Note: The above was not the right game-score, look for the corrected game-score near the bottom of this post.
I had never lost to Ken before, but he had tied for first last month. In this game, I got into time-trouble, then took a guess that 20…Qa5 would work. After the game I said that I was winning, but blundered with …Qa5, which is exactly right.
The correct defense is 20…Rf4, and then Black can sac the exchange if Nh4, and then play …e5. I had the feeling I could sac the exchange and still be winning (keeping the bishop-pair), and I also wanted to play ..e5. I never saw ..Rf4 though, didn’t know if I should defend by attacking or by defending.
This was a game where I needed the 40/2, G/1 time-control. I spent a lot of time calculating complications up to move 20, and it took me too much time to do so. It took me half an hour to figure that 16…bxc works, a critical moment in the game where a move simply needed to be calculated.
The game went according to plan. I knew that I was going to get attacked on the kingside with this variation, I just underestimated how much time that it takes to calculate on defense in complications.
I knew to look out for his Nh4-g6 mate, but was hoping my attack would stop that. When he played 21.Nh4, that is when I realized that Bxd4 loses to Nh4-g6+xRf8. So I grabbed the g-pawn with my queen, not even realizing that Ng6+ is protected by his queen. In my mind, I thought that the queen would have been pinned already by a …Rxc3 move, but I was a tempo short of being able to implement that plan.
After I gave up my queen, I thought I still had some chances, but didn’t have enough time compared to him, and blundered with two minutes left, but he was going to win anyway, was looking at strong moves like RxBg7 and Rh3.
I needed to find the defensive move first, and only then see if I could defend by attacking, but instead I was tempted by that big piece of cheese in the mouse-trap, and it always gets tempting to play a quick decisive-looking attacking move late in the game.
I considered that I could bail out of the position with 20..BxNf3, 21.NxBf3 RxNf3, 22. RxR Qg5+, but that seemed like it could conceed my advantage, and it is nearly +1 for White. It seemed as if Black should have something which doesn’t give away the advantage, chess-justice sort of thing, but I simply ended up being the optimist who wasn’t thinking about defense. Particularly after 19.bxc, that is when my pawn push plan is thwarted and I should go back to the bail-out plan, but I still wanted to find the win somewhere.
The other bad thing for about G/90 is that my opponent, as I had predicted, used my time to think so when I did play those defensive moves, his move immediately followed since he had used my time to do all of his thinking. It’s tempting to just move quickly to prevent the opponent from doing that, but that doesn’t work out. This is something which makes defense at G/90 even less practicable. One really has to save their time ahead of time as Black in these sort of situations.
White’s position falls apart pretty quickly after 20…Rf4. So close, and yet play that position wrong (i.e, not playing 20..Rf4) and then there all sorts of ways to lose for Black.
The problem with this sort of position, looking at moves 16-19, Black’s moves were non-obvious, but White’s moves were obvious. As the lower-rated player, this gave him a practical boost. To point out how difficult it would have been to reach an objective assessment on defense as Black, Crafty points out that 17..hxBg6!!:
A)18.hxg6 Qa5! wins.
B)18.h6 Bh8, 19.h7 Kg7 (I saw this during the game, didn’t like the looks of it for obvious aesthetic reasons), 20.Kb1 c3, 21.bxc Qc7, 22.Rc1 Rb8, 23.Ka1 Rfc8 is -4.5 in Black’s favor. This position looks paradoxical to the eye at first glance, and is counter-intuitive to find OTB in a stressful defensive situation. That would have been Kortchnoi-esque defending right there. Yes, this position qualifies as worthy fodder for a Stoyko-exercise! 😉 Yes, in this last line White has the h7 pawn for the light-bishop, but Black can win the h7 pawn with ..Nd7-f8xh7 and Black is up -3.
I am so not used to defending castled positions, and yet this was the challenge for me, why I liked this ..d6 variation. I got what I wanted, but winning advantages from these positions do not come cheaply.
So many subtleties. The Bxd4 variation doesn’t work after the c3 push and bxc. The reason that hxg6 is so strong is that White’s follow up move of g6xf7..Rxf7 gives Black the free tempo of ..Rxf7. During the game, I originally wanted to play ..Rf7 after ..h7xg6 Qxg6. This loses because Black has just drug the queen to g6 for free and spent a tempo on his/her own dime to play ..Rf7. ..Rf7 fails because of Rh8+..KxRh8, QxRf7 with Rh1 mate to follow (lucky for White the Nf3 is covering that light diagonal of Bb7-h1).
Here is another difficult move that I looked at. Instead of 16..bxc4, play 16..f5. There is a trick to this one. 16..f5, 17.gxf6 bxc4!, 18.Bxg6 Nxf6, 19.h6!! c3, 20.bxc3 h7xBg6, 21.h6xBg7 Kxg7, 22.Rg1 Qe8, 23.Rh3-g3 and it is curtains for Black. Naturally, I didn’t see anywhere near all of this at the board and simply had to pick the variation that looked best to me at that point in time at the board.
After this game I had one thought on my mind “Forget about the opponent, playing the board is what is hard in chess.” It’s not who I play, it’s how much time do I get for positions like that. I could quite likely be an Expert by now if all of my games were 40/2, G/1.
After I played 20..Qa5, I had 12 minutes left, and my opponent 52. It was so hard to not play such an obvious attacking move and instead, had I done it, try and solve the position. I need to be solving the position all of the time except when the advantage is very clear and in time-trouble.
Here is a variation that I almost played, but refrained, and then stuck with my original plan:
6..c5! (Crafty likes this move), 7.Nc3 cxd, 8.Nxd Nc6, 9.Nd4-b5 d5! – I didn’t see that this move was possible, I only saw myself as having the backward d-pawn on d7, when Nd6 is a clamp. This particular variation is -+, around -.7 or more.
After all that analysis, I remembered that I had had a choice between ..Kf7 and ..Kh8, and I spent exactly one minute choosing between the two, such that I had 11 minutes left afterward. So …Kh8 was the losing move. I knew that Ng6 mate had to be prevented should I play Kh8, but This is where I thought that Nh4 = Bxd4 in reply, and I did not calculate it at all. It seemed as if every move was a major decision and at some point I would have to play major decisions quickly, but this is not so, merely a distorted perception. The game does calm down into a reasonable blitz position soon after a move such as ..Kf7, when White is +1 but would probably lose the game in practical terms.
..Qa5 was also quite the mistake as well because I had ..hxg which is winning (It’s interesting that a defensive move is the winning one). This is quite an obvious blunder looking at it now, but at the time I missed that ..hxg, Qxg Rxc3+, Kb2 RxNf3 – I hadn’t thought about that the Nf3 is being attacked three times and defended only twice, I had only noticed that RxN seemed to lose an exchange (and if Black doesn’t play ..RxNf3 there, then Black loses). I simply played too quickly here, it wasn’t time to blitz yet.
Lest we think it’s a wrap after Black plays ..RxNf3, it would still lose the game to then continue Nd2xRf3 Bb7xRf3??, Qh7+ Kf8, Rxf3+ Ke7, QxBg7+, etc. OTH, Nd2xRf3 RxNf3? only leads to a draw by perpetual with best play from White (although I doubt he would find best play, and same goes for me). So the immediate ‘bowing and scraping’ to immediately regain the piece on f3 does not work. Instead, after Nd2xRf3, Black needs to play ..Qb6+!, Ka1 Rxf3, Qxe6+ Kf8, RxBf3 BxRf3, Qf5+ Ke7, QxBf3 Bd4+, RxBd4 QxRd4+ and Black is up a knight for a pawn – exchange queens and wins. If Kc2 instead of Ka1, then Black doesn’t even worry about regaining the piece on f3 (White has rook and pawn for two minor pieces) but rather uses the bishop pair to help in a king-hunt mating attack where Black is up nearly -6 after playing the nice idea ..Rf8-c8+ followed ..Nf8, kicking out the queen from g6 and also defending h7 and e6. I have to admit that I had I still been looking to make quick moves without calculating, I feel that I would have played the Bb7xRf3?? move. So what likely would have happened then? I think there’s a good chance he misses Qh7+ winning and plays Qxe6+ Rf7, g6 Qb6+ where Black can force a draw by repetition with queen attacking king, and if king moves toward the center it gets mated. But I think he may have found Qh7+. I told him to study tactics a couple months back, and he is seeing them quickly now.
The hard part to imagine is that even had I played all the way to ..Rc8+ and instead of playing the ..Nf8 idea on the next move, which does not readily spring to mind, but instead gotten lazy at this point and played ..BxNf3 (just to play a quick move, right?) then either Qh7+ or Qe6+ are majorly winning for White.
To further clarify the complexity of the position, ..Rxc3+ is not even the best move. The best move after 19..hxg6, 20.Qxg6 is ..Qa5 winning. 21.c4 (everything loses quickly here) Qf5 (trading queens leaves two pawns for the piece and not for long at that). 22.Qh5 Nb6 (-6) when the cheeky 23.Na4 fails to QxRh3, ..g6 (threatening mate on h7) 24.Rxc3+ NxRc3, 25.Qc3+ Kb1, 26.Be4+ Rd3, 27.BxRd3 mate.
When I played …Qa5, I still wasn’t sure what to do yet on my next move, I was playing hope-chess in terms of being able to think on his clock time, which didn’t happen until the game was already in the tank. I had wondered if this were going to be another one of those games where I could win if I played it down to the nubs on the clock, and it probably would have, but I obviously didn’t want to go down that road, though should have.
Now I know why it is so much more difficult for me playing Black at G/90. One big reason is that, for example, it’s difficult for Black to whether ..Bxd4 will work in response to Nh4. Black has to spend time calculating whether it will work or not. OTH, if White merely notices that Ng6 is mate, even an 1100 player could play Nh4 (good chance they don’t notice ..Bxd4 reply anyhow) and after ..Bxd4, they would HAVE to play Ng6+ followed by NgxRf8 anyway, whether it works out or not. It’s up to BLACK to avoid this variation altogether, like many other variations that were avoided.