Think

It was my hope that after one full week of playing bishop endings against Stockfish on Chess.com, I would be able to report success. I had a completely winning position in the fifth and last exercise in the set yesterday morning when a single hasty move threw away the win. I can blame the distraction that caused me to look away from my iPad, but by now I should know to look at the board before moving when I look back.

This morning, the computer threw me several curves in exercise number three and I failed each time. This was one instance.

White to move

I played 1.a4?? and after 1...Bd8+ 2.Kc6 Bxa4, there was no way to prevent the loss of the g-pawn, too.

Later, I set the position up in the Stockfish app and won easily.

Several times this week, I have played 1.a5 in the position below. The resulting draw is a position that I should recognize by now.

White to move

Overworked Piece

I failed my bishop endgame this morning. It was number two in the series described in "The student should work this out". I was doing fine until I reached this position. Then, I went from winning to losing.

White to move

What would you play?

"The student should work this out"

In Chess Fundamentals by José Raúl Capablanca, the author frequently states that a given position, "should be worked out by the student", and other comparable phrases. This assertion vexes some readers and some coaches are loath to recommend the book. If the student does not understand something, how will they come to an understanding without assistance?

In some cases, especially with simple pawn endgames in chapter 1 of Chess Fundamentals, I believe that Capablanca provides the necessary resources in the explanations that he offers. Still, I have seen others struggle. They need more help. Capablanca recommends a teacher.

When Capablanca wrote Chess Fundamentals (he was checking the proofs in England in 1920 while negotiating conditions for the World Championship Match that took place in March and April 1921--see "WCC Havana 1921" in this site's index), there were fewer resources than we have today. I believe the standard endgame book then was Johann Berger, Theorie und Praxis der Endspiele (1890), a book in the German language that was probably not readily available or accessible to the majority of Capablanca's readers.

Today, however, we have an abundance of books, websites, YouTube, ready access to engines, and even tablebases. The danger to the student stems not from lack of help, but rather too much help. Easy access to answers cause many to fail to develop self-reliance. Capablanca's call to develop the habit of working things out may offer a more certain formula for strengthening a player's skills.

I often claim that I have zero natural ability at chess, but that I have learned a modest amount from books (see "My First Chess Book"). Many years ago when I started reading chess books, I found much that was confusing. The book I remember most clearly is Irving Chernev, The 1000 Best Short Games of Chess (1955). Often it was not evident to me why a game ended. Sometimes Chernev explained the reason, but even then I had questions. I sought answers to my questions by moving the pieces on a chess board, exploring variations. Certainly, with so little skill at the time, most of my fantasy variations were rubbish. Even so, the effort developed a habit of thinking for myself about chess.

The past few days I've found myself back in the classroom, seeking understanding. I started playing some bishop endgames against the computer on chess.com. Failure in my first effort got me hooked. My time struggling through five positions against the silicon monster engages me for hours every day. At the Learn/Endgames/Minor Pieces/Bishops tab on chess.com, there are five practice positions. To see the second one, you must checkmate the computer in the first. The first is simple and takes me from 12 to 24 seconds most times. 

The second one varies in difficulty, depending on how the engine plays.

White to move

Number Two

1.Bf5

I always play this move. It might be best.

1...h5 2.f3

Now the engine has played 2...a5, 2...Ke7, and 2...Kc6 most regularly. 2...Kc6 has proven most testing but I am reliably winning most of the time. Success grants me the opportunity to see number three.

White to move

Number Three

I found the pawn sacrifice quickly and now get to my seventh move in two seconds or less. Sometimes I reach this point with considerably less than two minutes total time elapsed for checkmating the computer in the first two and playing these seven moves.

1.c5 dxc5 2.Bc4 Bg4 3.Kxe5 Ke7 4.Kd5 Kxf7 5.Kxc5 Ke7 (the computer threw me a curve with 5...Kf6 at least once) 6.Kxb4 Kd6 7.Bd5

Black to move

The computer has played at least five different moves here.

a) 7...Bf3 proved challenging the first time I saw it. That line now often continues 8.Ka5 Be2 9.b4 Bg4 10.Kb6 Bh3 11.b5 Bg4 12.Ka7 and the computer must give up the bishop for my b-pawn.

Black to move

12...Kc5 has given me trouble when I am moving too fast, which the exercise encourages by tracking total time for the series and making me play simple checkmates all the way to the end. Sometimes I underpromote to practice bishop and knight or two bishop checkmates.

13.Bc6 Bc8

Here, 14.b6?? fails, but both 14.Kb8 and 14.e5 succeed.

I have seen most of the computer's responses several times, but this morning's effort presented me with a position I did not recall seeing before. 

b) 7...Bd7 8.Ka5 Kc7 9.Kb4 Kb6

White to move

10.Kc4!

There is more than one way to get rid of Black's bishop.

10...Be8 11.Kd4 Kb5 12.e5 Kb4 13.Ke4 Kc5

White to move

14.Bg8 Bc6+ 15.Kf5 Kb4 16.Kf6 Kc3 17.Bc4 Bd7

This move is Stockfish surrendering it seems.

White to move

18.e6 Bc6 19.Kf7 

Black must give up the bishop for the e-pawn. A few moves later, I underpromoted the b-pawn to a bishop. When I was nine or ten moves from mate, I made a hasty move that stalemated the engine. Now, I must start over at number one. I've played number five half a dozen times and have some ideas how to proceed. My current problems are two: 1) the engine lures me into a pawn ending that is lost foe me, and 2) I fail two, three, and four too often.

Have you tried these exercises? How did you do?

Checkmate Challenge

My online chess camps this summer have been competing with one another in the checkmate challenge. The challenge consists of 30 minutes of solving as rapidly as possible checkmates in one move. I use a database of 101 exercises that I downloaded from the web so long ago that I no longer know where or when. I believe it may have been Gunther Ossimitz's now defunct chess site (see "Gunther Ossimitz PGN Files").

The first ten exercises are mate with a rook. Then there are ten with a bishop. These are followed in succession by mates with a knight, a queen, and then pawns. Three involving en passant captures usually slow the students down, and I often need to explain the rule.

White to move

Mate in one

Some pawn mates are the pawn delivering the final check. Often, the pawn promotes to a rook or queen. In one case, the pawn must become a knight.

After the en passant mates, there are some double checks mixed with other themes. The exercises grow slightly more difficult. I like the structure of the database, which resembles the structure at the heart of the work of Viktor Khenkin. See "Two Old Books (and one new)" for more about Khenkin.

Many of the exercises are composed, and the early ones have few pieces. Some are from familiar games, such as Morphy's Opera Game. 

My students have included a handful of strong youth players, but most are beginners. Some of them have been playing chess a week before camp begins. I ask the students to shout out answers or type them in the Zoom chat. Speed matters. Wrong answers are ignored if the right answer comes fast enough.

Many groups have solved 60-70 of the exercises in the 30 minutes. Three groups have solved all 101! Today, a group solved the whole set in 27 minutes. Last week, a group did so in 26 minutes. One group with some skilled players whipped through the entire set in 18 minutes.

These are some of those that slow the students in their quest. It is always White to move.

Coaching: Constructing a Lesson

I had been coaching youth chess for a couple of years when I decided to keep a clear record of what I did each week with the students at the elementary school where my youngest was no longer a student. While he was enrolled there, I was a parent volunteer helping with the after school chess club. When he moved on to another school, I returned to his old elementary as a paid chess coach. The year was 2004. Chess club started in late September.

I began the year with a position from a game I had played a day or two earlier on the Chessmaster Live server. Although Black had a one pawn advantage in a king and pawn endgame, it should be drawn with correct play.

White to move

My opponent erred with 46.Kc5?? We both promoted pawns, but then my opponent blundered theirs away. Had we continued in a queen ending, White might have held out long enough to run me short of time. Black was technically winning with QPP vs. QP and Black's king closer to the pawns.

Instead, play might have continued:

46.Ke4 Kf7 47.Ke5 Ke7 48.Kd4 Kd6 49.Ke4

Black to move

Here Black can err with the same flawed idea that White pursued in the game: 49...Kc5?? Correct play would be either 49...Kd7 or 49...Ke7 and a technical draw.

In the lesson, I sought to introduce to the students the concept of opposition in pawn endings. I followed this position from a recent game with a number of other positions, mostly composed, that had fewer pawns and presumably simpler continuations.

The school chess club ends the year with the Washington State Elementary Chess Championship in April. The last session before state, I showed the students a pawn ending that had been played at the Spokane Chess Club the week prior.

Black to move

After the game ended as a draw, I spent some time with others at the club, including the one who had Black, arguing about whether Black had a win. The next day, I checked some of our ideas on my computer. If we had the same argument today, someone would pull up Stockfish on their phone or a tablebase site and end the debate. My phone then was a RAZR flip phone. It's chess app was a version of Chessmaster that I could beat on its top level.

Had any of us read with sufficient comprehension José R. Capablanca, Chess Fundamentals (1921), we would have known how to use triangulation to secure the win after 1...g5. I had nearly forgotten this lesson until I was playing through an ending presented in C.G. Van Perlo, Van Perlo's Endgame Tactics, new, improved and expanded edition (2014). The ending came up in my study as I was comparing endgame books for "To Know a Position", which I wrote last weekend.

For the children headed to state, I hoped to reinforce some lessons I thought they should have absorbed in September. It is always hard to measure results, but both the elementary team (grades 1-5) and the sixth graders from the middle school brought home trophies--individual and team. School officials perceived me as a successful coach and I have continued to coach youth players.

The lessons I employed that first year as a chess professional at an elementary school were compiled into a booklet that I printed at the school and then had bound at Kinkos. I have learned a lot in the years since. My skills as a teacher have improved. My chess playing ability also rose. I have also learned to print my lessons through Amazon, where I get better quality binding at lower cost, and also am able to make them available to others.

Grigoriev Study

As mentioned in "To Know a Position" yesterday, a 1921 study by Nikolai Grigoriev appears in several books, especially Dvoretsky's Endgame Manual and Alex Fishbein, King and Pawn Endings. However, I did not find the study with that date in Harold van der Heijden, Endgame Study Database VI. Rather, the position is flipped and the date given is 1924. Otherwise, it is the same position.

White to move

White must play precisely to draw.

To Know a Position

Two pawns against one when the stronger side has one that is both passed and protected is generally a win. There are exceptions when the pawns are too close to the edge of the board with one of the pawns (stronger or weaker side) on its starting square. Having a pawn on a rook file also limits the stronger side's chances. The position of the kings often proves critical.

This position, ending no. 107 in Pandolfini's Endgame Course (1988), also arises while playing through examples 8 and 9 in José R. Capablanca, Chess Fundamentals (1921). Josef Kling and Bernhard Horwitz, Chess Studies, or the Endings of Games (1851) presents a study that reaches this position after the first two moves.

White to move

Sacrificing the forward pawn allows White to win Black's pawn after a few moves using opposition and outflanking maneuvers, resulting in a standard position that Bruce Pandolfini covers earlier in the book (endgames 60-62).

The diagram position is important enough in Lev Alburt's view that it appears in Chess Training Pocket Book: 300 Most Important Positions and Ideas (1997). Alburt writes that to become a strong tournament player one needs to know only 12 key pawn endings, promising that the right 12 are in the book (9). A similar position appears in Rashid Ziyatdinov, GM-RAM: Essential Grandmaster Knowledge (2000), but did not make the cut for Thomas Engqvist, 300 Most Important Chess Positions (2018).

Lest there be any doubt about the practical value of understanding how White should play, consider the ending from Polgar,I.--Ciocaltea,V., Baja 1971, which is presented in C.G. Van Perlo, Van Perlo's Endgame Tactics, new, improved and expanded edition (2014).

White to move

54.Kg3

54.Kg5 fails to make progress, Van Perlo notes.

54...Kf7 55.Kf3 Ke7 56.Ke3 Kd7 57.Ke4 Ke6

White to move

Flip the side to move, remove White's f-pawn and place a Black pawn on f6, and the position is identical to one that appeared at the Spokane Chess Club in April 2005 between two of the group's strongest players. Black pushed the unopposed pawn and the game was dead drawn, yet the game continued for many moves more because the weaker side has very little time left on the clock.

58.g5!

"Depriving the black king of some vital squares in the tempo battle" (Van Perlo 25).

58...Kd6 59.f5 Ke7 60.f6+

Now White has a passed pawn.

60...Kf7 61.Ke5 Kf8

We have reached the position at the top of this post.

62.f7

Black resigned.

We can move the initial position right or left, up or down. In most cases, the stronger side has an easy win.  Sacrifice of the forward pawn is the easiest method when the blocked pawn is on the fifth rank, but this fails when the pawn is further back.

Two key positions are analyzed in Johann Berger, Theorie und Praxis der Endspiele (1890), referencing the earlier work of George Walker, A New Treatise on Chess (1841).

No. 483

Das Spiel ist leicht zu gewinnen

It is an easy win, regardless of who has the move, Berger explains.

No. 484

Weiss gewinnt nur mit dem Zuge.

White wins only with the move.

Mark Dvoretsky, Dvoretsky's Endgame Manual, 5th ed. (2020) offers one full page of analysis with one winning and two drawing positions, plus a well-known 1921 study by Nikolai Grigoriev earlier in the chapter. Grigoriev's study, "underscore[s] that a system of corresponding squares certainly does not have to always be 'straight line', as with the opposition. Each case demands concrete analysis" (23). Alex Fishbein, King and Pawn Endings (1993) offers two positions, both of which are Grigoriev compositions. Reuben Fine, Basic Chess Endings (1941) has seven diagrams over a little more than three pages. Karsten Müller and Frank Lamprecht, Fundamental Chess Endings (2001) offer two positions from games, one winning and one drawing. Comparable positions do not appear in Jeremy Silman, Silman's Complete Endgame Course (2007), unfortunately.

Paul Keres, Practical Chess Endings (1974) offers the most thorough discussion with 11 diagrams over six pages. He begins with Berger's no. 483 (see above).

After nine moves in Keres' analysis, the following position is reached.

White to move

"White would spoil everything with 10.b6?" Keres notes (34).

10.Kd5! Kc7 11.Ke6 Kb6 12.Kd6

This outflanking idea is another standard technique.

12...Kb7 13.Kc5

Keres then turns to the most important technique for endgame book authors: shifting the position. Shift everything up on rank and White can no longer win because of the same stalemate threat noted after 10.b6? Shift one square down, and White still wins. Two squares down and it matters who has the move. White must have the move to win.

Moving the position one file to the right and one rank up, White wins, even when Black has the move.

Black to move

1...Kd6 2.Ke4 Ke6 3.Kf4 Kd6 4.Kf5 Kc7 5.Ke6 Kc8

White to move

6.c7!

The pawn must be sacrificed, as we should know from Capablanca and Pandolfini.

Continuing the shift to the right and this time down a rank, Keres sets the kings in a manner optimal for Black. After a few moves, the following position is reached.

White to move

To secure the win, White must use threats of moving to either side of the mass of pawns to perform the opposition and outflanking maneuvers.

6.Kc1!

Working out the rest is good training. Keres book is helpful, too. It was when I reached this position in my reading yesterday morning that I started playing against Stockfish from this position and then many other variations with the same arrangement of pawns across the range of possibilities.

Shift the starting position one rank down and this resource no longer exists for White.

Knowing a position well is knowing an arrangement of pieces that may be, in fact, many hundreds of possible positions. Some of these have arisen in my games in the past and can be expected to do so again. 

The Quiet Move: Origins

Quiet Move: A move which is not forcing, i.e. a move which does not directly attack or capture an enemy piece. In tactics problems, a quiet move is often used to control important squares or guard your own pieces from future capture, before launching a more direct attack in subsequent moves.
"Chess Tactical Motifs and Themes", Chess Tempo

The quiet move in chess first appeared in Medieval problems where checkmate must be accomplished in a set number of moves. These contrasted with older Arabic problems requiring a series of checks to achieve victory because the other side threatened mate on the move. In the Arabic problems, forces often were balanced. Not so in these early European compositions.

Consider this position that is said to have occurred in a game played sometime in the ninth century. An Arabic manuscript examined by H.J.R. Murray, author of A History of Chess (1913), asserts "this happened to Abū’n-Naʿām, and he used to boast of it" (309).

Black to move

Black must sacrifice the knight and one of the rooks in order to checkmate White with the other rook on the third move.

Contrast that problem with a composition appearing in a collection of manuscripts dating as early as the early fifteenth century. The problem specifies that each White piece moves once and mate must be delivered on the third move.

White to move

Medieval compositions were not considered "cooked" when there were multiple solutions, as this one. That standard is modern. Also violating modern standards, the mate can be accomplished in less than three moves.

According to Murray, the the pseudonym Bonus Socius was used by the compiler of an important manuscript of a group of several that all were copied from another work (618). This problem appears as no. 62 in the MS. The same arrangement of pieces shifted back one rank appears as no. 35 with the same instructions, but an altogether different set of solutions.

White to move

There are eight first moves where the solution meets the specified criteria and two more that do not, although mating on the third move.

Murray writes that this work represents an effort to organize existing material according to the number of moves in the solution. He credits these simpler problems with exploring "the powers of a single piece, or the combined powers of a few pieces" (649).

This one, for example, could serve to teach beginners the habit of cutting off the king from the center of the board. That rooks can checkmate a king in the middle of the board comes as a surprise to beginners. The restraining power of the king is an important element.

White to move

After the only move that leads to the required mate in four, White has several ways to proceed. Murray's solutions from the Bonus Socius MS is 1.Rg4 Ke3 2.Rc3+ Kf2 3.Rd3 (or a3, b3) Kf1 4.Rf3#.

Another set of manuscripts expand upon Bonus Socius. The most important of this second set is known as Civis Bononiae, also a pseudonym. There are 194 chess problems in the former and 288 in the latter. One from this group strikes me as particularly useful for teaching elementary skills in the use of the king. Although there is more than one way to achieve checkmate, achieving it according to the specified criteria requires opposition and outflanking--a standard technique in pawn endings, but in this case it is a mate with rook and king against a lone rook.

White may move the rook only once. Mate in 12 or less.

White to move

Playing against Stockfish 15, I delivered checkmate on the ninth move.

Solve This!

The position below arose in Stean,M.--Mestal,A.J., Canterbury 1973.

Black to move

Michael Stean analyzes this game in Simple Chess (1978), where he gives great credit to his opponent's play.

The tag, "Solve This", links to posts on Chess Skills where the move played is not part of the post. They are offered to encourage engagement.