Visit from an old friend

A couple weeks ago I got an e-mail from an old friend of mine, Robin Cunningham. We used to play chess together in North Carolina, when he was in high school and I was teaching at Duke University. When we first met we were roughly even in strength, but of course because he was younger he improved faster than I did. His peak rating (which I just looked up) was 2448.

Robin hasn’t played very much chess in recent years, partly because he is married with three kids, and partly because he succumbed to the lure of poker. He made as much as $50,000 in a year playing poker, versus … well, a whole lot less playing chess. However, chess is intellectually a much richer game than poker, in my opinion. (Robin agrees.) So he told me that he is now interested in getting back into chess.

The last time I saw Robin was in 2006, when I lured him out of “chess retirement” to play for a team I organized for the Bay Area Chess League. This league only existed for two years, sad to say. I needed a master to play first board for my team (which was called “Eight is Enough,” because we ended up with a roster of eight players). Robin generously volunteered, even though he was completely out of practice. He said he realized that his role on the team was simply to occupy the first board and hopefully pick up a half point here and there, while the rest of us won the matches. And that was pretty much how it worked. We won the league championship in a playoff with a team called “Mike’s Maters.” The whole season came down to Robin’s game. We were leading the match, 2-1, but because we had finished behind Mike’s Maters in the regular season, we had to get a draw in Robin’s game. If he lost, the match would be tied, 2-2, and Mike’s Maters would win the league title on tiebreaks.

Robin had an endgame against Mike Pearson (the captain of Mike’s Maters) that was right on the razor’s edge between a draw and a loss. I would have given about 50-50 odds of either result. But one of Robin’s great skills in chess is saving positions where he is suffering. With all the pressure on him, he pulled out a miraculous draw. It was really one of the most exciting moments in my chess career, even though there was no money at stake. (The only monetary prize was a free entry into season three — which turned out to be irrelevant because season three never happened!)

So anyway, I was very glad to hear that Robin is getting back into chess. He wanted to play a couple of training games with me. As for me, I wanted to play a couple of training games at a game/60 time control, because I bombed out so badly last month at that time conrol at the Cabrillo quads. So it was a perfect opportunity for both of us.

We played our two training games at my house on Saturday. Here is a picture that Kay took as we were getting started:

Our first game was probably the better of the two games. One interesting thing about it is that my best move of the game and my worst move were the same move! Here is how that happened. We got to this position (Robin is White, I’m black) after 1. Nf3 Nf6 2. d4 e6 3. Bg5 c5 4. c3 cd 5. Nxd4?! d5 6. g3 Nbd7 7. Bg2 h6 8. Bxf6 Nxf6 9. Nd2 a5 10. N4b3 Bf5 11. O-O Qd7 12. Re1 …

White’s last move pretty much announces his intention to play e2-e4, and so the question is what should Black do about it? At this point Robin got up to go to the bathroom, and as he told me after the game, “I figured that you would look hard at 12. … O-O-O but you would play 12. … Rd8.” And that is exactly what happened! It seemed to me that Black has a comfortable, if very slight advantage — really, it’s almost as if Black is playing White. So there was no need to jeopardize it by subjecting my king to a queenside attack.

After 12. … Rd8! Robin played, of course, 13. e4 de 14. Nxe4?! Nxe4 15. Qxd7+ (Note that 15. Bxe4?? would hang a piece after 15. … Qxd1 16. Rxd1 Rxd1 17. Bxd1 Bxe4. However, Black couldn’t play this zwischenzug on the previous move, with 14. … Qxd1, because White hits him with the counter-zwischenzug 15. Nxf6+! Does anyone know how to say counter-zwischenzug in German?)

15. .. Rxd7 16. Bxe4 Bxe4 17. Rxe4 Bd6.

Black definitely has a slight advantage, because he has the superior minor piece (bishop versus knight). The fact that the king is stuck in the center and White has a pin on the e-pawn is fairly insignificant, because the rook on d7 can easily swing over to e7 to break the pin. Also, an immediate 18. f4 could be met by 18. … f5 followed perhaps by …e4.

In our post-mortem, it turned out that White should have played 14. Qe2! instead. Robin said that he had been a little concerned about the reply 14. … Qd3, but after 15. Nxe4 Qxe2 16. Rxe2 Nxe4 17. Bxe4 Bxe4 18. Rxe4 Bd6 we would get to exactly the same position as above — except Black’s rook is on d8 instead of d7! That tiny change makes a world of difference. With the rook on d8, Black has no advantage.

From the position above, the game continued 18. Rd1 Bc7 19. Rxd7 Kxd7 20. f3 f5 21. Kf1 g6 22. Rd2.

Now I made what I think was my only significant mistake of the game — and it is exactly the same move that was good 11 moves ago! I played 23. … Rd8? when I should have played 23. … Re8.

This move is interesting psychologically because I made it very rapidly and with very little thought. I have always said that my worst moves are (a) the ones that I spend the most time on, because I talk myself out of making the best move; and (b) the ones that I spend the least time on, because I think they are “automatic.” This was the second type of move.

In the game up until this point, every exchange of pieces has made Black’s advantage bigger. That is because there were two imbalances in Black’s favor: superior minor piece and more space. There were also two imbalances in White’s favor: the somewhat exposed position of Black’s king, and Black’s lag in development. Every exchange until now has emphasized Black’s advantages while neutralizing his disadvantages. So I played 23. … Rd8 on autopilot. I was still thinking that all exchanges are good for me.

But in fact, as Robin pointed out, this position is where the exchanging should stop. Black’s rook can serve a useful purpose on the e-file, supporting an advance of the e-pawn. This move also emphasizes the difference in king position; Black has the better king, and White will be a little bit nervous about putting his king on e2 (although he’s got to do it). Robin also felt that Black’s rook and bishop will be a better pair than White’s rook and knight.

Robin backed up his analysis by drawing the game after I played 23. … Rd8? After the rook trade, we got to a position where my bishop really didn’t have many entry squares or many things that it could do. Robin had to “suffer” a little bit, but he’s good at that, and we eventually got to an endgame of K+B+RP versus K+N, which was drawn. The complete game score, if you are interested, is here. It’s possible that he would have drawn after 23. … Re8 as well, but I would definitely have had more winning chances. Again, those are the subtleties of master-level chess.

Interestingly, when I analyze the game with Fritz, the computer prefers 23. … Rd8? to 23. … Re8. Computers have no clue.

Our second training game had more mistakes in it, which were mostly made by Robin, so I won’t show you the whole game. However, there was one interesting position near the end. Of course, I’m White this time.

White is up an exchange and also has better king position, so it seems as if he should be winning. But how would you proceed here?

First I’ll tell you what I did, which was not best. Robin and I were both down to about two minutes on our clocks, but of course that is no excuse. I played 41. Rh6 — not bad, but missing the real point. And here Robin, quite uncharacteristically, mixed up the order of moves and played 41. … Nb4?? (remember, he had just two minutes left), after which I played 42. Rxb6 Kc7 43. Rxb4 (the easiest) and he resigned. But after the game, he pointed out that after 41. … Kc7 42. Kd5 Nb4+ it’s actually very hard for White to make progress. 43. Kd4 Na6 Black has basically an ideal blockading position. He said that he once had a position like this against a mutual friend of ours, Matt Noble, back in the era before sudden-death time controls. Matt played on for several hours but was unable to win. Certainly, with two minutes on my clock there is no way that I would have been able to win this if Robin had not blundered.

So have you solved the above position yet? The clearest winning move is 41. c5! The point of this temporary pawn sacrifice is that it allows White’s king to penetrate to the fifth rank and stay there, unlike the above line where I got chased away immediately by … Nb4+. After 41. c5! play might go 41. … bc 42. Kd5 Nd4 43. Rh3. Black’s c-pawn and a-pawn will fall, and White should be able to win the ensuing K+R+P vs. K+N endgame, although it may take a little bit of work.

By the way, the same trick would work after 41. Rh6 Kc7. White can still win with 42. c5! After 42. … bc 43. Kd5 Nd4 White plays 44. Kc4! Kb7 45. Rh5! and wins the c-pawn with the rook. Fortunately, the little trick 45. … Kb6 46. Rxc5 Nxb3? does not work for Black because White has the zwischenzug 47. Rb5+!

I think that the probability of my finding 42. c5! would have been somewhere between 0 and 1 percent. So I learned something from this training game — and isn’t that what a training game is all about?