A Tidbit Versus a Computer

A slew of end of summer obligations have kept me away from the keyboard for the week just passed. I have in hand a couple of bad games I played at the recent NYS Championship, and today’s game; a casual contest between a version of the famous strong computer program Rybka and Tim Wright.

Wright has been making good progress this last couple years improving his results. Recently he obtained a demo copy of Rybka and has been sparring with program. Even stripped down, Rybka is not an easy opponent. The usual formula for winning against these electronic beasts is to avoid tactics. In this case, Tim out-combines the machine.

(35) Wright,Tim - Rybka Demo Version [C19]

Casual Game September 2010

1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Qc7

I don’t know the official name of this variation. I call it the Botvinnik line because the World Champion of my youth, M.M. Botvinnik favored it when winning and defending his title from 1948 through 1961

7.Nf3,..

Sharper is 7.Qg4, and at the top level that has been successful as a winning attempt for White. Some very good players have used the text; Karpov, Shirov, Svidler, Spassky and Nunn, apparently as an alternative to 7.Qg4.

7...Ne7 8.Be2,..

This is maybe too restrained. The Grandmasters preferred moves such as 8.Bb5+; 8.h4; and even 8.a4; over this rather quiet continuation.

8..., Nbc6 9.0–0 0–0 10.Be3 Nf5

The game is now away from the positions that masters have explored. My Deep Rybka electronic helper sees things as dead level. Black has his usual possibility in the Winawer variation of the French Defense; trade pawns on d4 and create pressure down the c-file. White has his forces aimed at the Black King, again as usual in this line. So, the exact position has not come up at the higher level of play, but it is typical of the Winawer.

11.Qd2 Nxe3

I think most human players would not do this. Why give up a perfectly good Knight for a Bishop that is not doing much right now? Deep Rybka recommends the text. The notion behind it is the elimination of a potential attacking piece.

12.fxe3 Qa5

Another move that most people would not play. I expect human players would be thinking of putting the Bishop on d7 and trading pawns on d4, then building up on the c-file by bringing a Rook there. Rybka says that method is OK, but White gets a small edge. After the text, Rybka marks the game as even.

13.Bd3 h6 14.Qe1 c4 15.Be2 b5 16.h3 Bb7 17.Nh2 Ne7

Black is slightly better. The White Queen and the Ra1 are occupied with guarding weak pawns at a3 and c3 threatened by the Black Queen standing on a5. That is good for Black, but I don’t see what Black can do with his free Rook other than playing .., f7-f6; and trading on e5. White has no intention of going along with that idea.

18.Ng4 Nf5 19.Bf3 h5 20.Nf6+?!..,

A bold decision. Making combinations with computers is not to be recommended, they do not lose their composure and calculate accurately. The obvious safe alternative 20 Nh2, and Black is just a bit better than White with tension continuing. Now Black is winning.

20..., gxf6 21.exf6 Qc7 22.Bxh5 Qg3 23.Qd1 Nxe3 24.Qe2 Nxf1

This is the first material cost of the flawed combination White has undertaken.

25.Rxf1 Rad8?

Most uncharacteristic of a computer program. Black needs to take some actions to guard the weaknesses the sacrifices have created around the Black King. Best is 25.., Kh7; and a reasonable alternative is 25.., Qg5. White now has a chance to make some progress.

26.Rf3 Qg5 27.Bg4?..,

Missing the opportunity to test the program with 27 Qf2, which seems to force a draw after 27..., Qxh5 28 Rg2+ Kh8 29 Rg4. One has to admire Tim’s spirit of adventure. He must have believed there was a way to make his attack stronger.

27..., Bc6 28.Qf2 Kh7 29.Rg3 Rg8 30.h4 Qc1+ 31.Kh2 Qxa3?

Computer sees hanging pawn, computer grabs pawn?? This is not the way the more sophisticated versions of the Rybka play. Safe and sound keeping a winning advantage is 31..., Qh6. White now has a shot that turns the tables.

32.Bf5+!,..

The only way to stay in the game. The move is strong, so strong it wins outright.

32..., exf5 33.Qxf5+ Kh6 34.Rg5 Qd6+ 35.Kg1 Qxf6

Computers don’t resign and the game goes on for a few moves.

36.Qxf6+ Rg6 37.Qxd8 Be8 38.Qxe8 Rxg5 39.hxg5+ Kxg5 40.Qxf7 Kh6 41.g4 b4 42.Qf6+ Kh7 43.g5 b3 44.Qf7+ Kh8 45.g6 bxc2 46.Qh7, Checkmate!

A very nice demonstration of Mr. Wright’s chess imagination.

More soon.

One More Game From the Lower Boards

Here is a game from the night of the big Albany versus Schenectady Match. This was not counted in the final score because both players are SCC members. Mike and Jeff are adult players laboring away in the middle of the pack trying to improve their chess.

Stanley, Mike - Capitummino, Jeff [D53]

Albany - Schenectady Match SCC, 08.19.2010

1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.Nf3 c6 5.Bg5,..

All the K’s; Kasparov, Korchnoi, Karpov, Kramnik, have played so, as well as just about every other GM, except those constitutionally opposed to 1.d4, so this can not be bad.

5..., Be7

Lots of Grandmasters have played this move, but not the elite cited above. Kasparov liked 5..., Nbd7; Kramnik and Anand prefer 5..., h6; with only occasional forays with 5..., Nbd7. There are many subtleties about move order and many transposition possibilities in this opening.

6.e3 dxc4?!

This move is not the best. Black usually moves the Knight on b8 to d7 first and then develops the Bc8 to b7, after .., b7-b6; the Tartakower variation of the QGD. The text gives White a somewhat better development than is customary with a bit more space in the center. Jeff seems to have mixed ideas from Slav and the Cambridge Springs variations with the Semi-Slav. I did something similar against GM Har-Zvi and paid the typical price; White builds up pressure on the c-file eventually winning a pawn leaving Black with a hopeless endgame when playing anyone who knows the classics. In this contest the technical knowledge is not so high as against a GM and there is the chance White will miss his opportunity.

7.Bxc4 b5 8.Bd3 Nbd7 9.Qc2 h6 10.Bh4 Bb7 11.0–0 a6 12.Rae1?!,..

This looks clumsy. More logical is 12.Rd1, or 12.Rc1. The game is about equal according to Rybka.

12..., c5!

This move, or possibly 12.., Rc8; are correct. Black gets rid of his c-pawn dissolving a weakness and opening lines for the Bb7. Mike did not realize how important it is to restrain .., c6-c5. White could have, earlier in the game, captured the Nf6 with his Bishop expecting .., Bxf6; just to be able to prevent the advance of the c-pawn.

13.dxc5 Nxc5 14.Rd1 Nxd3 15.Rxd3 Qc7 16.Bg3 Qc8

More forceful is 16..., Qc5; keeping c8 free for a Rook, and thinking of retiring the Queen to b6 if the Nf3 goes on the tour Nf6/d4/b3. It is possible to consider 16..., Qb6; right away, but then Black has to worry about White doubling his Rooks immediately and calculate if he has the time to prevent a White Rook from getting a place on the 7th rank

17.Qd2,..

White is alert to threats developing on e4; if 17 Rfd1 Be4; pins the Rd3, and the Nc3 can’t help for he is pinned over his Queen. Other than some confusion about the theory of the QGD, these guys are playing very well. They are tactically alert and have some ideas.

17..., 0–0?!

A sensible routine move that misses an opportunity. Black is worried about the White Bishop getting on d6, probably feeling if he trades Bishops there, the White Rook on d6, backed by the Queen and eventually the other Rook at d1, his game would be worse. What Jeff did not see is; 17.., b4 18 Na4 Qc6 19 b3 Ne4; hitting the Qd2 and eliminating the Bg3. He then has time to castle, and with the center open, his Bishop pair have prospects.

18.Rd1?!,..

White has been wanting to triple on the d-file and now he has gotten his wish. But, why not make some hay out of the awkwardness of the Black major pieces with 18 Rc1? Where does Black put the Queen? Likely Black has to trade a Bishop for air for the queen with 18..., Bxf3 19 gxf3 Qb7 29 e4 covering f3 and taking d5 away from the Knight leaves White well placed. After the foregoing, the danger of the two Bishops is gone. The game is nearly equal now.

18..., Rd8 19.Rxd8+ Qxd8 20.Qxd8+ Rxd8 21.Rxd8+ Bxd8

Mike elected to trade off all the major pieces on the d-file. He could have varied with 20 Bd6, when after 20..., Bxd6 21 Qxd6 Qxd6 22 Rxd6 b4 23 Na4 Rc8; while the Bishop pair has been broken up, the Black pieces are well placed to maintain the initiative. As it is, Black is slightly better and White has chances to hold the game.

22.Ne5?,..

But not this way. Black now turns the game into a battle between two Bishops and two Knights. Better is 22.Bd6, giving his Bishop more space in which to operate.

22..., Nh5 23.Bf4,..

Why spend a move to get a pawn on f4? Letting the trade happen on g3 and using the tempo gained to move his King towards the center is a good option.

23..., Nxf4 24.exf4 Bf6?!

It is Jeff’s turn to miss-deploy a Bishop. At f6 the Bishop only can exchange itself for the Ne5, instead at b6 the Bishop threatens f2 and cuts through the White position. Long ago Steinitz laid down the strategic rules for fighting Knights; take away their squares. With the Bishop at b6, Black can kick the Ne5 with .., f7-f6; not so with a Bishop standing there.

25.Nd7,..

White may have been worried about 25.., g5; and so decided to vacated e5. It is better to try 25 a3, then 25.., g5 26 Nd3 Bd4 27 fxg5 hxg5 28 Kf1, and White is holding.

25..., Bxc3?

Why surrender the Bishop pair? Best is 25.., Bd4; working to restrict the available squares for the Nd7.

26.bxc3 Bd5 27.a3 Bb7 28.Nc5 Bc8

White has equalized. The game score had some mistakes in it around this point. It appears one or two moves were left out. After several attempts the following are the moves I was able reconstruct.

29.Kf1 29.Kh7 30.Ke2 Kg6 31.g4 f6 32.Ke3 e5 33.f5+ Kf7 34.h3 g6 35.fxg6+ Kxg6 36.Kf3 f5 37.Kg3 Kg5 38.a4?,..

After doing good work for several moves. Mr. Stanley loses his way. Without any notes on the clock time situation, it can not be said for sure that time trouble was the cause, but the move looks as if worry about the clock provoked the error. With 38 f3, or 38 h4+ and 39 f3, White is well on his way to simplifying the game to a draw. The text tips the game in favor of Black because a pawn is lost on g4.

38..., bxa4 39.Nxa4 a5 40.c4 fxg4 41.hxg4 Bxg4 42.c5 Bd7 43.c6 Bxc6 44. Resigns.

Time trouble or a terrible reluctance to face tension over the board are the explanations most often advanced for the kind of errors seen in the final phase of this game. It is interesting to see these gentlemen developing their chess ideas. If they fix the holes in their opening knowledge, and play a bit more often to improve their tactical awareness, both can be dangerous opponents for most Schenectady Club members.

More soon.

A Game From the Lower Boards

Here is a game from one of the lower boards of the match. It came to my attention when Charles Eson and I went over the last few moves of the game at the Albany Area Chess Club. Charles did not have the game score at hand, and he tried to reconstruct the position from memory. The discussion was interesting and educational; interesting because of a couple of points that came to light from what Eson remembered of the position as compared to the actual game position, and educational because observing how he re-created the position on the board at the club and his perception of the critical points of the game.

Dipre, George - Eson, Charles [B59]

Albany - Schenectady Match SCC, 09.02.2010

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 e5 5.Nb3,..

Not a particularly sparkling choice. The natural positional continuation is 5 Bb5+, and if 5..., Bd7 6 Bxd7 getting rid of Black’s better Bishop. If Black plays instead 5..., Nd7; his development will be slower than in the game.

5..., Nf6 6.Nc3 Nc6 7.Be2 Be7 8.0–0 Be6 9.f4,..

Black is playing to get in the advance .., d6-d5; achieving equality. The game move is best way to make that unpalatable for Black.

9..., Bxb3?

This is not a sensible move. Better is 9..., Qb6+ 10 Kh1 exf4; or 9..., exf4; straight away putting aside concerns about the weakness of the backward d6-pawn on an open file. Bronstein wrote about how to use such a pawn in his masterpiece on the Zurich Inter-zonal of 1953, in the Averbakh - Gilgoric game where Black demonstrated the dangling the d6-pawn can be effectively defended by threats on other parts of the board letting Black get all his forces well mobilized. The Averbakh - Gligoric game was a KID, but the ideas apply in this sort of Sicilian also.

Black has been playing to break with .., d6-d5. It is not quite possible to do so just yet safely; 9..., d5?! 10 f5 Bc8 11 exd5 Nb4 12 Bg5 13 Nxd5 Nxd5 14 Bb5+, Bd7 15 Qxd5 Qb6+ 16 Nd4 Qxd4+ 17 Qxd4 exd4 18 Bxe7 Bxb5; and notwithstanding the Bishops of opposite color, Black has a difficult game. His King is stranded in the center and the d-pawn is hard to defend. So, why trade a well placed Bishop for a Knight on b3 that has no immediate prospects? It seems to be another example of a club level player not being willing to face ongoing tension in a position.

10.axb3 0–0 11.f5 a6 12.Be3 b5 13.Qe1 Qc8 14.Bf3 Nb4 15.Qf2 Qb7 16.Rfd1,..

Putting the other Rook on the d-file is more in tune with the needs of the position. If there is a solid reason for the advance of the f-pawn to f5, it has to be the idea of advancing the K-side White pawns at some point. That suggests d1 and f1 are the right posts for the Rooks.

16..., Rfd8 17.Bb6 Rd7 18.Ba5?,..

Why not the logical 18 Rd2, and doubling on the d-file? White may have miss-calculated the line; 18.Rd2 d5; forgetting how powerful the Bf3 is, or he could not make up his mind about where he wanted to double the Rooks, the a-file, or the d-file. Whatever is the case, this move and the next give Black the advantage.

18..., Nc6 19.Nd5 Nxd5 20.exd5 Nxa5 21.Rxa5 Bf6?

Missing completely a wonderful opportunity. Winning quickly is 21..., Bd8; when the threat of 22..., Bb6; pinning the White Queen over his King requires White to give up the Ra5.

22.Kh1,..

White has seen the threat and takes precautions. The upshot of the recent maneuvers by White is the pressure on d6 is masked by the pawn on d5, that is not good for White. Black could now shift his Bishop to active duty on b6, or make some threats down the c-file by playing 22.., Rc7.

22..., Re7?

Black is swimming. The text obstructs the Bf6 and puts the Rook on the least active square available. I suspect Mr. Eson ran out of ideas thinking “I’ll pass” and see what my opponent intends. That is very dangerous. One technique to use when the idea-well runs dry is list in your mind every legal move that is on the board before making a choice. GM Jacob Aagaard suggests this as useful when you are in a quandary and also as technique to steady your nerves in positions of great tension.

23.Be4 Ree8 24.b4 Bd8

Black found the right idea but too late. White now has a route open along the 3d rank to the K-side for his Rook. Since Black does not have the best resources for the defense of his King on the board, and his available resources are not well placed for K-side defense, danger looms.

25.Ra3 Bb6

This move drives the White Queen to where she wants to go. It can’t be criticized too much if Black has accurately assessed the position.

26.Qh4 f6 27.Rh3 h6 28.Qg4 Kf8 29.Qg6 Qc7

A move that leads me to think Black has not evaluated the position accurately. Better is 29..., a5; taking the opportunity to open the a-file. White really needs a dark squared Bishop, or a Knight to sacrifice at h6 or f6 to have real zing in the attack he has built. Nevertheless White has the initiative. Accepting he does not have the right force to break through directly on the K-side, White could shift focus here and play 30 Rc3 Qe7 31 Rc6 Bc7 32 Ra1, then if Black panics and plays 32..., a5?!; when White has the sophisticated line: 33 Qh7 Ra7 34 bxa5 Rxa5 35 Rxa5 Bxa5 36 Bf3!, threatening 37 Bh5. In this way White could have used the initiative and the principle of two weaknesses (the Black Q-side pawns and the squares around the Black King) to obtain a near winning advantage.

30.Qh7 Qf7

The best move. If Black gets ambitious with 30..., Qc4?; White has a long winded and winning attack beginning 31 Rg3 Ra7 32 Qh8+ Ke7 33 Qxg7+ Kd8 34 Qxf6+ Kd7 because 34..., Kc7/c8; allows the immediate 35 Rc3, pinning the Queen. When the Black King goes to d7, the White Queen and Rook will range along the 6th and 7th ranks sowing destruction and ruin on the Black position leaving no time for Black to safely capture the Be4.

31.Rg3 Rec8 32.c3 Rc4 33.Bf3 Rh4?!

After avoiding several tactical dangers and arriving at an equal position, Black becomes overly ambitious. I can not see how this sortie with a Rook and a Bishop is going to improve his position. Better 33..., Qg8 34 Qg6 Qf7; repeating the position and making a draw, or 33..., a5; opening the a-file when both sides will have to be very careful, but Black looks to be getting his nose in front.

34.Rg6 Bf2?

A bad mistake. Necessary is 34..., Qg8; White then could be tempted to go for broke with 35 Rxf6+, which is met by 35..., gxf6 36 Qb7 Re8 37 Qxb6 Qg3! 38 Qxd6+ Kf7 39 Kg1 Qxh2+ and Black is winning although certainly he would have to be watchfully careful of tricks. After the game move White can get back on top with 35 Rf1.

35.Rd2?,..

This move with same intention fails to do the job.

35..., e4?

Better 35..., Be3; and the Black Bishop has just enough free squares to avoid the White Rook’s attempt to capture the cleric. The game move gives away a pawn for no good purpose. Worse is that the tactical tricks White has in hand remain potent.

36.Rxf2 exf3 37.Rxf3 Re4 38.h3 Ra7?

Required is 38..., Qg8; to force off the Queens. After the text the best Black can hope for is an endgame with poor chances to hold.

39.Rfg3 Qg8?

This should lose the game.

40.Qxh6?,..

I guess both sides were short of time the last several moves. White misses 40 Rxf6+ Rf7 41 Rxf7+ Kxf7 41 Qg6 + Kf8 42 f6, winning. If you think Black has other moves in the sequence cited, look again. Other than the spite check on e1, all alternatives lose quickly and the check just delays things for one move. The move played leaves White with a substantial advantage, but makes one wonder if George is seeing things clearly in time pressure.

40..., Re1+ 41.Kh2 Ree7?

Charles is having the same problem. This move does nothing to prevent 42 Rxf6+ Ke8 43 Rxd6, and White has too many pawns for Black to even consider 43..., gxh6 44 Rxg8+ Kf7 45 Rgg6. Better is 41..., Rf7; when things look bleak for Black, but where there is life there is hope, and the clock is running.

42.Qf4 Qxd5 43.Rxf6+?,..

White has been forced to move quickly by rapidly disappearing time. The move played is utterly wrong, but it shows the right ideas were in the player’s mind. That has been the case for both sides throughout the game. White could have brought home the point with 43 Qh4, then if Black plays 43..., Qf7 44 Qh8+ Qg8 45 Rxf7+ Rxf7 46 Qh4, and now all roads lead to a White victory. Black can not stand for any sequence that trades off the Queens and Rooks - the extra White pawns on the K-side are unstoppable. White could have used that fact to win more material or mate the Black King. The game move drops a whole Rook without gaining even a short lived initiative.

42..., gxf6 44.Qh6+ Ke8 45.Re3 Qf7 46.Rxe7+ Rxe7 47.Qd2 d5 48.Qf4, and White lost on time.

Recalling Charles working to reconstruct the position from about move 40 and his comments during the effort enlighten me in several ways; first, it appeared he did not realize that many critical errors happened earlier, most particularly the two-move sequence on move 21 (21.., Bd8; instead of 21.., Bf6;) that were as important as the final positions. Secondly, Eson had questions about the position but had not sat down to investigate the position even two weeks after the event. And, thirdly, he was unclear about just what the position was in fact.

What can be learned from these observations?

First, an almost absolute requirement to achieve improvement in chess skill is analysis of your games. For us non-professionals, time is not always available for study, but two weeks is a very long time to let a game where you have questions to linger. The sooner you put a game you don’t quite understand under examination the better will be your progress.

Second, less successful players have trouble recalling positions, more successful players do so with less effort and fewer mistakes. The most recent experience I have had with this phenomenon was working with GM Ronen Har-Zvi. Time after time the GM astounded me by recalling games he played fifteen or twenty years before, and other associated GM games when we’d come across a related position in analysis. Was this just the difference in talent between a GM and the rest of us? Perhaps, yes, but does this not tell the more average player that working on fixing in our memory critical positions is an important skill? To that end, it is useful to at crucial points in a chess game to use a couple of minutes to memorize the board position. That is; don’t think about; moves, threats, imbalances, hopes or chess fantasy, focus on the location of pieces enough so that you can recall the position at will. Jacob Aagaard sees this as a vital technique to improve calculation skill.

Tonight the NYS Championship begins for those of us playing the four day schedule. I may not get games other than my own for the blog. This tournament always is a struggle for me, more than one per day poops me out. I hope to see you there.

More in a day or two.

Another Albany - Schenectady Game

An important game in the recent match was the encounter between Dean Howard, one of the “old guard” Experts locally, and Philip Sells, a newly minted Expert. It was very nearly the last game to finish, and had Mr. Sells not carried the day, it would have led to a most surprising upset win of the match for Albany.

Howard, Dean - Sells, Philip [B22]

Albany - Schenectady Match, SCC, 19.08.2010

1.e4 c5 2.c3 d5 3.exd5 Qxd5 4.d4 Nf6 5.Nf3 Bg4 6.Be2 Nc6

The c3 Sicilian is one way for White to avoid the heavily analyzed body of theory in the mainstream Sicilian. Philip Sells is a notorious “Siciliano” who has excellent results in both the open lines (3 d4) and the closed lines (an early Nb1-c3 with no d2-d4). The c3 Sicilian, otherwise known as the Alapin variation, shows in my databases only a small edge for White - 1180 wins, 1150 drawn and 1090 lost in general. For this particular move order not quite so good results are reported; White won 198, drew 180, lost 210.

7.c4!?,..

More normal is 7. Be3.

7..., Qd6!?;

Both players are striving for originality. With 7..., Qf5 8 d5 Bxf3 9 Bxf3 Nd4 10 Na3 0-0-0; Black could have obtained a comfortable game. I believe neither side wanted to see off a pair of minor pieces just yet; they want to keep the tension in the position.

8.d5 Ne5 9.Nc3,..

White is somewhat better placed than is Black.

9..., a6

I like the way Philip plays chess, it is “hard chess”. He knows that risks have to be taken if you want to win. Objectively, trading a couple of pairs of minor pieces at f3 might bring the game closer to the friendly splitting of the point, but maintaining tension keeps alive winning chances.

10.0–0 Nxf3+ 11.Bxf3 Bf5

One pair of minor pieces go off, but Philip retains his light squared Bishop for both attack and defense. Howard now has a solid positional edge. He will develop easily, the e-file is naturally in his hands and he controls a bit more space.

12.Bg5!?..,

It looks natural for Black to castle long and 12. Qb3, is a natural answer to that possibility. Another try is 12 Re1, taking control of the e-file. The text is another valid path. It makes both sides make difficult positional judgments over the next several moves.

12..., g6?!

Now White can get a firmer handle on the position if he reads the position correctly. Black would do better with 12..., 0-0-0.

13.Qd2?!..,

Piling up on the e-file with 13 Qe2, then 13..., Bg7 14 Rae1, will make castling difficult for Black. If not impossible.

13..., Bg7 14.Bf4 Qd7 15.d6!?..,

White decided taking immediate action in the center before Black can castle is the best course. Another way to go forward is 15 Rfe1,

15..., e6!?

Black continues to court danger in order to play for victory. Maybe better is 15..., 0-0-0; I suspect Sells did not care for the immediate tactical dust-up after 16 Nb5. Of course 16..., axb5 17 Qa5 exd6 18 Qa8+ Kc7 10 Qxb7 is mate, but the alternative 16..., e6 17 Qe3 Ne4 18 Bxe4 Bxe4 19 Qxc5+ is so dangerous that Black will have to give up too many pawns to stifle the raging attack White can make. One other try in this line should be examined; 15..., 0-0-0 16 Nb5 e5?!; while this move is flawed, White would have to calculated 17 Bxe5 Rhe8 18 Bxf6 Bxf6 19 Qa5 Be4 20 Bg4 Bf5 21 Bxf5 gxf5 22 Rad1 Re4 23 Rd5 Rxc4 24 Na3, and White is better. This is a line with many complications and sub-variations, quite a challenge to find one‘s way through. It may be Black’s best chance.

16.Qe3?!..,

Missing the strong move 16 Qd1, threatening 17 Na4, hitting both the pawn on c5 and the tender b6 square, as well as preparing 17 g4, which picks up a pawn after 17..., g5 18 Bxg5 Bg6 19 Bf4, when White is clearly has a significant advantage. After the text move White retains some of his edge.

16.., Rc8 17.Rad1 0–0 18.b3 b6

At this moment in the game White had 17 minutes remaining and Black 13 minutes of the 90 minutes allotted. Serious time trouble was near for even these two “time trouble specialists“.

19.Na4 Qa7 20.d7?..,

White trades the pride of his position, the d-pawn for the Black b-pawn giving up just about all of his advantage. Better is 20 Be5, then if 20..., Ng4 21 Bxg4 Bxg4 22 Bxg7 Kxg7 23 f3, embarrasses the Bg4. And, worse is 20..., Ng4 21 Bxg4 Bxe5 22 Qxe5 Bxg4 23 f3 b5 24 d7 Rcd8 25 Nxc5, when a pawn is gone while the Bishop is still in trouble. Taking a different track is the immediate 20..., b5; hoping to make an outlet for the Bf5 via c2 and a4. This is the best bet for Black, but White retains a substantial advantage.

20..., Nxd7 21.Rxd7 Qxd7 22.Nxb6 Qd8 23.Nxc8 Bd4?;

An error that could lose the game, or at best leave Black struggling to hold the draw. The straight forward 23..., Qxc8; is best. Has Black forgotten about the problem Bishop at f5 to say nothing of the potential weakness of his K-side? It seems so.

24.Qe2 Qxc8 25.Rd1?,..

Here the best move for White is 25 Bh6. If the Black Rook moves away, 25..., Rd8; we enter an absolutely fascinating tactical sequence; 26 Rd1 (not 26 g4 Bd3! 27 Qxd3 Bxf2+; is winning for Black) 26..., Qd7; (if 26..., e5 27 Rxd4!, followed by 28 Queen to the e4-square with mate soon) 27 Bg5 R-moves 28 g4, and the hapless Bf5 is trapped. It is probably best for Black to give up the Exchange with 25..., e5; then 26 Bxf8 Qxf8 27 Be4 Bd7 28 Bd5 Kg2; and there will be a long technical struggle for White to demonstrate the superiority of his Rook and Bishop over the two Bishops. Given the small amount of clock time each player had at this point in the game, it certainly is easier for Black to hold the draw than it is for White to show the win.

25..., e5 26.Bh6 Re8 27.g4,..

Now the Bf5 has options and this move introduces an unnecessary weakness in the White formation. Better 27 h3, or 27 Qd2. The game is just about dead even now.

27..., Bd7 28.h3 Bc6 29.Bd5?!,..

Making another weakness unnecessarily. Better 29 Be4, 29 Be6, 29 Bd2, or 29 Bxc6. Even though his time is dwindling, Mr. Howard is trying to win the game. This transaction trades the newly minted d-pawn for the Black a-pawn creating a passed outside pawn for White. The idea is good if properly followed up.

29..., Bxd5 30.cxd5 Qd7 31.Qf3?..,

Time pressure was by now severe for both sides, something like a minute and change. Dean made this move almost instantly and unfortunately it is an error. Better 31 Qxa6, then 31..., Qxd5 32 Qf6, (setting up a mate threat that limit’s the choices for Black) 32..., e4 33 Qg5, and the game is headed towards equality.

31..., e4 32.Qe2 Qxd5 33.Qxa6??..,

This now drops the Exchange.

Bxf2+ 34.Kxf2 Qxd1 35.Qf6 Qd4+

It is likely Dean did not see this reply when he chose 31 Qf3. The mate is prevented and the Queens are forced off leaving things hopeless for White.

36.Qxd4 cxd4 37.Bf4.

Bill Townsend records Howard’s clock fell in the next couple of moves. Dean said he resigned. My notes indicate both happened at about the same moment.

This was a fighting game with honor for both sides in their stern resolve to go for victory. The all out effort by the contestants, as well as the importance of the game to the results, made the game interesting and well worth the time to play it over, analyze it and make it part of this blog.

I thank the participants for their considerable efforts.

My score of this game, recorded while looking over the shoulders of the participants, was not quite accurate after move 33. In the fury of time pressure I missed a couple of moves. The generous help Bill Townsend extended allowed the correct reconstruction of most of the final moves of the game. My thanks to Bill also.

More soon.