Some News and Endgame Thoughts

The State Championship marks the beginning of the chess year here in the Capital District. The next usual steps are the various club events. I have heard from Richard Chu, the President of the Schenectady Chess Club that the dates are set for the beginning of their season. The organizational meeting will held on September 25 at the Club on Aqueduct Road. The Championship tournament is to begin on October 9. If you are considering playing in the Schenectady event, it is advisable to attend the organizational meeting. It is there the format and rules for the tournament are hashed out, and you can have your say in how the event will be conducted.

The 136th NYS Championship took place over the Labor Day weekend. It was well attended with 231 participants. The Open Section was headed by two famous names; Gata Kamsky and Maxim Dlugy. GM Kamsky is a New York State resident and with a 5 ½ – ½ score he took the title and won the tournament is fine style. GM Dlugy was his only opponent to hold Gata to a draw. That happened in Round 4. Dlugy and Bonin tied for second place with identical 4 ½ – 1 ½ scores. Jay Bonin lost to Kamsky, and Dlugy to IM Tim Taylor, both in Round 5.

Of the local players Patrick Chi was the highest finisher, 4-2 points tied with five others just behind the top three. A good result for Patrick and he picked up a handful of rating points.

In the Under 2100 Section Andrew Ardito won going away with a 5 ½ – ½ tally. He was so far ahead of his competition that he was able to take a half-point bye in the final round and secure the victory. The best local finisher was Jeremy Berman at 4 – 2. His only loss was to Ardito.

The Under 1800 Section was won Alan Geiger with 5 ½ – ½. Mr. Geiger was something of a surprise winner. Going in his rating was 1485! He had not played in USCF rated event sine 1997. His long break must have been well used to achieve such a victory.

Mingrui Liu won the Under 1500 Section with a 5 – 1 score. This is quite an accomplishment with less than one year‘s experience. Liu began the event with a just 1284 rating and finished with 1461!

The Under 1200 Section was won by Nippon Makkar with a perfect 6 – 0 score. This was Nippon’s first rated tournament. It will be hard to improve on that standard.

At serious chess events, matches and tournaments, we see the players diligently recording their games. I wonder how many players analyze these laboriously recorded games? Many study only master games, opening monographs, and endgame manuals. I think we, the local non-masters, are missing the boat. Our own games can provide grist for the learning mill. Just playing over our triumphs and failings while looking for specific improvements is not the only or best way to use our own games for study. Here is some thoughts on that subject:

A few years back Chuck Eson of the Albany Chess Club and I worked together on some of his games. We were specifically trying to improve his results. Chuck seemed to do well up to a certain point in many games, then disasters happened. From one particular game we found a position that I dubbed: the Eson Position:

Mr. Eson misplayed this ending and permitted the following formation to come about:

Beginners often fall into this sort of positional problem; the Black pawn is so far advanced that it cripples White’s K-side majority. Even players with some experience encounter the problem occasionally. If White tries to advance his g-pawn, Black then takes on g3. The thing that allows this error to be made easily is that White’s King is not too far from the scene, and White can err by relying on this closeness in his calculations. In the Eson position the existence of the passed e-pawn is decisive after the Black h-pawn goes to h4. Then the advance of the White g-pawn permit’s the creation of passed Black e and g-pawns, and they are immune from capture freeing the Black King to go after the outside passed pawn.

Up to this point the Eson position is only moderately interesting to the more advanced player. The standard solution for White is: 1 g3 a5 2 h3, (White does not need to move his Q-side pawns, but Black must move his. Every step forward the Black pawns make bring them closer to the White King.) 2…, b5 3 g4 h4 4 a4, (A timely use of a tempo. White does not want to rush the creation of a passed pawn until he is well placed to harvest the Black Q-side pawns.) 4…, b4 5 c4 b3 6 c5 Kd5 7 g5, (Drawing the Black King back like a magnet.)

7…, Ke5 8 g6 Kf6 9 Kxe4 Kxg6 10 Ke5 Kg5.

The c6-pawn falls and the finish is a matter of counting. It will take White six moves to capture the c-pawn and Queen, while Black requires seven moves to Queen the h-pawn.

The ace-in-the-hole for White from a calculation point of view is, in the worst case, he might have to trade Queens, but his King is perfectly placed to win another pawn on the Q-side and again make a Queen there. Actually, if you calculate precisely, White has enough time to prevent Black from Queening. Having multiple winning possibilities is important practically. In most of our local events the time controls are Game in x minutes with a five second delay, and such endings are most often conducted in some kind of time trouble. Having the knowledge of a fall-back winning procedure gives White an insurance policy and permits quick decision making.

The above ideas are well known. As Chuck and I worked on the position I posed the question: What if there were other pieces on the board? Adding Rooks or minor pieces may change the evaluation, and if so, how much and in which direction?

Say there are Rooks on the board:

With Rooks on, White’s most forceful try is: 1 Rd4, and if Black defends pawn with 1…, Rf4; a tactical situation comes about where inaccuracy will be costly. For example: 2 g3?? Rf3+; wins for Black. Similarly, after the better move: 2 h3,.. Black may err with 2…, c5?; when 3 Rd4+, forces the Rook trade. Then the attempt to hold after 3…, Kxd4 4 Kxf4 h4 5 g3 hxg3 6 Kxg3 Kd4 7 Kf4 e3 8 Kf3,

shuts out the Black King, and the distant passed pawn will win by drawing the Black King away. White then goes after the Q-side pawns. If Black then does the same, all White has to keep in mind is not letting the Black a-pawn get beyond a5. This is done by advancing the White a-pawn to a4. When all is said and done as far straight forward pawn captures goes, the White King will be in front of his own a-pawn and the Black King behind it. That is won for White.

Instead of 2…, c5?; Black can try to create counter-play on the Q-side for his Rook with 2…, b5; and if White tries to stifle Black with 3 b4, possible is 3…, Rf1; with chances for counter-play. Very probably Black is lost even in then, but it will take twenty or so moves to demonstrate this. Black can make things more difficult for White by avoiding trading Rooks.

To sum up: White has good winning chances with Rooks added to the basic Eson position, but Black certainly has chances for counter-play.

Moving on to adding Bishops to the position. This is much more complex.
The possibilities are numerous: Same color Bishops, two flavors: light squared or dark squared, and opposite colored Bishops, Black with light and Black with dark. Because of the size of the problem set up by the question about Bishops, going into detailed lines is daunting. I will limit my comments to the following:

GM Alex Yermolinsky in his ICC series “Every Russian School Boy Knows” has covered in fine detail much about the various Bishop versus Bishop endgames. If you want to practical examples of Bishops versus Bishop battles, this series of lectures is an excellent place to begin. One thing is consistent across all the several different kinds of Bishop endgames in Yermo’s lectures is the absolute need to calculate position by position, and to do so accurately. Some general principles are good guidelines in Rook endgames, but it does not seem to be so in Bishop endings. The key factor is: while Rooks can cut off pawns on any square of a pawn’s file, Bishops have one-half the chances to do so. This fact seems to let any passed pawn move forward faster than the same pawn can do so in a similar Rook ending. The speed of advance for passed pawns in Bishop endings requires exact calculation to avoid unpleasant surprises.

We now come to the Knight versus Knight situation. The standard piece of chess wisdom is Knight endgames are very like pure pawn endgames in that they are very concrete. Dvoretsky says: “… even the tiniest change in the position generally alters the shape and outcome of the struggle.” As in pawn endings the creation of an outside passed pawn is a major factor. The inferior side, with the presence of one more piece in the mix, has more resources with which to hold the position, or alternatively, perhaps make balancing threats on some distant part of the board.

With Black to move.

As in pawn endings the relative King positions are critical. Also a great deal depends upon which side’s Knight is better placed. In this first instance White is threatening the e4-pawn. Black needs must defend his pride and joy dynamically because the other support squares, c5 and h5, can be attacked by pawns immediately. To make this study more interesting Black is on the move. Here is a sample line:

1…Nf4 2.Nxe4 Nxg2+ 3.Kf2 Kxe4 4.Kxg2 Ke3;

and we have a position where most of us can visualize much of the subsequent play. It is a counting exercise. How many moves to make a White Queen and how many moves to make a Black Queen? For White the count is nine moves. For Black the count, at first glance seems to be eight moves. However, White can try to throw a monkey wrench into the count with c2-c4 when the Black King steps onto d2. Play could continue:

5.Kh3 Kd2 6.c4 b5 7.b3 Kc3 8.Kh4 b4 9.Kxh5 Kb2 10.Kg5 Kxa2 11.h4 Kxb3 12.h5 Kxc4 13.h6 b3 14.h7 b2 15.h8Q b1Q;

when Black chances are better than White’s. A whole new type of endgame is at hand: Queen versus Queen with extra pawns. In light of this line White is better served by 2 g3, and trading his a-pawn for the Black e-pawn.

What if the Black Knight is somewhere else?

This time we will return to the situation of the basic Eson position with White on the move.

1 c4, White plays to take away squares from which the Black Knight can check the White King, 1…, a5 2 b3 Kf5 3 h3 h4 4 a3 b5 5 cxb5 cxb5 6 Kd4 Kf4 7 Nd1 a4 8 b4 Kg3 9 Ne3 Kf2 10 Nf5 Kxg2 11 Nxh4+ Kxh3 12 Nf5 Kg4 13 Ke5 Nd5 14 Kxe4 Nc3+ 15 Ke5 Kf3;

and while the position still requires careful attention, it should be drawn. What can’t be lost sight of is the sacrifice of the Knight for the last pawn. That, along with a Knight sacrifice to break free an unstoppable run of a pawn to Queen, are typical tactics in Knight endings.

It is not surprising that even strong local players might find such an endgame challenging, and particularly so when there is little time remaining on the clock. Both Bishop versus Bishop and Knight versus Knight endings have a high premium on exact calculation. What about the dreaded Bishop versus Knight ending?

The Bishop versus Knight version of the Eson position can be of four different types: 1)Black with the poorer Bishop and White with the Knight on some reasonable square, 2) Black with an aggressive Bishop and White with a reasonably placed Knight, 3) White with the aggressive Bishop and Black with a reasonable placed Knight, and finally White with a less aggressive Bishop and Black with the Knight. It should be noted that whatever color of Bishop White has, it will have considerable scope. The same is not true for Black. A Bishop on the light squares is poor for Black because it can only defend e4 while being restricted by that very pawn.

For the first situation:

Here is a sample line of play:

1.Ne2 Bf7 2.Nf4 c5 3.b3 c4 4.b4 Be8 5.Nh3 Ba4 6.c3 h4 7.Ng5 Bc6 8.Nf7+ Kf6 9.Nd6 b5 10.Nxe4+ Ke5 11.Nc5 Bxg2 12.Nxa6 Bc6 13.Nc5 Kf5 14.Kf2 Kg4 15.Ne6 Kh3 16.Kg1 Bd7 17.Nd4 Be8 18.Nf3 Kg4 19.Kf2 Kh3 20.a3 Bh5 21.Nd4 Be8;

and the game is roughly equal. The Grandmasterly wisdom I have heard quoted by someone commenting on the recent St Louis tournament: “Bad Bishops guard good pawns” seems to apply here.

For the second situation:

Here is a sample line of play:

1.Nc3 Bg5+ 2.Ke2 b5 3.Nd1 h4 4.Ne3 Bf4 5.Kf2 Kf6 6.g3 hxg3+ 7.hxg3 Bc7 8.Ng4+ Ke6 9.c3 a5 10.Ke3 Kf5 11.Nf2 Bb6+ 12.Ke2 Bd8 13.Nd1 Bc7 14. Ne3+ Ke6 15. g4 Bd8;

and clearly Black is somewhat better, but White may well hold the slightly inferior position.

For the third situation:

Play might continue:

1.g3 Ng4+ 2.Ke2 Nxh2 3.Bxb6 Kd5 4.b3 Nf3 5.Ke3 Ke5 6.Ba5 Nh2 7.Bc3+ Kf5 8.Bg7 Nf3 9.Bh6 Ne1 10.c4 Nc2+ 11.Kd2 Nb4 12.a4 Nd3 13.Bf8 Kg4 14.Ke3 Kxg3 15.Kxe4 Nc1 16.b4 h4 17.b5 axb5 18.cxb5 cxb5 19.a5 Ne2 20.Bd6+ Kg2 21.a6 Nc3+ 22.Ke5 h3 23.a7 h2 24.a8Q+ Kg1 25.Bc5+ Kf1 26.Qh1+, and White wins.

For the fourth situation:

A likely line of play is:

1.Bxa6 Ng4+ 2.Kd2 Nxh2 3.Be2 Kd4 4.c3+ Ke5 5.Bxh5 Nf1+ 6.Ke1 Ne3 7.Kf2 Nc4 8.b3 Nd2 9.c4 Kf4 10.Ke1 Ke3 11.Be2 Nb1 12.c5 b5 13.Bxb5, and White is winning.

An alternative for the fourth situation is:

This also looks to be good for White:

1.b4 Ne6 2.Bxa6 Nf4 3.Kf2 Kd4 4.Be2 e3+ 5.Kf1 Kc3 6.g3 Nd5 7.b5 cxb5 8.Ke1 Nb4 9.Bxh5 Nxa2 10.h4 Kxc2 11.Bg6+ Kc3 12.h5. Also winning for White.

In the struggle of unbalanced minor pieces it appears White has very good chances in many cases, but not all. The potential passed K-side pawn is quite important to this judgment. Also contributing to the evaluation is in this situation is the somewhat exposed advanced Black e-pawn.

A final summary of the Eson positions: The pure pawn endgame is won for White. Adding Rooks gives Black chances to find counter-play, but White can win with accurate play. Adding Bishops creates a very tough endgame battle where the better calculator has the chances. Adding Knights for both sides brings a similar conclusion. The Bishop versus Knight scenario – unbalanced minor pieces – favors White, particularly if White has the Bishop.

A couple of general rules are helpful. White should be careful about moving his Q-side pawns. With his a, b and c-pawn standing on their original squares and his King on e3, the Black King has a hard time getting at targets there. In so far as possible White wants to keep the possibility of creating a passed pawn on the K-side a viable option for as long as he can. Maybe the most important general point is: Endgames such this come up when the clocks are nearing the time limit. It is unlikely there will be sufficient time to calculate everything even though objectively that is what is required. The player is forced by circumstance to rely on his intuition. That does not mean making hasty, careless or pointless moves. Long ago Matt Katrein taught me that every move, even those made under great time pressure should have a point. Preferably it should be a threat, even a small threat is better than no threat.

The Albany Tourney Nears the Finish and an Endgame

The last local Club Championship is getting closer to a finish. If all goes as planned the very last game will be played May 7th for the Albany Chempionship. Wednesday evening just two make-up games were played:

Alowitz 1-0 Stephenson. This was some kind of mixed bag combining the Sicilian and the Ponziani, and that is no easy thing to do. By move 19 Alowitz had collected two extra pawns and was cruising towards a win. It took another twenty moves, but the issue was never in doubt, and Mr. Alowitz took the full point just after move 40.

Henner 1-0 Perry. The opening here was the Meran System of the Semi-Slav. They stayed in the book up to move 8 or 9. Then Mr. Perry improvised just a bit with the move order, and the game left the well known trail of the theoretical. Mr. Henner came up with an idea of clamping down on the Q-side since Glen did not get in the c6-c5 break when he had the chance. The Q-side situation was stabilized with White in the better position. Peter then switched to play on the K-side before he had all of his pieces in play. That didn’t work out too well, but with a too tempting offer of a Q-side pawn White obtained an advantage. Through the early endgame Mr. Henner’s edge grew until by move 31 it was near winning. Shortly thereafter Henner missed a combinational chance to finish the game. Time was getting short for both sides mostly because both sides were willing to play complex and testing lines. The notoriously difficult Rook plus the f&g-pawn versus Rook ending came about. According to Dvoretsky these endings are usually drawn. To do so you need a high level of endgame knowledge and some time on the clock to work out the details. That was not the case here, and Mr. Henner won in 56 moves. This was one of the most competitive and entertaining games of the Albany event – a close contest right down to the end.

The standings in the Albany event now are:

1-2 Wright 10½-1½ with Berman to play
1-2 Berman 10½-1½ with Wright to play
3 Howard 9-3 with Henner to play
4 Magat 8½-3½ with Perry to play
5 Mockler 8½-4½ finished
6 Denham 5½-5½ with Eson and Perry to play
7 Henner 6½-5½ with Howard, to play
8 Perry 4½-5½ with 3 to play
9 Lack 6-6 with Perry to play
10-11 Jones 5½-7½ finished
10-11 Northrup 5½-7½ finished
12 Alowitz 3½-9½ finished
13 Stephensen 1-12 finished
14 Eson 0-12 with Denham to play

The final standings in the Albany Championship are taking shape. It is Wright or Berman for the top spot. I am told unofficially that Tim Wright has the better tie-breaks if their game is drawn. The 3rd and 4th places will go to Magat or Howard depending on the results of their last games. Michael Mockler will finish 5th , no one below him in the standings can match his score. Steady and consistent play by the two top guys separated them from the field. Regardless of the outcome of their head-to-head encounter, both Mr. Wright and Mr. Berman can look back on a very fine performance this year.

Chuck Eson has had a tough tournament in this year’s AACC Championship. It has been difficult for him to score even one half-point. Mr. Eson has made progress in his chess notwithstanding his score. Today’s tidbit shows how close he came to making an upset in a recent game.

Magat, Gordon – Eson, Charles
AACC Championship 2013–14 Guilderland, NY, 09.04.2014

Playing against one of the highest rated contestants, Charles held his own up to about move 30. Gordon had tried out the Polish Opening; 1 b4, and 2 Bb2. Using good sense and careful play Eson kept the balance until he dropped a pawn around move 30. Even then he did not panic, and he was able to get enough activity for his two Rooks to almost balance the pawn minus. Black’s active Rooks induced White to trade down to one pair of Rooks. The game then reached the following position:

Eson Magat 1


So far so good. Neither side was in bad time trouble. Eson had about 30 minutes remaining on his clock. That probably is enough time to negotiate even a tricky Rook and pawns ending if you have correctly evaluated the position, and you know a few principles.

One of the rich veins of chess culture is the field of Rook and pawn end games. On the surface they appear to be fairly simple. Maybe they are to those gifted with great chess talent. For the rest of us, Rook and pawn endings are like icy roads to night time drivers; nerve-wracking and dangerous. The least bit of inattention can land you in a ditch.

2.Kf3 Rb6?

A brief conversation after the game was over leads me to believe Chuck had calculated things out for several moves and arrived at an incorrect evaluation. I believe he thought with the Rooks off the position was drawn.

Maybe the first principle of Rook and pawn endings is that trading off the Rooks is a critical decision. Once they are off the outcome depends on the resulting pure pawn ending. Having the Rooks on allow for such techniques as cutting off the opposing King from his passed pawns, creating threats to the opponent’s pawns and the defense from a distance of your own weaknesses. Without Rooks on the board everything depends on the pawn position and where the opposing Kings stand. So, deciding on a trade requires accurate calculation and visualization of the positions after the trade. Once the Rooks are off the die is cast, you can’t go back.

In training chess players pure pawn endings are often used as exercises to improve calculation. I think this idea was first brought to general popularity by various trainers from the Russian School. It is so because pawn endings are simpler than positions with several kinds of pieces and pawns. In them there are just two types of pieces that have to be manipulated; Kings and pawns, but in many practical positions fairly long sequences have to be visualized and compared to arrive at the correct evaluation. Studying pure pawn endings is a way a player can train himself to handle long strings of data, and to comparatively evaluate positions distant from the current board position.

Progress in understanding chess is not a smooth process. We individually move forward in fits and starts. Sometimes it takes a great disappointment to drive home a particular point. Today’s example was such for Chuck. He’d held one of the strongest of our Club’s players even for a couple of hours and even fought back from one error. Now all that is required is to get through a “simple” Rook and pawn ending and secure the draw.

With the game move Mr. Eson jumps into a lost pawn ending because he misread the resulting pawn endgame. He’d be better served by keeping the Rooks on with either 2…, h3; or 2…, a5. Both moves fit a general endgame principle; passed pawns must be pushed. This brings us to the question of the general evaluation of the position. Black’s two passed pawns don’t look particularly strong. They are isolated and lack immediate support. Their wide separation, oddly enough, gives them extra value. To stop and maybe to capture them will require both of the White pieces to take some action. The farther the Black pawns advance the greater the pressure on White to decided what to do. That fact should give Black time enough to: first, eliminate a White pawn, probably the c-pawn, and second, to set up a drawn position around the pawns on f4/g5/g6.

It is my suspicion that Black feared the line: 2…, h3 3 Rxg6 Rxc4 4 Rh6:

Analysis Position
Eson Magat 2

And Black may have been thinking the h-pawn is doomed. There is a trick here however. If Black sees clearly enough he can play 4…, Rc2!; and if then 5 Rxh3? Rc3+; when the White King ends up on h3 after the Rook trade.

Analysis Position
Eson Magat 3

That is outside of the “square” of the Black a-pawn. Then the moves: a7-a5,a4,a3,a2,a1 (Q) wins long before White can get his King and two connected passed pawns anywhere near making his own Queen.

So, the capture of the h-pawn on the 5th move is no good. What else can White do here? Push a “passer” of course. Since sending ahead the g-pawn, 5 g6, is easily met by 5…, Kg7; and after 6 Rh7+, it is clear the game will become a Rook+ 1 pawn versus Rook+ 1pawn ending that is drawn. That judgment leaves only 5 f5,

Analysis Position
Eson Magat 4

then: 5…, h2 6 Rh7.., and now Black has to find just the right move to make a fight out of it.

Analysis Position
Eson Magat 5

The natural 6…, a5; is wrong because: 7 f6! a4? 8 g6, setting up a mating position and forcing 8…, Kg8; then 9 f7+ Kf8 10 Rh8+, when White makes a Queen with check. Even improving slightly with 7…, Kg8; does little good after: 8 g6 h1(Q) 9 Rxh1 Rc6 10 f7+ Kg7 11 Rh7+ Kf8 12 Rh8+, Queening once again with check. Black’s best is 6…, Kg8!;

Analysis Position
Eson Magat 6

first forcing the g-pawn forward, and then chasing the White King to f6 where he obstructs the f-pawn; 6…, Kg8 7 g6 Rc5 8 Kg4 Rc2 9 Kf4 Rg2 10 Ke5? a5 11 Kf6 Rb2 12 Rc7 Rb6+ 13 Kg5, and the Black h-pawn Queens. Finding these moves over the board if you is difficult even if you have studied Rook and pawn endings diligently. White can bailout by giving up his connected passed pawns when Black attacks them with the Rook. White then captures the h-pawn, and since his King is closer to the a-pawn than is Black’s King, there little trouble blockading the a-pawn and then winning it. The bare Rook ending is obviously drawn.

3.Rxb6! axb6

Eson Magat 7

The situation is now very different. The passed a-pawn is gone. The h-pawn will be taken with relative ease, and Black is left defending a pawn down endgame.

4.Kg4 Ke7 5.Kxh4 Kd6 6.Kg4,..

Eson Magat 8

The h-pawn is gone, and the White King is ready to support the f-pawn’s advance. That will bring into being a passed g-pawn which can draw the Black King to it like a magnet. If the Black King goes towards the g-pawn White then pushes the f-pawn recovering it by virtue of zugswang then the White g-pawn occupies the Black King long enough for the White King to race over, clip the b-pawn leaving the White King ideally positioned to escort the c-pawn to the 8th rank.

6…, Kc5

It could be Mr. Eson calculated going this way would allow him to obtain his own passed pawn. The problem is the pawn is not far enough forward to offer adequate counter-play.

7.f5 gxf5+ 8.Kxf5 Kxc4 9.g6 1-0

Eson Magat 9

Mr. Eson resigned here. If the game were played out it might have continued as follows: 9…, b5 10 g7 b4 11 g8 (Q)+ Kc3 12 Ke4 b3 13 Qd5 b2 14 Qd3+ Kb4 15 Kd4 Ka4 16 Kc4, and the pawn falls in short order.

I have in my library five works that touch on Rook and pawn endings. The earliest is Reuben Fine’s Basic Chess Endings, 1941, and the most recent is Dvoretsky’s Endgame Manual, 2006. In none of these wonderful books was I able to find the piece of advice I gave early on in this article: In Rook and pawn endings deciding to trade Rooks is a critical decision. Memory is a slippery thing when you older, and I can’t recall where I first heard that fragment of chess wisdom, or from whom. For High Experts, Masters and titled players it is probably so obvious as to not need saying. For the rest of us it is something usually learned the hard way, as Mr. Eson did in this game.

More soon.

Some Thoughts on Rook and Pawn Endgames

Lately I have been writing about the Albany – Schenectady match, and there is still more to come. Today, however, we will take a break from the match and talk a bit about endgames. Wednesday evening the Albany Club met as usual. In the course of an evening of skittles play, I had the opportunity to discuss some endgame play with Chuck Eson and Mike Laccetti. Continue reading “Some Thoughts on Rook and Pawn Endgames”