Not All Equals are Equal

by admin on July 29, 2020

Here’s a position I reached in a recent rapid game against Shredder. I’m White, and it’s my move. The computer is set to a rating of 2321. One of the nice things about Shredder is that you can stop the clock even during a “rated game,” and I like to do this at most once per game as a training technique. (I’ve written a lot of blog posts about this before; Google “matrix chess.”)

Position after 44. … Ke8. White to move.

FEN: 4k3/2n5/6R1/pp1p1P2/3P1q2/1P1B1N2/P3K3/8 w – – 0 45

What do you think about this position? Who stands better? What is White’s plan? Before I stopped the clock, I had four candidate moves: 45. f6, 45. Rc6, 45. Rb6, and 45. Ne5. Which of these would you choose?

If you’re Fritz 17, this is a trick question. One move is clearly worse for White, but every other move scores at 0.00. All three of the options I gave you, plus many more moves (45. a3, 45. a4, 45. b4, 45. Rg7) all get the “0.00” treatment. So if you’re a computer, you just pick one at random. For human players, however, I think that one of these moves really stands out. Not all “equal” moves are equal.

Let’s start with the Mike Splane Question: How am I going to win this game? The answer is, most likely by pushing the f-pawn, either queening it or forcing Black to make concessions (like the two queenside pawns) to stop it. So the move 45. f6 is a sensible one to look at, but it lets Black activate his knight too easily: 45. f6? Ne6, when the d-pawn falls and White’s position crumbles.

45. Ne5 is a very interesting idea. Of course, the d-pawn is poisoned because of 45. … Qxd4?? 46. Rg8+ Ke7 47. Nc6+. However, I was not able to find a convincing way to escape Black’s checks after 45. … Qh2+. Running to the queenside is a good way to lose: 46. Ke3 Qh3+ 47. Kd2 Qh2+ 48. Kc3?? b4 mate! For a while I thought that White could stop the checks by bringing his king to f2 and then interposing his rook on g2. Thus, for example, 45. Ne5 Qh2+ 46. Kf3 Qh3+ 47. Kf2 Qh2+ 48. Rg2 Qf4+ 49. Ke2, and Black is out of checks. But if we compare this to our original position, has White really made progress? I don’t think so. He’s improved one piece (the knight) but “de-improved” another (the rook) and, most importantly, White has not gotten any closer to the proposed winning plan of f6. I decided to play this variation only if I could not find anything better.

If you like Nimzovich-style prophylactic chess, then 45. Rc6 is the move for you. Black’s knight has no moves and his queen has almost no moves. Unfortunately, the king does have moves, and after 45. … Kd7 the rook is just going to have to move again. Alternatively, I could play 46. Ne5+, but again I had the feeling this move was premature. White’s king has even less shelter from the checks than it did before.

So 45. Rb6! was my last best hope, and it’s a pretty darn good hope. Here are all the great things about this move:

  1. It puts the rook on the “good” side of the f-pawn, where it prevents … Ne6.
  2. Therefore it threatens 46. f6, which will be White’s response on most “do-nothing” moves by Black.
  3. It also threatens the b-pawn. This is not meaningless; if I can win both the a- and b-pawn, even if it costs me the f-pawn, I will have chances to win the game.
  4. It also gives my rook access to the back rank, which is very important because it takes me just two moves (f6 and Ne5) to get mate threats. Those threats are particularly important in line (A) below.
  5. It’s also worth remembering something I said in another recent post: When you can tie down your opponent’s stronger piece with your weaker piece, then you should try to attack on the part of the board his stronger piece can’t get to. Here, White’s knight and bishop are doing a nice job of immobilizing Black’s queen. So White should try to get the most he can out of his force advantage (R versus N) on the queenside.

So what could possibly go wrong after 45. Rb6? Well, one thing we have to look at is the possibility of active defense for Black by

(A) 45. … Qc1. Although I could play 46. Bxb5+ here, I think that the more incisive way to go is to let Black’s queen eat all the pawns it wants while I set up a checkmate: 46. f6! Qb2+ 47. Kf1! Qa1+ (… Qc1+ also runs out of checks) 48. Kg2 Qxa2+ 49. Kg3. After 49. … Qxb3 or any other Black moves, the mate threats after 50. Ne5 are too strong.

(B) Another really cool and really thematic variation occurs after 45. … Kf7. On the surface, a very logical move because it intends to meet 46. f6 with … Ne6. Now, however, the time is right for White’s knight to leap into the fray with 46. Ne5+! Suddenly Black’s queen is going to be the target of a blizzard of forking threats. The king has to go to the g-file, because either 46. … Kf8 or 46. … Ke7 would be met by 47. Ng6+, and 46. … Ke8 47. Rb8+ once again forces Black’s king into a fork. (Again we see the usefulness of being able to check on the back rank!) After either 46. … Kg7 or 46. … Kg8 White plays 47. Rg6+. I’ll leave you to work out why both moves to the h-file lose for Black. The only other possibility is 47. … Kf8, which is met by 48. Rf6+! This move was somehow hard for me to see, stepping in front of my own pawn, but the tactics all work. If Black’s king goes to g8 or e8, then White will play 49. Rf8+!!, sacrificing the rook to set up Ng6+. So 48. … Kg7 is the only move to avoid all the forks. But now White plays 49. Rf7+ Kh6 50. Rxc7! This cold-blooded win of a piece also wins the game. Black no longer has a perpetual check because my king will run to either b2 or b1 and then I will interpose the rook on c2 or c1! What a fantastic, unbelievable line!

(C) After finding forced wins in variations (A) and (B), which seemed like Black’s two most thematic responses, I was really pumped. However, I then came to the move that is really best for Black: 45. Rb6 Kf8! This modest move vacates the e8 square for the knight, which is now prepared to come into the game via e8, f6, and then either e4 or g4 (the latter being especially dangerous for White). The best answer I could find was 46. f6 Ne8! 47. f7! The f-pawn is dead anyhow, so I might as well create some trouble with it. Now 47. … Kxf7? loses to 48. Ne5+. I thought also that 47. … Nf6 48. Ne5 looked very strong for White, because it looks as if White escapes the checks after 48. … Qh2+ 49. Kf3 Qh3+ 50. Kf4, marching right up the board. Unfortunately, my analysis was flawed: Fritz comes up with 50. … Nd7, which supposedly saves the draw (although against a human I would probably play on). Finally, on 47. … Qxf7 48. Rxb5 Black’s queen cannot defend both the a-pawn and the d-pawn, and I thought that this endgame too would give me some winning chances. (Fritz again disagrees and says it’s a draw.)

Bottom line: 45. Rb6! has a clear winning plan (two of them, even), and it leads to a forced win if Black plays either of his two most obvious moves, and even on the computer-like defense 45. … Kf8, White may still have winning chances. It’s by far White’s best chance to win.

Now let’s see what happened!

45. Rb6! Kf8!

Of course, the computer plays the computer-like defense.

46. f6 b4??

My jaw dropped. Of course a human would play this move, because humans would worry about losing the b-pawn and a-pawn. But after my lengthy timeout analysis I knew that 46. … Ne8 was Black’s only hope. Since my opponent is a computer, I can only conclude that either Shredder evaluated the endgame in variation (C) differently from Fritz, or it didn’t play the best move because it was “dumbed down” (but only slightly) to 2321.

Either way, I make no apologies. My play gave Black the most chances to go wrong. A human probably would have gone wrong, and even the computer went wrong. Black is now a tempo too slow to defend.

47. Bg6 Ne8 48. f7 Nf6 49. Ne5 Qh2+ 50. Kf3 Qh1+

Or 50. … Qh3+ 51. Kf4, which is exactly like the line in (C) above except White is a tempo ahead and wins after 51. … Nd7 52. Nxd7+ Qxd7 53. Rb8+.

51. Kf4 Qh4+ 52. Kf5 Kg7 53. f8Q+! …

The triumph of White’s strategy.

53. … Kxf8 54. Rxf6+ …

At this point both computer programs say that White is easily winning. However, I was sweating bullets for the rest of the game. Endgames with rook, knight, and bishop versus queen are not exactly an everyday occurrence, and I was scared to death that I would blow it. Fortunately, the winning plan is not very subtle: surround and ingest the d5-pawn with my minions, and then push the d-pawn to victory. I’ll give the rest of the moves with no comments.

54. … Kg8 55. Bf7+ Kh7 56. Ng4 Kg7 57. Be8 Qh3 58. Rg6+ Kf8 59. Ke6! …

Okay, one little comment: Never forget about the power of checkmate threats, even in the endgame. They often allow you to get away with “impossible” moves, as in this position where White is able to leave his bishop on e8 unprotected.

59. … Qh7 60. Nh6 Qe7+ 61. Kxd5 Qe2 62. Rg8+ Ke7 63. Nf5+ Kf6 64. Rf8+ Kg5 65. Ng3 Qxa2 66. Ne4+ Kh6 67. Ba4 Qc2 68. Ke5 Qc7+ 69. Kd5 Qe7 70. Re8 Qc7 71. Nc5 Qf7+ 72. Re6+ Kg5 73. Kc6 Qa7 74. Re8 Kg4 75. d5 Qf7 76. Re4+ Kf5 77. d6 Qg6 78. Re6 Qg2+ 79. Kc7 Qd5 80. Kb6 Kf4 81. d7 Kg4 82. Rc6 Kf5 83. Rc8 Black resigns

After 83. … Qd6+ 84. Bc6 Black has no more checks and no way to stop the promotion of the d-pawn. I like the way that all of White’s pieces are working together.


  1. The Mike Splane Question is a great way to orient yourself in any position. Fortunately, computers do not know about this question. Many humans don’t know about it either.
  2. When you can immobilize a strong piece with weaker pieces, you should be able to use your force advantage on a part of the board the strong piece can’t get to.
  3. Checkmate is a surprisingly strong weapon in the endgame — usually not as an end in itself, but as a threat that enables you to make “impossible moves.” See move 45, variation (A), and move 59.
  4. Timing is everything in the endgame, and there are billions of positions where a tempo is worth more than a pawn. See Black’s move 46, which saved a pawn but lost the game.
Print Friendly, PDF & Email


First Live Chess in Four Months!

by admin on July 26, 2020

Yesterday I played my first live, in-person chess game since the coronavirus pandemic started, in early March. To be precise, March 3 was the last date when the chess club at Aptos Library met. Most likely I played one of the kids that day, although I don’t specifically remember a game. The next week, March 10, I cancelled the club meeting, after my assistant (Shan Crockett) said that he wasn’t going to come. I haven’t sat across the board from a person since then — only a computer.

However, I’ve been meeting with two of my favorite students, Emmy and Ryder, via Zoom for the last three weeks. They recently celebrated their 12th and 9th birthdays, respectively. Hooray for them!

Their parents, Scott and Shelley, mentioned that they had a plum tree that was producing lots of fruit, and I mentioned that I had tomato plants that were practically bursting, and so the idea of a plums-for-tomatoes trade was hatched. Here was my contribution to the swap.

Where’s Waldo?

We decided also to try a chess lesson, with safe social distancing. Scott and Shelley set up a table outside, with two large boards at opposite ends of the table so that I could use one board and Emmy and Ryder could use the other. (In fact there were several boards at the table. They have lots of chess boards.)

First, we played a consultation game (me against Emmy and Ryder and a little bit of contribution from Scott), which led into a pretty good lesson. Then Ryder wanted to play 5-minute chess, and that’s when things got interesting!

Black is ahead by one chocolate-chip cookie.

I had actually seen a photograph on Facebook of people playing in a chess club in Europe with exactly this setup — two boards and a clock in the middle. We put the clock roughly where the chocolate chip cookie plate is in this photo. I’ll call this setup Coronavirus Chess. When you make your move, you have to tell your opponent what it is (although in practice, it was just as easy to see the move).

My first game of Coronavirus Chess was not exactly a thing of beauty. It took longer than I expected to go through this procedure of making the move, saying the move aloud, reaching halfway across the table and pressing the clock button. No matter how fast I tried to play, my time was just melting away. Ryder and I both got in time pressure. Then we found out an even bigger problem with Coronavirus Chess.

Position after 39. Ke3. Black to move.

FEN: 8/1p2k2p/1p3p2/1P2p3/8/3bK2P/8/8 b – – 0 39

Ryder is playing White and I’m playing Black. Although I’m obviously winning on the board, the clocks are another matter: I was down to 8 seconds. Ryder had about the same amount of time. I played 39. … Bxb5. But somehow or other, Ryder didn’t make the move on his board, and so he (thinking my bishop was still on d3) played “40. Kxd3.” I tried to explain that it was an illegal move, but I saw that I was down to 4 seconds and so I played “40. … Bxd3.” Now we have two different positions on two different boards, one with his king on d3 and the other with my bishop on d3!

Now as we’re trying to figure out who is on d3, and I’m trying to explain at the same time that in speed chess (unlike tournament chess) it’s legal to take the king, my flag falls. So, who wins? Me, because I took his king? Ryder, because my flag fell? And what exactly do the Rules of Chess say to do when the two players have different positions on their boards?

Well, it’s all beyond me. This was just a friendly game, after all. I’d rather focus on how well Ryder played. He didn’t make any really serious mistakes in this game until move 33, when his time pressure was already pretty severe. I think he played better than me for quite a while. Keep in mind that his rating is under 1000 and mine is close to 2200!

Ryder – Dana: 1. e4 d5 2. Nf3?! … (Ryder is used to playing against Emmy, who always plays 1. … e5, and I don’t think he even looked at my move before touching his knight. Actually, 2. Nf3 isn’t all that bad, and he follows it up very well!)

2. … de 3. Ng5 Qd5 4. d3! … (Forced, but good.)

4. … ed 5. Nc3! … (Very nice. Gaining tempi, not worrying too much about material.)

5. … Qe5+ 6. Be3 e6 7. Nf3 Qa5 8. Bxd3 Bb4 9. Bd2! Nf6 10. a3! … (He’s playing exactly the moves I would play as White. By now White has pretty good compensation for the pawn.)

10. … Bxc3+ 11. Bxc3 Qb6 12. O-O Nbd7 13. Qe2 O-O 14. h3 Nd5 15. Qd2 … (Probably 15. Bd2 is better, recognizing that the pawn on b2 is poisoned. Fritz says 15. Bd2 Qxb2 16. c4! Nc3! 17. Bxh7+ recovering the pawn, with a slight advantage to White.)

15. … Nxc3 16. Qxc3 Nc5 17. b4?! … (Not only does Ryder not avoid the trade on d3, he actually provokes it. To me this was a little bit of a “beginner mistake,” but it led me to relax and make a much worse blunder of my own.)

17. … Nxd3 18. Qxd3 c5? (A really bad move on my part. After 18. … f6 preparing … e5, Black has a healthy extra pawn.)

Position after 18. … c5. White to move.

FEN: r1b2rk1/pp3ppp/1q2p3/2p5/1P6/P2Q1N1P/2P2PP1/R4RK1 w – – 0 19

This is a somewhat ironic position, where the “beginner move” for White might be better than the “sophisticated move.” Fritz points out that 19. Ng5! is quite effective. Either 19. … g6 or 19. … f5 would give me huge weaknesses. For example, after 19. Ng5 g6 20. Qc3 cb 21. ab f6 22. Ne4 e5 23. Rfd1, White has the very inconvenient threat of Rd6. Fritz says that it’s completely equal.

Instead Ryder plays a more sophisticated move, which is exactly what I would have played. I will give it an exclamation mark even though the computer disagrees.

19. Qe3! cb 20. Qxb6 ab 21. ab Bd7 22. Rxa8 Rxa8 23. Rd1 Bc6 24. Ne5! … (Several times in this game I had the experience of thinking, “It would be annoying if Ryder played X,” and then he immediately played X. It happened on moves 5, 9, 10, 19, and now.)

24. … Kf8

This is another move where I have to disagree with the computer. Fritz recommends the very ugly 24. … b5. Maybe in some ideal computer sense it’s right, but I actually think that 24. … Kf8! was an important decision and a key turning point. There are times in a chess game when you have to say, “Okay, this is not working out,” and hit the re-set button. That’s what I’m doing here. I’m going to improve my king position so I don’t get mated, and then I am going to try to win whichever endgame I end up in.

Position after 24. … Kf8. White to move.

FEN: r4k2/1p3ppp/1pb1p3/4N3/1P6/7P/2P2PP1/3R2K1 w – – 0 25

White’s decision here is pretty tricky. I think that Ryder’s choice was not the best, but very reasonable.

For White, the best shot at fully equalizing is 25. Nd7+! Fritz thinks (and I think) that White is doing fine after either 25. … Ke7 (which is what I probably would have played) or after 25. … Bxd7.

Many people might be tempted to trade pieces the other way: 25. Nxc6 bc. After 26. Rd7 Ra1+ 27. Kh2 Rc1 28. Rd2, Black has the better of the rook endgame, as his rook is active while White’s is passive. More interesting would be 26. Rd6! Rc8 27. Rd7. The point of this two-step is to make Black put his rook in a more passive location. But that would be a pretty sophisticated idea for an under-1000 player to find!

Interestingly, Ryder chose not to trade pieces, which is counter to his usual tendency. I think he must have felt, intuitively, that his knight was a better piece than my bishop, so he didn’t want to give it up. I don’t want to criticize his next move too harshly. First, I think he has been too prone to mindless exchanges in the past, and I’m glad to see him thinking before taking. Second, he has a concrete plan. I like to see planful play from my students. It’s just not the optimal plan for this position.

25. c4?! Ke7 26. b5 Be8

To Ryder, this move probably looked like a retreat. But actually, Black’s position is getting more compact and more solid. The bishop will re-emerge on f7, and it will be a much better piece than White’s knight. Black’s extra pawn will finally start to make a difference after … f6 and … e5.

27. g4 f6 28. Nf3 e5 (For the the first time all game I really feel comfortable with my position.)

29. Re1 Bf7 30. g5 Bxc4 31. gf+ gf 32. Rc1 Bd5 33. Rd1? … (Until now, Ryder has played 32 moves that have not had any serious faults. He’s made mistakes, but they were subtle ones: misjudging the value of certain trades, coming up with plans that were good but not the best. Now, with both players under 30 seconds for the rest of the game, his haste finally gets the better of him. But all in all he played very well.)

33. … Bxf3 34. Rd3 Rg8+ 35. Kh2 Rg2+ 36. Kh1 Rxf2 37. Kg1 Be2 38. Kxf2 Bxd3 39. Ke3 Bxb5, Game ends in confusion.

And now we have arrived at the position we started with. Ryder played “40. Kxd3” and I played “40. … Bxd3” and then my flag fell while we were trying to figure out what was going on. Oh well.

One lesson I’ve learned from this is that if you’re going to play Coronavirus Chess, you have to do it with a time delay. Next time I’ll ask for at least a three-second delay.

In spite of this strange interlude, I enjoyed being outside, talking with the kids and their parents, breathing in the fresh air, and playing some version of chess. I think you don’t appreciate what you have until you lose it for a while.

Print Friendly, PDF & Email

{ 1 comment }


July 19, 2020

For every good, instructive game that I play against the computer, there must be two or three really awful games. Here’s one that I played this morning that is so bad that it’s good. I’m playing White, Fritz (with its rating set to 2025) is playing Black, clocks are set at 10 minutes for 40 […]

Print Friendly, PDF & Email
Read the full article →

Winning the Queenside Battle, Winning the Kingside War

July 12, 2020

Have you ever had a game that you won and you didn’t understand why? I call it a Billy Preston game, in honor of the singer who had a #1 hit in the 1970s called “Will It Go Round in Circles.” The lyrics go: “I’ve got a story ain’t got no moral / Let the […]

Print Friendly, PDF & Email
Read the full article →

The Audacity of Hope, 2020 Version

July 5, 2020

Going off-topic today, with apologies to my chess fans. I. Last night Kay and I had movie night, a family tradition every Friday. Sad to say, I can’t tell you what movie we watched last week, or the week before that, but I can tell you that last night we watched Hamilton and it was […]

Print Friendly, PDF & Email
Read the full article →

Back in Business!

June 30, 2020

My computer issues have been taken care of, thanks to one of my readers, Marshall Polaris. Marshall is a Kolty Club member, a fan of this blog, and was also a participant on the Kolty Club team in the U.S. Amateur Team West tournament back in February. (Back in the long-ago days when you could […]

Print Friendly, PDF & Email
Read the full article →

A Weird and Wonderful Game

June 18, 2020

Just when I think I’m wasting my time playing so many games against the computer, along comes a game like this one! I’ll show you the end first. FEN: 2kr4/ppp5/6p1/2Q3P1/4PP2/5KN1/PPp4q/R4B1r b – – 0 23 I’m playing Black against Fritz, with its rating set at 2025. What would you do? If you want a little […]

Print Friendly, PDF & Email
Read the full article →

Computer Issues (Alas)

June 16, 2020

I just heard from Mike Splane that there was a comment on my last post that did not display when he clicked on the link. Here is what the comment said, from Larry Smith: “A great post! I can only imagine how much time went into putting this together. Here are several comments: First, when […]

Print Friendly, PDF & Email
Read the full article →

A Notorious Endgame

June 14, 2020

In the past I’ve mentioned the “Four Endgames of the Apocalypse,” the four supposedly “elementary” yet ferociously difficult endgames that even grandmasters screw up sometimes: K+Q versus K+R; K+B+N versus K; K+R+B versus K+R; and K+Q+RP versus K+Q. I’ve encountered each of these in my chess career, some of them more than once. But there […]

Print Friendly, PDF & Email
Read the full article →


May 31, 2020

There will be a short-lived pause (which you won’t even notice) in this blog as I de-chessify. The reason is that I’ve gotten entirely too much sucked into playing against Fritz on my computer. My wife finally said enough is enough, and took away my laptop for a week! Let me hasten to add that […]

Print Friendly, PDF & Email
Read the full article →