominating the squares around the black king (actually stalemating the king from moving temporarily, which is usually a bad (or good) sign). 39...Ng3 (39...Qc5 can't be played due to 40.Qe8#) 40.Bxg3 hxg3+ 41.Kxg3 c2 white should win relatively easily 42.Qf5+ Kg8 43.Re7 Qc3+ (not 43...c1Q?? 44.Qf7+ Kh8 45.Qxg7#) 44.Kg4 Qf6 45.Rc7] 39...Qb5 [39...Qd5 was better according to F but still a slight plus for white (and not 0.00 as the computer dumbly evaluates due to the repeated moves, as white can play something else this time 40.Qa4 Qf7 41.Rb1 a little bit better for white according to F] 40.Rb1?! apparently inaccurate [40.Qe4] 40...Qd5 41.Qe2 F agrees with both of the player's last moves 41...Qc6?? a serious mistake; this immediately loses the game according to F (giving 12 pawns to white in evaluation) [41...Bh6 threatening Bx+ 42.Rb8+ Kf7 apparently about half a pawn to white] 42.Rb8+ devastating 42...Kf7 only move, other than giving up the queen for nothing 43.Qh5+! generous to give this an exclamation mark as it is easy to see with the positition in front of you, but in combination with 42.Rb8+ it is a killer 43...Kf6 44.Qg5+ Ke6 45.Qg6+ Bf6 all spot on so far from white according to Fritz. At this point I was in chronic time trouble, with about 2 minutes to finish the game, and I just couldn't find the winning moves from here on, and played inferior moves to avoid stalling and losing on time, but I knew soon that I had let it slip 46.Re8+? Given enough time on the clock this still would have been an easy win for white, (about 3.5 pawn advantage according to F), but I needed a checkmate soon realistically or I was going to inevitably lose on time [46.Rf8! totally crushing. This move was pointed out in the post game analysis by my teammate Ian Hardwick and he was spot on; it was hard for me to find this in the time trouble situation; it sometimes seems unintuitive to play a move that is not a check just in case you let your opponent back in, but I assume (at least I would like to think) I wouldn't have had too much trouble finding this with a bit more time to spare 46...Ne7 47.f5+ (47.Qxf6+ not as accurate 47...Kd7 48.Bxh4 Qe4 49.Qxc3) 47...Kd5 48.Qxf6 Qc7 having looked at this I don't feel quite as bad; I was expecting a forced mate but it looks like there never was one and I had to be satisfied with a crushing material advantage eventually leading to a win] 46...Ne7 47.f5+?? I already knew on the last move I had screwed it up a bit, but after this the position is at best equal [white should play the crushing 47.Bxh4 Kd7 if the king moves to d7 then 48.Qxf6 Kxe8? 49.Qxe7#] 47...Kd7 48.Qf7? [it would have been roughly equal after 48.Qxf6 Kxe8 49.Bxh4 Qe4 but inevitably I would have lost on time soon] 48...Qe4?! [48...c2 was pretty crushing 49.Qxf6 Kxe8 50.Bxh4 Qc7 51.Bg5] 49.Qxf6?! [49.Rf8 threatening Qe8+ 49...Qf4+ Fritz gives the repetition (49...Kc6?? 50.Qe8+ Kd5 51.Qb5#) 50.Kg2 Qe4+ 51.Kh2 Qf4+ 52.Kg2 Qe4+ 53.Kh2 Qf4+] 49...Kxe8 50.Qxc3 Qxf5?! It is better to take with the knight and activate the position more. It is remarkable that despite all of the huge tooings and froings of evaluation throughout the game, we now reach a very interesting endgame which really is roughly equal, (although I prefer white with a bishop in this open position). The only catch is that I had about 2 minutes left to finish the game and my opponent had at least 15 minutes, possibly half an hour or more, so I was inevitably going to either lose on time or fatally blunder. I don't have access to many of the remaining moves but I made every move almost instantly as my opponent moved. I eventually resigned with my flag due to fall any moment after I thought I had blundered, but thinking back to it we both had a pair of queens, and I had had just promoted to a queen and assumed I had blundered as it was connected to one of my opponents' queens, but it's counterpart was next to it so it was not en prise at all; I think this was a trick of the mind in chronic time trouble, but even if it was sound it would have been a complicated 2 queens each endgame and I would have had to lose on time soon. Still I should perhaps not have given up immediately until my flag had fallen. The real lesson for me from this game was not to get in time trouble in the first place, even if the position appears to be winning, as this game demonstrates how rapidly the evaluation can fluctuate one way or another in sharp positions. I was impressed with how well Peter played practically after I got a winning advantage from the opening, but creating difficult complications and playing on my time trouble; of course I have to take some of the credit for letting myself get into bad time trouble, but the counterplay that Peter created caused me a lot of difficulties and wiped the proverbial grin off of my face assuming I was in for an easy win. [50...Nxf5 F evaluates this as equal] 51.Qh8+ Kd7 52.Qxh4 white (in 10 or 15 moves) resigns 0–1
No comments:
Post a Comment