In my last blog, I wrote about two of my games as black in Navalmoral where I changed my standard opening choices at the board without full preparation. In this one, I’d like to focus on my lone loss of the trip, to GM Sergey Fedorchuk. Fedorchuk was the top seed in the event and at 2619 FIDE, even seemed a little underrated. He had been as high as 2671 in April 2008 and was 2655 back in July 2009 before hitting a big cold streak to drop to 2619.
This was the evening round of a double-header, so neither of us was probably as rested or prepared as we might have liked to be, but such was the schedule for the event. As it was, there wasn’t really enough time to prepare anything new, but I did have some time to look at some of his games and figure out what I wanted to do.
I ended up choosing the Graf Variation of the Ruy Lopez, not the most solid line, but it’s the line I have played the most within the main line Ruy Lopez. Given the fact that I didn’t have much time to prepare, it seemed best to go with what I knew the best.
After 14.Qd1-e2 (a novelty), we reached the following position:
I didn’t really understand the main point behind 14.Qe2 at this point in the game. Before this, White had two standard plans – 14.Nf1, which was normally met with 14…f5, and 14.b4, which was normally met with 14…cxb4 15.cxb4 Nac4. In both cases, Black’s results were fine.
I played 14…f5 after a few minutes of thought, and he replied with 15.b4. If I continue along the standard path with 15…cxb4 16.cxb4 Nac4 17.Nxc4 Nxc4, I thought that 14.Qe2 was a high-class waiting move, as now he can try something like 18.a4 Bd7 19.axb5 axb5 20.Rxa8 Qxa8, and given that the e6-square has been weakened by …f7-f5, he can choose from 21.exf5 gxf5 22.Ng5 or 21.Ng5. That was sort of along the right paths – while the rather direct plans I looked at with Ng5 don’t seem so dangerous now, he can instead play 18.a4 Bd7 19.Bh6 Rf7 20.exf5 with a small advantage. Compared to the normal situation without Qe2/…f5 included, Black’s kingside is a bit weaker and his pieces will be stretched a little thin covering both the queenside and kingside.
However, in the normal lines with 14.b4 (instead of 14.Qe2), I knew that there was also a rare idea with 14…Nb7. Black’s idea there is to play …a5 and force White to capture on c5 or a5 at some point, thereby freeing the knight back up for active duty.
Thus, when I played 14…f5, I planned 15.b4 Nb7, but now he played 16.a4!, and I began to understand the main point behind 14.Qe2. It’s not common for White to want to give up his bishop for a knight in the Lopez (as he would do after 16…Nxa4 17.Bxa4 bxa4) but then, after 18.Rxa4, Black’s got a real problem. His knight on b7 has no active prospects and the a6-pawn is difficult to defend. If Black then plays …a5, White steps past with b4-b5, getting a past pawn and leaving the Nb7 to its fate. Black can exchange on b4 before playing …a5, but then White uses the Queen on e2 to good effect. He will exchange once (or twice if Black recaptures with his bishop) on f5 and then play Nf3-d4! If Black leaves the knight on d4, then the e6- and c6-squares are extremely nice for the knight, while if Black takes the knight, White takes the bishop on e7. Black is left with weak, doubled d-pawns and bad knights.
I’m not sure how much staying power 14.Qe2 will have once people understand what it’s about, but it’s a pretty subtle idea and it was quite annoying to face over the board.
Fast forward and we reached the following position after 28.Bb1-a2:
He was disappointed that this was all he got from the opening, as he thought his 16.a4 idea was good enough for a clear advantage. Here, though, it’s not clear at all what’s going on. White has a rook and pawn for 2 knights, but Black’s kingside is a bit exposed and his minor pieces aren’t so well coordinated. If Black’s king was safe or his minor pieces were working well together, then Black would definitely be better here.
I had 11 minute left before this move (no second time control, but a 30-second increment), and I spent over 9 of them trying to decide what to do here. In the end, I played 28…Bg5!?!?, which really threw him for a loop.
I had spent most of my time on 28…Nc4-e5, but maybe it was the fatigue or something else, but I didn’t do a very good job of calculating here. The knight retreat is the most desirable move, since it improves my kingside defenses and coordination, but I was worried about 29.Rxf6 Qxf6 30.Qxe5 Qxe5 31.Rxe5, when I wasn’t sure what was going on. I looked at 31…Rc2 32.Re7! Rxa2? 33.Rg7+!, when Black either gets mated after 33…Kh8 34.Rxd7 and 35.Rd8 or ends up down a piece after 33…Kf8 34.Rxd7+ and 35.Rxd6.
I didn’t seem to appreciate 32…Bb5!, though, after which White has to find the very tricky 33.Rg7+ Kh8 34.Bd3!, with the idea of 34…Bxd3 35.Rd7, threatening the knight and mate on d8. He did see this idea, but as it turns out, the simple 35…Kg8 avoids the mate and leaves Black with a definite advantage in the endgame. He missed 35…Kg8 as well, for what’s it worth.
Having missed this sequence, though, I played 28…Bg5 to force a resolution to the kingside tension. If he takes on g5, then …Qxg5 will lift the bind on my king, while after 29.Rf8+ Qxf8 30.Bxf8 Kxf8, I thought my 3 minor pieces would be able to hold his queen off. I banked on there not being so much of an attack now, as White has no entry on the e-file and without his dark-squared bishop, he can’t use the weak kingside squares as well.
It wasn’t a horrible idea (just not the best), and it took a series of excellent moves from Fedorchuk to avoid letting me regroup and then go to work with my better minor pieces. The game continued 31.Qf3+ Kg7 32.Qg3! Nf7 33.h4 Bf6 34.g5 Be5 35.Qf3 Bf5 36.d6! Ncxd6 37.Qd5!, reaching the following position:
White has shed one pawn but activated his queen and bishop. Black’s clump of minor pieces look nice, but the bishop on e5 is under attack and if it moves, White’s rook invades on e7. Thus, I played 37…Re8, thinking that on 38.f4, I’d get a chance to play …Bd4+ at some point, breaking the pin and grabbing the rook on e1. Unfortunately, I forgot that …Bd4 check is always met with Qxd4 check!
After 38.f4 Be6 39.Qd2 Bxa2 40.fxe5! Nxe5, White had a choice. At this point, we were both down to playing on the increment (I generally fluctuated between having 35 and 45 seconds after making my moves and getting the 30-second bonus, while he was a bit closer to a minute), so we didn’t have much time to make our decisions.
As it turns out, 41.Kh1 is now the best move, but it’s not an easy move to come up with in time pressure. He chose the more natural 41.Rxe5 Rxe5, and again had a big choice – take the knight on d6 or the bishop on a2? After the game, the post-mortem, and a brief look on my own, I’m still not sure what the correct decision is or what the correct result is in either case!
In the post-mortem, we were joined by GMs Korneev and Shchekachev. Both of them thought the endgame was a draw in either case, but that leaving me with the bishop was a much simpler draw for Black. I was not sure what was going on, but I figured I would have better drawing chances with the bishop than with the knight. Fedorchuk wasn’t completely sure, but he was strongly leaning towards a win if he had left me with the bishop, but possibly a draw if he took the knight!
He had 30 seconds to decide, and in the end, he took the bishop, leaving me with a rook, knight, and pawn against his queen. After 42.Qxa2, I played 42…Nf7, hoping for 43.Qxa4 Re6, when I think Black might have a fortress of sorts. If White exchanges the b-pawn for the a-pawn, I’m pretty sure it’s a draw. But if he doesn’t exchange, Black’s rook goes to d6, and then he just shuffles with his king, as everything is protected. White’s king can’t enter Black’s position, so I’m not sure how White can win.
Instead of this, he played 43.Qc4!, which was a very strong move and a nasty shock. Now I didn’t know how to try and set up a fortress. I played 43…Rb5, guarding the pawn, but after 44.Kf2, he walked his king over to the b4-pawn, and then went after my a6-pawn with his queen. Eventually he broke down the defenses and pushed his b-pawn through.
Instead of 43…Rb5, I now think that maybe Black can draw with 43…a5!. After 44.b5, Black now plays 44…a3 45.b6 Re1+ 46.Kf2 Rb1, getting behind the b-pawn. White’s queen can help the pawn advance, but then Black’s a-pawn will cause serious trouble. But if White goes back to get the a-pawn, then the b-pawn will fall. Thus, I think that it’s a draw, but this isn’t the most intuitive way to do it – Black takes his rook away from the safety of his pawns or knight and hangs on by a thread.
Still, I’m not very sure about this endgame, or the one with the R+B vs Q. While Korneev and Shchekachev were certain it was a draw, they were unable to prove it to us in the 20-minutes or so that we spent on it in the post-mortem. White always had a series of tricky checks to make small inroads into Black’s position.
Just wanted to drop you a line and say thanks for doing this blog! You always have interesting stuff to write and I enjoy following it.
Complicated game vs. Fedorchuk. I would tend to agree with you that that the R+B vs. Q ending gives you better chances to draw than the R+N vs. Q ending. At the same time, that 43…a5! resource you discovered seems to hold up, so I guess I dunno either :) Great analysis regardless.