But I'm not here to write about any of the good stuff. I'm here to write about my sole loss from the tournament, against FM Jorge Leon Oquendo, USCF 2473, FIDE 2380. The game started kind of strangely, as you will see, and became very complicated. Mikhail Tal once wrote, "You must take your opponent into a deep dark forest where 2+2=5, and the path leading out is only wide enough for one." This I did! And though the balance swung a little this way and that, it was my opponent who eventually stumbled and fell! But two moves later, in a fit a utter carelessness, I missed the winning move and lost in turn. Yes, I lost a game I should have won against a player rated almost 2500.
But rather than feeling crushing disappointment at this missed opportunity (this is only the second time I have even played a senior master in a rated game), I feel quite good. After all, I did play mostly well enough to win against such a good opponent, and it gives me hope that I am once again starting to get better at the game. It has been many years since that was true, but circumstances have given me the means and opportunity to improve my game even at the advanced age of 48.
I've placed the game, with my hand-written notes, below the fold. I have partly supplemented them with a few computer suggestions. I'm not going to put it in the ChessBase viewer this time, so this post will be long. I will include plenty of diagrams so that the game can be followed. (The time control was Game in 45 minutes, with a five second delay. Numbers in parentheses after the moves show time remaining in minutes, and later in minutes & seconds.)
November Chess Challenge, Daytona, 2016 November 19
White: Jorge Leon Oquendo (USCF 2473)
Black: Todd Durham (USCF 1795)
1 d4 Nf6
2 Bg5
A minor surprise. I would have expected a senior master to play something more mainstream. But this occurred just a few days after Magnus Carlsen opened his World Championship match against Sergey Karjakin with a Trompowsky, and I guess it doesn't get more mainstream than that!
2 ... e6
I've experimented with 2 ... g6 against a couple of members of my club and various speed chess opponents online. If White doubles the f-pawns the game takes on a very different look strategically. It's more playable as Black than one might think. And if White doesn't take the Knight, Black takes the game into more KID- or Grunfeld-like waters, which White usually doesn't care for if he opens this way.
But here I just went with something more normal by my standards. With a shorter time control I wanted to rein in some of my more outlandish instincts.
3 e4 Be7
4 Bd3
Another surprise!
4 ... d5
5 e5 Nfd7
6 Be3 (42)
And we reach a rare position. I only have nine examples of this in my (incomplete) database, and Richard Palliser is the only player one might recognize in the bunch.
6 ... c5 (44)
Black has other options, but this seems most thematic & reasonable to me. We have a pseudo-French Classical developing, perhaps? I felt as though Uquendo was trying to out-fox me in the opening, but here I'm very comfortable, as the French is my main weapon against 1 e4. The position is level here.
7 c3 Nc6
8 Ne2 (40)
Now the game kind of resembles a French Advance variation in which White has developed his bishops first. I have exactly one example of this position in my db, between a 1900 & 2100 player. Thin gruel!
I can't believe, and didn't during the game, that the Bc1-g5-e3 maneuver can be considered anything other than a waste of time. Black essentially used the tempo gained to get his g8-knight to d7, which seems useful to me.
8 ... Qb6 (41)
I considered 8...Qc7, but that seemed wrong - illogical even. 8...O-O is also a reasonable looking move. I had started looking at ...f6, but it's just wrong now. I was fantasizing about maybe saccing a rook on f3 after some possible sequences involving Nb1-d2-f3 for White and ...O-O, ...f6, and fxe5 for Black. Most likely White plays f4 before the knight ever gets to f3, though. This last is pure fantasy, of course, and didn't even come close to happening in this game, but of such fantasy plans are formed.
Also, 8...Qb6 is a novelty, for those keeping score at home.
After 8 ... Qb6
9 b3 (39)White could play an interesting pawn sac here with 9 Nd2 c4 10 Bc2 Qxb2 11 a4 Qb6 12 O-O. This is a Stockfish suggestion. The idea seems to be to get in some more development while Black wastes time winning a pawn. The move played has a slightly better evaluation, but the sacrifice might be hard to meet against a superior opponent, which Oquendo obviously is. (Ratings don't lie, my friends.) These guys typically know how to handle superior development and an initiative!
9 ... cxd4 (38)
10 cxd4
White spend no time considering other recaptures. I had spent just enough time on my ninth to decide that I wasn't worried about other recaptures, which are worse.
10 ... Nb4
Playing for the two bishops. That may seem a little silly given the center & my light-squared bishop, but it's a thing. And I WILL be getting the better of White's two bishops. 10...a6 and 10...O-O both look reasonable, but once White moves his knight on b1 I won't be able to win that bishop, so it's now or never on that front.
11 O-O O-O (37)
Here I started thinking I had gone from equal to a little worse. (I haven't.) My position started to look a little over-extended to me, and my pieces didn't look well-coordinated.
But this is a false impression. I look over extended because of my b4 knight, which is about to be exchanged and is causing White some minor problems (see note next move), and ...f6 will leave my position looking very coordinated if White plays exf6, as ...Nxf6 follows.
After 11 ... O-O
12 Nc3 (33)
White spent six minutes on this move, over 15% of his remaining time. He'd probably prefer to play knight to d2 & hence to f3, but the d3 bishop would hang. And 12 a3 to force the knight to do something would mean weakening his queen-side pawns. (An eventual b3-b4 wouldn't help in that regard, and would leave a juicy c4-square for Black's remaining knight.) This is why I didn't take the bishop on move eleven - let him solve a problem! Oddly enough, both White bishops are slightly loose. For example 12 f4 f6 13 a3 Nc6 14 Nd2 fxe5 15 fxe5 Ndxe5. No, that line isn't ideal play from either side, but does serve illustrative purpose.12 ... Nxd3 (36)
Now or never! (One of my grandmothers had another expression suitable for the occasion, but not suitable for a family friendly blog.)
13 Qxd3
After 13 Qxd3
13 ... a6 (35)For some reason I can't rightly explain, I became worried about Nb5-d6. This is nonsense, of course, 13...f6 or 13...Qd8 probably make more sense, but 13...a6 does provide an eventual anchor for my light-squared bishop. (For what it's worth, Stockfish preferred five moves over 13...a6, though just barely: 13...Qa6, 13...Nb8, 13...Kh8, 13...Qd8 and 13...f6, with the first being for choice. Frankly, 13...Qa6 just looks wrong to me! But the moves are all basically even, evaluation-wise.)
14 Rac1 (32)
After 13...a6 I didn't really understand this move. Seizing the only open file makes basic positional sense, but he won't be able to do anything with it for several moves, and Black is many moves away from being able to seize the file himself, as I need to move my bishop on c8 first, which really means moving my knight on d7 before that, and that isn't easy. I would think that if White wants to play on the king-side, which his next move implies, now is the time to do it. Perhaps he was planning Nc3-a4-c5?
All that kvetching from an 1800 aside, Stockfish evaluates the move as basically as good as any other move for White, including including Rfc1! It does give an ever-so-slight preference to 14 f4, however, so I can salvage some ego from this minor debacle of a note....
14 ... Qd8 (32)
With White's queen on d3, a protected pawn on b3, and his queen's rook on c1, Black's queen is no longer doing anything useful on b6 other than potentially presenting herself as a target. 14...f6 didn't seem right for some reason. I kept thinking of lines like 14...f5 15 Na4 Qd8 16 exf6 Nxf6 17 Nc5, with an eventual Nf4 forcing me to to exchange my dark-squared bishop for the c5 knight, and now Rac1 starts looking good. 17...b6 would stop that, of course, so strike this one up as an example of fear overriding reason.. Which doesn't make the move I played bad, of course, just that my reasoning wasn't completely sound.
(Actually, Stockfish tells me 14...f6 is inferior, though probably far from losing, for tactical reasons after 15 Nf4.
Analysis position after 15 Nf4
Now Black has to decide between 15...Ba3 and 15...Qd8. Lines such as 15...Qd8 16 Nxe6 Nxe5 17 dxe5 Bxe6 17 exf6 Bxf6 18 Rfd1 need to calculated, and the resulting positions accurately assessed. Frankly, I'm glad I skipped out on all of that!)
After 14...Qd8
15 g4 (30)When I showed this game to Paul Leggett after the round, he was impressed with this move. "It stops ...f5," he said. "Does it?" I asked in turn.
15 ... f5 (31)
After 15...f5
It's funny, my notes above only mention ...f6 as an option. But ...f5 is also a standard idea in such positions and had been in my thoughts. Frankly, if you aren't considering ...f5 in such positions, you shouldn't play the French Defense. But to me in the positions above, ...f5 hadn't seemed as important, or perhaps I should say 'critical', an idea as ...f6 once White's light-squared bishop came off the board. That's probably a failure of imagination on my part, I confess, and quite likely a misunderstanding of the positions. But mostly I had been looking to pry open the center a little bit with ...f6. But 15 g4 made me adjust my plans.
As you can see from my time records, I made this decision quickly, after only a minute's thought. My thinking was that 16 gxf5 Rxf5 left me stronger on the king-side, while 16 exf6e.p. Nxf6 17 g5 Ng4 would with let me grab the e3 bishop with the center starting to open up, or drop the g-pawn. And 17 h3 would just be sad. You'll see why my thinking was wrong in a minute, if you don't already.
(Stockfish, incidentally, didn't think much of 15 g4, it not even being in it's top twenty choices after chugging for about 15 minutes. but it didn't think it much worse than it's twentieth choice, though it does slightly favor Black (=+) after that move. It really doesn't like my reply, though, barely putting that in its top twenty choices. It prefers 15...Ba3 16 Rc2 Nb8 17 Na4 Bd7 18 Qd2 and so forth. My move drops the evaluation by a full point, and White is now almost to +/-.)
The game now becomes a knife fight in a small room.
16 gxf5 (18)
When he spent 40% of his remaining time on this move, I felt good!
16 ... Rxf5 (30)
17 Nxd5
Played instantly, as one might expect after the big think the prior move. I felt chagrined. I was also reminded of the other time a played a senior master, and missed something at the end of a long tactical sequence.
Still, I pulled myself together quickly and kept playing. First, as I have written on the club blog and expounded endlessly in person, I have come to believe in fighting to the bitter end. So fight I will. Second, as I was out-rated by 700 points, I had nothing to lose, right?
After 17 Nxd5
17 ... Ba3 (28)
Black has lost a critical center pawn. Worse, the d5 knight threatens to move to c7 and win the e6 pawn.I found my next move quickly, using something Dvoretsky calls "The Principle of Exclusion", if memory serves. At least, this is what I think he means when he writes that!Basically, I looked at moves and eliminated them as prospects once I saw something wrong with them, and played what was left. First, if I try reinforcing the e6 pawn with 17...Nf8 or 17...Nb6, White can take on e7 (or even b6 in the latter case) and be comfortably better. (White may have better, but that's unimportant. What is important is that he already has something good.) 17...Bg5 accomplishes nothing after 18 f4. 17...Rg5 is the kind of move that could only work in a speed chess game, and even then only if the other player were drunk or otherwise prone to panic.
So we come to 17...Nc5 and 17...Nxe5. Both White's d5 knight and queen are loose, but the moves given obviously fail. 17...Nb8 and 17...Bh4 don't even merit consideration.
That left 17...Ba3, which at least makes a threat while removing itself from immediate danger.
18 Nfd4 (17)
A move that reinforces the e2 knight, gives two layers of protection to the knight that moved, and given the queen some protection. It also threatens to take the e6 pawn, eliminating the defender of the f5 rook and collapsing Black's defense.
After 18 Ndf4
18 ... Nb6 (18)
Best, I thought at the time, and I believe the computers agree! (When defending, the best move isn't always the best move, if you know what I'm saying. And if you don't, never mind.)White is threatening 19 Nxe6 when Black's position collapses. 18...Bxc1 19 Nxe6 Qe7 (or whatever) 20 Qxf5 Bxe3 21 fxe3 Nf6 22 exf6 Qxe6 23 Qxe6 Bxe6 24 fxg7 is winning for White. Both sides can play better, I'm certain, but that seems representative of the best Black can hope for after 18...Bxc1. 18...Qg5+ seems to do less than zero after 19 Kh1. But 18...Nb6 lends support to the d5 square, lets the c8 bishop defend e6 and dream of escape, and keeps White's queen and rook from occupying c4. (Not that I can see a reason they would want to now, but this position is complicated.)I also thought 18...Rxf4 gets nowhere. Upon further reflection, that's correct. White's worst case scenario is 19 Nxf4 Bxc1 20 Rxc1 when he's up a sound pawn with a better position. That's enough to stop thinking about it.
I spent a long time on this move, 35% of my remaining time. At the end I was happy because 18...Nb6 eliminated a certain amount of disharmony in my position.
(NOTE: After a long think Stockfish prefers both 18...Bxc1, and especially 18...Qg5+ to my move. The lines appear to me to be even more complex than what happened in the game, however. For example, not my "whatever 19...Qe7" in the 18...Bxc1 line above, but 19 Nc5 when it's a fight. A complicated 2N vs R+B ending arises, when White has a pair of extra pawns as compensation for the exchange. The 18...Qg5+ are even more obscure. Silicon & carefully sorted electrons be damned, but I prefer my own move, despite the slightly better evaluations of the other two moves! The interested reader should explore them himself.)
19 Rcd1 (15)
A tough decision to make. This abandons the file, but that's a minor consideration - the pieces are where it's at. This move adds support to the queen and adds more force down the d-file. Stockfish seems to slightly prefer 19 Rc2, keeping the file and, more importantly, reinforcing the e2 knight. Don't laugh, that last is important!
Here I felt he had the advantage, but that I had enough fight to make it interesting.
19 ... Bd7 (16)
I considered 19...Qg5+ first, but didn't like it - yet. For example, 20 Ng2 left me with a decision about where to place my queen - 20...Qg4 21 f3 didn't look inviting! (Yes, I missed 21...Rxf3, but after 22 Rxf3 Qxf3 23 Rf1 Qg4 24 Rf4 is doesn't look appetizing, though objectively Black is fine. But 20 Kh1 leaves White much better. But that's part of the problem, White has three responses to evaluate.) So I decided to save the check and increase the activity of my pieces. I saw that a bishop on c6 could make life unpleasant for White's king and a bishop on b5 might skewer three pieces at once. I realized that the knight on e2 would only be defended by the queen in some lines, and would hence be a little on the loose side.
I rejected 19...Nd5 out of hand as the exchange on d5 only seemed to help him. Yes, that was my actual thought process! His pieces are kind of tripping over themselves, and that seemed to benefit me. That said, 19...Nd5 is probably objectively best - if you're a computer, playing against a computer. Subjectively, I thought I was farther behind than I was, and I wanted to keep the game complicated. It does no good worrying about the other player being better than you are, even when (as here) they are. You've got to try to make and take your chances.
Now seems a good time for another diagram.
After 19...Be7
20 d5 (8:57)Here it comes! During the game this seemed very scary, but I also felt my last three moves (...Ba3, ...Nb6, and ...Be7) had done more for my position than his previous two moves (Ndf4 & Rcd1) had done for his - so I couldn't be worse off than I had been before!
Objectively speaking, White's last move gave away all his advantage and we're dead level again. 20 Qe4 and 20 Ng3 were both better.
20 ... Bb5 (15)
This first, to weaken the e2 square.
21 Qe4 (8:25)
21 Qd4 Bxe2 22 Nxe2 (22 Qxb6 Qxb6 23 Bxb6 Bxf1 and Black is up a rook) 22...Nxd5 and Black is better, though I'm not sure how much. 23 Nf4 Qg5+ solves all my problems on the d-file. (I'm only a little better, it turns out, but it looks easier to play for Black at that point. White could also back up with 21 Qd2, but it's hard to imagine a human player backing down at this point, especially when he out-rates the opponent by 700 points!)
21 ... Qg5+ (14)
Now!
I felt very good about my position!
After 21...Qg5+
22 Kh1 (5:26)In fact, now I'm better. (Stockfish gives Black an advantage of about a pawn.) 22 Ng3 was better, though very complicated. The position would then be even. 22...Rxe5 23 Qd4 Bxf1 24 Qxb6 might follow.
Analysis position after 24 Qxb6
Fun! 23...Rxe3, and many other moves at several points, also merit consideration. I love chess like this!
After 22 Kh1
22 ... Rxe5 (13)
I took a minute here to make certain I wasn't missing anything obvious.Here I was so nervous I got up and took a walk for a few minutes to try and settle my nerves. I knew I was close! But the walk didn't help.
23 Qd4?? (4:19)
This is a blunder, though not obvious, and should have cost him the game. Take some time to ponder the position, look for the winning combination. 23 Qf3 was best, though Black is still significantly better.
After 23 Qd4??
23 ... Nxd5 (7:25)
The correct, and obvious, first move.24 Nxd5 (3:28)
24 Rg1 was the alternative, but now the weakness on the a8-h1 file decides the game. 24...Bc6 25 Rxg5 Nxf4+ 26 Kg1 Nxe2+ 27 Kf1 Nxd4 28 Rxe5, and yes, Black is winning, though I'm not sure I can prove it against a 2473-rated player!
Analysis position after 28 Rxe5
This position will be a training position for me to play out against Stockfish.
In the position below, can you see the winning move for Black now?
After 24 Nxd5
I've dropped plenty of hints along the way in my analysis, and that's by design. In a real game, we don't come upon positions fresh, with only the knowledge that someone has the move and should win. We build up a certain understanding and feel for the positions along the way. Sometimes those feelings get in the way, as conditions change, but sometimes they don't. I've spent a few hours writing my notes to this game, both in an effort to understand what I did correctly and incorrectly (both in a strictly chess sense, and in the sense of how I thought about and acted on thoughts during the game), but also to try to give a flavor of what it was like.
Here I will give one more hint, of a general nature. In positions in which the other side is making multiple threats (here to win a piece if I don't recapture on d5 or elsewhere, or to win my queen if I do), it is necessary to respond with moves that make at least as many threats as the other side is making, or to make the ultimate threat (mate).
If you don't have the answer, please stop and think about it for a few minutes. I spent 15 seconds on this move, which was not enough. Most of those seconds were spent on sitting back down, writing down the move, and settling myself before moving. I've always been impulsive at the board, and here is the best example ever. If I had checked my calculations, I would have seen that my intended line wasn't good enough, and would have looked for something better. If you feel your position is good, there should be something better! But I failed to look,, and thus failed to convert the biggest upset of my chess career, and arguably ruined the best game I've ever played.
24 ... Rxd5?? (7:15)
Ruined! At move 27 I looked at the board and said, 'How'd that happen?" I didn't realize until then that I was losing a piece! Later, showing the game to Garry Day, he first played as I did and also wondered how he ended up down a piece. It's easy to see now, at the analysis board, but I missed it at the board in my haste. Paul didn't see the winning move either when we looked at the game after the round, but Garry did find the winning move with a second chance. I didn't realize what I had missed until I ran through Stockfish once - and then I kicked myself. I'm going to put the rest of this note in white font. To see it highlight the space below with your cursor and the text will appear.
24...Bxe2 simply wins. That's right, ignore the captured knight on d5, ignore the threat to my queen, and make some threats of my own! After all, it's not the pieces that come off that matter, but the ones that stay on. 24...Bxe2 does four things at once. First, it gets the piece back. Second and third, it threatens to capture the rooks on d1 and f1. And finally, it threatens mate! 24...Bxe2 25 Bxg5 Bf3+ 26 Kg1 Rxg5+ 27 Qg4 Rxg4 mate. Not bad, mating a senior master in 27 moves as Black. Alas and alack!
The most likely variation, though, 24...Bxe2 25 f4 Rxd5 26 fxg5 Rxd4 27 Bxd4 Rd8! (not 27...Bxf1 28 Rxf1 when Black has a lot of work left), and now Black will emerge up a whole bishop. Yes, I do believe I would have found all this if I had looked. For one thing, you don't have to find 27...Rd8 until it's the 27th move! For another, and again, once one realizes that 24...Rxd5 fails, it's easy to look for alternatives.
But what really chaps my posterior is that I had seen the move earlier! The whole point of 19...Bd7 and 20...Bb5 was to exploit exactly this kind of possibility! And I had seen the move when I played 23...Nxd5. GAH!
Regardless, that's how I let the second strongest player I ever face in a tournament off the hook. The rest of the game doesn't really merit comment.
25 Bxg5 Rxd4
26 Nxd4 Bxf1
27 Rxf1 e5
28 Nc2 Bb2
29 Rad1 Rc8
30 Rd8+
and here even I thought resigning was warranted. 1-0
Nonetheless, I still take some pride in this game. I did play well, save for one awful blunder, and it's good to know that I wasn't the first one to lose the game. But I was the last person to lose it! I look forward to playing more games like this.
No comments:
Post a Comment