50 Years of Chess: Year 39

In 2010, according to the USCF online database, my rating hit its lowest point ever. At the 2010 CalChess Labor Day Classic (the northern California state championship) I withdrew after four rounds with a score of 0.5-3.5, culminating in a loss against a player rated more than 200 points below me. This came after a disastrous result in the U.S. Open in August 2010, and a semi-disastrous result in the Chicago Open in June, and it dropped my USCF rating to 2075.

Of course my real lowest rating came years and years earlier, way back in the Fischer boom era. My first rating was 1226, and that was also my lowest. But that’s a little bit like saying that the lowest floor you’ve been on in a skyscraper is the first floor. Well, duh. That’s where you entered the building. My 2075 low is more meaningful and interesting because it was the end of a gradual multi-year decline and the beginning of a gradual multi-year improvement.

As I’ve already said on multiple occasions, I am convinced that the main reason for the turnaround was that I started going to Mike Splane’s chess parties. The first one I went to was in August of 2011. They didn’t make an overnight difference. At the end of that year my rating was still only 2089, not a big improvement. But over time, there was definitely an upward trend, and I think it’s because the parties showed me how to think about strategy and how to be more aware of “dynamic imbalances” (to use Jesse Kraai’s term).

This is not to say I don’t still have some things to work on. I’m sure we will see that when I return to tournament chess later this year. Some of my Facebook friends are starting to play again (Michael Aigner, Mike Zaloznyy) and they’re finding it hard to get back to their old level overnight. One reason is that the young players have continued to play online and improve during the pandemic, but their ratings have not changed because of the year-long hiatus in rated over-the-board tournaments. So there are lots of underrated youngsters looking to steal some rating points. We’ll see some weirdness as we get back to over-the-board play in 2021.

But I digress! Let’s go back to 2010. Here’s a game that really epitomizes why my results were so bad. My approach to chess at that time was something like this: sprinkle gasoline in a circle around me. Try to lure my opponent inside the circle. Then set it on fire and hope I can get out before he does.

Against 2400 players (and even against reasonably careful 2000 players) this “strategy” does not work. In the game below, International Master Dmitry Zilberstein pre-empts me by tossing the first match and watches as my position goes up in flames. This game was played in round one of the 2010 CalChess Labor Day Classic, the tournament where my rating hit its nadir.

Dana Mackenzie — Dmitry Zilberstein, 9/4/2010

1. d4 Nf6 2. c4 g6 3. Nc3 d5 4. h4 …

I had been looking for new ways to play against the Grunfeld, and this was an invention that I tried in this game for the first time. The idea, if Black continues developing normally with 4. … Bg7, is to play 5. h5!, taking advantage of the fact that Black needs his knight at f6 to defend the d-pawn. So he has a choice between defending it with 5. … c6 and having his bishop sent back to f8 after 6. h6, or wrecking his kingside pawn formation with 5. … gh.

Actually the move h4 has become quite popular, and has been played often by top-level players like Mamedyarov and Morozevich. However, they wait one move and play 4. Nf3 Bg7, then 5. h4. There’s a good reason for this: the immediate 4. h4 can be met by 4. … c5! 5. dc d4. By bringing the knight to f3 first, White prevents this variation.

If my opponent played 4. … c5, I was prepared to play 5. cd Nxd5 6. dc Nxc3 7. Qxd8+ Kxd8 8. bc Bg7 9. Kd2. This position is not new at all — it was played by Igor Zaitsev against Vassily Smyslov in 1963. And Zaitsev won the game! You might wonder why 4. h4 is so rarely played, when in its debut game White defeated a former World Champion. The reason, I think, is that the position looks so dull and unpromising. Zaitsev won the game in spite of the opening, not because of it.

Be that as it may, my opponent had different ideas. He played the Slav-like move … c6 instead of … c5.

4. … c6 5. Bg5 Bg7 6. e3 Be6

A saucy move, trying to force me to make a decision about my c-pawn. But this is potentially an awkward place for the bishop, where it can be harassed by a knight coming to f4.

7. Bxf6 …

Really there is a move here for every chess style. You like to sacrifice your pawns? Play 7. Nge2 cd 8. Nf4, aiming for the most harmonious development. You like to grab your opponent’s pawns while falling behind in development? Play 7. Qb3 and threaten to take on b7. You like more static, even positions? Play 7. Bxf6 and 8. cd. Usually the first would be my style, but I was a little bit cautious due to the 300-point rating difference.

7. … Bxf6 8. cd cd 9. Qb3 Qb6?!

Zilberstein thought a very long time on this move, and I don’t think he made the best decision.

FEN: rn2k2r/pp2pp1p/1q2bbp1/3p4/3P3P/1QN1P3/PP3PP1/R3KBNR w KQkq – 0 10

Up to now the game has gone quite well for me, but now I start going into self-destruct mode.

10. Bb5+? …

I hate this move. First, it shows that I’m psyching myself out. Against most players I would trade queens without hesitation. But somehow, because he’s an IM, I think he must have some reason for trading queens that I don’t see.

But let’s start with making a plan. What does White want to do in this position? That’s easy: he wants to play Ne2 and Nf4, pressuring the most sensitive spot in Black’s position — the pawn on d5 — and also eyeing the bishop on e6, which cannot leave because then d5 would fall.

Does 10. Bb5+ do anything to enhance this plan? No. In fact, it helps Black. If the bishop takes Black’s knight on c6, it allows the queen to take on c6, where it defends both d5 and b7. If the bishop doesn’t take Black’s knight, then it’s misplaced. It’s not doing anything useful.

By contrast, after 10. Qxb6 ab 11. Nge2 Na6 (or any other move) 12. Nf4 Nc7 13. Nxe6 fe 14. f4! we get to a position where White is totally in control. The key thing to observe is that Black’s bishop is permanently locked out of the queenside. Jesse Kraai would call it a “Wintered bishop,” after the famous game Winter-Capablanca where Capa imprisoned Winter’s bishop on the kingside, then switched his attack over the queenside where he had an extra piece. Presumably I could play a similar strategy here.

Now let’s discuss the queen trade. Does it help or hurt this strategy? I think it helps because it creates a weakness, the doubled b-pawns. With a weakness to attack, and an extra piece to attack with, I think that White would have really good winning chances. At minimum, he will have a very persistent advantage and be able to make Black suffer for a long time.

Finally, trading queens minimizes the risk — the possibility that Black may be able to confuse the issue with tactics. This was something that I completely missed, but my opponent did not. I don’t want to give away the surprise too soon, but let me just say you always have to be careful in positions where you are behind in development. If your opponent finds a way to open the position, you could be in big trouble.

So the best move here is 10. Qxb6. An alternative move, also good, is to play 10. Nge2 right away, keeping the possibility of a queen trade open but letting Black decide whether he wants to initiate it.

10. … Nc6 11. Ne2 …

The next three moves are where I lost the game — not because of the moves themselves, but how long I took to play them. I’ve already told you that White’s plan is to play Ne2-f4. Yet it took me 14 minutes to play my eleventh move, 18 minutes to play my twelfth, and 16 minutes to play my thirteenth! That’s 48 minutes for three moves, in a 30/90 time control. After move 10 I was actually comfortably ahead on time. I had 56 minutes left for 20 moves, while my opponent was down to 31 minutes. Three moves later, I had only 8 minutes left to play 17 moves. I can’t compete against an IM with that kind of time pressure.

The time consumption clearly speaks to my psychological state. I wasn’t sure what I wanted to do, and I did not trust the process. I didn’t even have a process.

11. … O-O 12. Nf4 …

FEN: r4rk1/pp2pp1p/1qn1bbp1/1B1p4/3P1N1P/1QN1P3/PP3PP1/R3K2R b KQ – 0 12

Now a psychological catastrophe occurs. I’ve already told you that I spent 32 minutes thinking about my last two moves. I have no idea what I was “thinking” about, though, because never once did I even consider the possibility that my opponent would play the move that he now plays.

12. … Bxd4!

An outstanding move, no doubt about it. Zilberstein set the board on fire before I was ready.

Now we can see why it was so bad for White to put the bishop on b5, and also why it was so bad to allow queens to remain on the board. If I take on d4 I run into a barrage of forks: 13. ed? Nxd4 (Fork #1: Q and B. Also threatening Fork #2: K and R.) 14. Qa4 a6 15. Be2 Qxb2 (Fork #3: R and N) and something’s gotta give.

This is the sort of situation where you have to remember Jesse’s saying that I’ve left at the top of this page for many years: “There will be mess-ups; that’s part of the game.” There is no question that White has messed up. But part of chess mastery is learning to recover from mess-ups.

Let’s take a look at this position. How bad is it, really? Black still has a weak pawn on d5 and several other targets: the pawn on e7, the bishop on e6, and of course the bishop on d4 that can’t be taken yet but which will certainly have to leave at some point, costing Black a tempo. In this position it was essential for me to regroup, control my emotions, and focus on playing sound and solid chess. Instead, what did I do? I thought for 16 more minutes, putting myself in desperate time trouble. I just threw more gasoline on the fire. The fact that I then played the right move is almost unimportant. I played the right move, but in a way that almost guaranteed I would mess up later.

13. Ncxd5 Qa5+ 14. Ke2? …

More gasoline on the fire! In a position like this, 14. Kf1 should be almost automatic. It does White absolutely no good to keep his king in the center of the board. I think I did this because I wanted to connect my rooks, but that is a much lower priority than king safety. In fact, this is one position where connecting the rooks really doesn’t matter much, because my king rook is probably best placed right where it is, on the h-file supporting the advance of the h-pawn.

I was probably calculating variations furiously and didn’t see any variation that would directly punish my Ke2 move. But you don’t need to calculate any variations to realize that it’s very hazardous to keep my king in the center.

14. … Bxd5 15. Nxd5 e6 16. Bxc6 bc 17. Ne7+ Kg7

FEN: r4r2/p3Npkp/2p1p1p1/q7/3b3P/1Q2P3/PP2KPP1/R6R w – – 0 18

There’s a time for calculating variations and a time for not calculating variations. This is definitely a time to buckle down and to some serious calculation. Both sides have desperado pieces, so things are possible that wouldn’t normally be possible.

The move 18. Nxc6 looks scary on general principles — White shouldn’t open lines with his king still in the center — but on concrete analysis it is actually pretty good. The fork on the Q and B virtually forces Black to play 18. … Qa6+ 19. Qd3. Then there are two possible paths for Black. If 19. … Qxd3+ 20. Kxd3 Bxb2 Black wins back his pawn, but after 21. Rab1 we’re getting close to an endgame where White’s king position will not be a liability but an advantage. Or if 19. … Qxc6 20. Qxd4+ Kg8 21. h5! White’s kingside attack, which has been moribund since move four, suddenly starts coming to life.

I’m not saying that my move was a mistake. But this was an opportunity to steer the game in the direction I want it to go. Especially when you’re low on time, it’s helpful if you can do more steering and less reacting.

18. ed Rfe8 19. Qc3 Qb5+ 20. Kd2? …

Compounding my earlier error of 14. Ke2. There was still time to head to the (relative) safety of the kingside with 20. Kf3.

20. … Rxe7 21. d5+ …

FEN: r7/p3rpkp/2p1p1p1/1q1P4/7P/2Q5/PP1K1PP1/R6R b – – 0 21

I greatly overrated the power of this move. It’s just one check, which doesn’t put Black in any serious danger, and it only opens lines against my own king.

21. … Kg8 22. dc Rc7 23. Rac1?! …

After sabotaging my own position for the purpose of connecting my rooks, why didn’t I at least take advantage of my connected rooks to play Rhc1? That way my rooks would remain connected after I retreat my king to e1. I guess the answer is that you can’t expect moves played in time pressure to make sense.

23. … Rd8+ 24. Ke1 …

All of Black’s pieces are working together. White’s pieces are separated and confused. There is no question any more: Black is completely winning.

24. … Rd6 25. Qf6 Rcxc6 26. Rxc6 Qxc6 27. Rh3 …

I feel sorry for this rook. The moment that it finally gets in the game, the game is over.

27. … Qc1+ 28. Ke2 Qc2+ 29. Kf3 Rd3+ 30. Kg4 …

My last move is an illegible scrawl on my scoresheet. I didn’t even write down what Zilberstein’s final move was. There are two likely candidates, and I think that the prettier one is

30. … Qc4+

Also good is 30. … h5+, but I like this version because it makes White build a mating net around his own king. Clearly 31. Qf4 Rd4 loses a queen, and 31. Kg5 Rd5+ walks into a checkmate, so White has to play 31. f4. Then 31. … h5+ 32. Kg5 Qc5+ 33. Kh6 Qf8+! is the beautiful point of the combination. I have to commend Zilberstein’s board sight, because if he doesn’t see … Qf8+ he doesn’t win. After this exquisite retreat/attack move, White has to play 34. Kg5 and then the quiet move 34. … Kh7! sets up an unstoppable mate with … Qh6.

Now that I was past the 30-move time control, I had all the time I wanted to work out the details of this combination, and when I was satisfied I stopped the clock.

31. White resigns

Lessons:

1. Time trouble always makes a bad position worse.
2. When something goes wrong, control your emotions first. Then look at the position with fresh eyes and calmly think about the position you have — not the one you used to have.
3. Don’t open lines with your king in the center, and don’t keep your king in the center if lines are about to open.
4. Don’t open lines when you are behind in development. If you are ahead in development, look for tactical ways to open lines. (Black’s move 12.)
5. You don’t need to calculate moves in every position. There are times to calculate and times not to calculate.

Extra Bonus of Interest to Nobody But Myself

Because 2010 was such a bad year for me rating-wise, I thought it might be interesting to make a table of my worst and best tournaments ever, purely on the basis of rating change. According to USCF I’ve played 146 rated tournaments since the beginning of their online database. Here are my top three flops and successes.

Worst Tournaments

1. -48 points (2221 –> 2173), 1994 Columbus City Championship
2. -47 points (2137 –> 2090), 2010 U.S. Open
3. -36 points (2173 –> 2137), 1994 Columbus Open

One striking thing here is that two of my three worst tournaments took place back-to-back in 1994. It’s surprising that I don’t remember that, but if you look back at Year 23 in this retrospective, you’ll see that a lot of other bad stuff was going on that year. Having two bad chess tournaments in a row was the least of my problems.

I would make an argument for the 2010 U.S. Open as my worst tournament ever. Even though I lost one more point at the 1994 Columbus City Championship, I was starting from a higher baseline.

Best Tournaments

1. +61 points (2086 –> 2147), 2009 Western States Open
2. +46 points (2113 –> 2159), 1999 Labor Day Chess Festival
3. +41 points (2164 –> 2205), 1995 Atlantic Class Championship
4. +41 points (2096 –> 2137), 2005 Far West Open

Very clear what was my best tournament since 1991! However, there were certainly some better ones before the USCF database went online. The two that I know about for sure are:

• +50 points (2187 –> 2237), 1989 Roosevelt Open. This was my second outright tournament victory, and I ordered the rating report ($5, in those pre-Internet days) from the USCF to keep as a souvenir.
• +169 points? (1669 –> 1838), 1976 Pennsylvania Championship. Back in the era when I was still young and improving, huge leaps were not so uncommon. I’ve given this a question mark because ratings were not updated after every tournament back then, so it’s possible that the 169-point leap was achieved in two or even three tournaments. However, I’m pretty sure that the Pennsylvania Championship accounted for at least 100 points. One other trivia item (which I may have mentioned before, sorry if I repeat myself) is that I completely leaped over the 1700’s; I never had a published rating between 1700 and 1799.