A Small Miss

In a tournament game a couple of weeks back I reached the following position as White, with Black to play:

Black to move

Here Black played 1 … Kc4 and after 2 g4 the game was quickly decided. I had already seen this breakthrough during the exchanges leading up to this position, so it was easy to play.

But I believe Black had a better practical try, namely 1 … f5. Surprisingly, White has one and only one path to victory, and he can even lose if he gets very careless. I will put the solution below the fold.

More endgame precision

Was looking around some files and stumbled upon an Anderssen-Kieseritsky rook endgame from London 1851. I then found something that surprised me. When I finally understood, I thought it might be worth sharing. The following position is from analysis following White's 53rd move. Blacked erred in playing 53 … Kf4. Had he played 53 … Kxh4 the following position would have eventually been reached:

Anderssen-Kieseritsky

London, 1851 (m1.2)

What surprised me when I saw this position is that Black has one and only one winning move. I thought it was either a draw to begin with, or that any reasonable move by the Black rook on the c-file would win. Not so.

I will put the rest below the fold for those that want to work it out on their own.

A Study

Via the Chess Unlimited twitter feed, I was directed to a reddit involving a problem posted on LiChess, apparently published in 1928 by a composer named Birnov. Here is the problem.

Birnov (1928)

White to play and draw

I saw a solution pretty quickly. Knowing that it is a composed problem means that everything on the board has a purpose, which directs one's attention to the e-pawn. So e2-e4 is likely to be involved in the solution. From there the rest is easy. But then I thought I saw a second solution. I am sad to report that I did not figure out why only one of the solutions is correct. I will post the answer below the fold.

Testing the Test

Today I stumbled across the website Elometer.net. Here's how they sum it up:

Dear chess friend,

we kindly invite you to the Elometer if you are interested in a quick measurement of your chess skills!

To obtain a psychometrically founded estimate of your playing strength based on a comparison sample of previous participants, all you have to do is to solve the following 76 chess problems.

...

In all of the following positions, you are playing White, and White is to move. Note, however, that the best move is not necessarily a brilliant one. Some of the following positions are rather simple, but some are very tough. Just try to choose the move you would play in a real game, and refrain from only trying to find a combination; this will yield the most precise estimate of your Elo rating.

The Elometer is part of a research project conducted by Birk Diedenhofen and Jochen Musch from the University of Duesseldorf, Institute of Experimental Psychology, Germany.

It took me a couple of hours to get through the test, but that was with stops due to supervising a 7 year-old. I would recommend doing it on a desktop computer instead of a tablet (as I did) or a phone, as that will minimize finger slips.

How'd I do? They estimated my rating at 2200 with a 95% confidence interval of about 2065 to 2330. Oops! Well, it isn't 100% confidence, is it? The format of the test also kept me from my usual errors of transposing moves and dropping random pieces, so I know where the 400 point difference is, alas.

Also of interest in a 48 position endgame test. All you have to do is state if White, with the move, can force a win in the given positions. I got 39 of 48 correct. Give them a try, and let me know how you did in the comments. The endgame tests include lots of basic stuff as well as more advanced material, such as the famous Cohn-Rubinstein ending from the St. Petersburg tournament of 1909, and Euwe's famous 1940 study of the pawn skeleton from the Ruy Lopez Exchange Variation. (Yes, I recognized both of those without having to look them up, although I did need to double check dates.)

I found the endgame studies interesting enough to enter them into a ChessBase database for further testing in the future. Aiming for no less than 48 out of 48.

A couple of endgames to ponder

First, a puzzle from LiChess, who take their puzzles from users games. Find the best move for White.

The answer later.

Second, another position with White to move.

Black's ...Kd7-e6 targets White's h-pawn. White clearly has to move his king now. Which way should the White king go?

King & Pawn Endgames: Answer Time

The other day I presented this position and asked a couple of questions.

St. Amant - Staunton, Paris 1843 (m11)

After 43 Kc3-c2

What is the proper assessment of the position? How many acceptable plans does Black have?

The answers are that Black has a won position, and he has exactly ONE winning move. In fact, all moves other than 43...Kxd4 lose. Staunton chose to play 43...Ke4 and duly lost. I can only assume he originally intended to follow-up with 44... Kf3 and then promote his e-pawn. However, that plan does not work because of St. Amant's move, 44 Kd1, and now the White king will blockade the Black e-pawn, whilst Black has to run down White's d- and g- pawns, allowing White to easily win Black's queenside pawns. I don't know anything about the particulars of this game other than that Staunton was leading the match by the score of 8+, 1-, 1= prior to this game. Perhaps Mr. Staunton merely had a lapse.

But the winning line isn't 100% straightforward, either. Black still has to worry himself about that pesky g-pawn. How can he catch it and not let White clean up all of Black's pawns? The answer is that he can't. But that he doesn't have to, because White can be deflected from his own queenside pawns by the Black e-pawn, and that plus the unfortunate (for White) geometry between c4, e2 and g8, gives Black a winning position.

After 43...Kxd4 White's best is 44 Kd1 (If 44 g4, then Ke4, and Black will be able to round up the g-pawn and return in time to guard the e-pawn, and if Black tries breaking through on the queenside, Black's e-pawn queens far too quickly for White to get his pawns up the board):

 After 44 Kd1

44      ...     Kd3

45     g4      e2+

46    Ke1    Kc2!

 After 46 ... Kc2!

47    g5          

(47 Kxe2 transposes to the main line)

47      ...     Kxb2

48      g6     c3     

49      g7     c2     

         50      g8(Q)      c1(Q)+

51     Kxe2  Qc4+

After 51 Qc4+

And there's an unfortunate triangle for White. Another interesting fact is that all but one of Black's moves from 43 to 50 are only moves - anything other than the move given above loses. The one exception is that Black can play 44...c3 and still draw. These endings are precise!

...

On a programming note, I hope to return to semi-regular blogging, at the very least. Soon we will have final news on the on-going Club Championship, and there may even be a game or two worth noting.

King & Pawn Endgames

Below is a position from the eleventh game of the 1843 match between Pierre Charles Fournier de Saint-Amant and Howard Staunton.

St. Amant - Staunton, Paris 1843 (m11)

After 43 Kc3-c2

What is the proper assessment of the position? How many acceptable plans does Black have?

I'm not going to give the answers just yet. This is largely because I don't quite feel like typing it all up, but if I don't post something, I'm not going to get back to writing! But this is also a chance for the reader to work out the problem for himself - practice makes perfect, after all. So, have at it, reader, and enjoy!

Tricks with Knights & Rooks

In an earlier post, I had a messy position from the Eickelman-Slade game in the club championship. Working my way through it I came to a position wherein White has to win with a Rook, g- & h-pawns versus a Knight & h-pawn. It turned out that winning the position wasn't too hard, but it put me in mind of the famous Em. Lasker-Ed. Lasker game from the New York, 1924 tournament. Looking up that position in my trusty copy of Fundamental Chess Endings, I came across the following study.

J. Moravec, La Strategie, 1913

White to move and win

This will eventually resolve itself into a R vs N endgame. If the knight is close to the king, the defending side can often draw the game, as in the Lasker-Lasker game above. However, sometimes the knight can be separated from the king, even when they're close together at the start. To see some endgame magic, check below the fold for the solution.

Wow.

From Rashit Ziatdinov's Facebook page: White to play and win. The first move is obvious. After that it gets hard! This one requires a little technique, as well. Unfortunately I don't know the origin of the study, but it's a beauty!

Show you drawing zone? I show you drawing zone!

After the simul on Saturday, I played a few blitz games against Hleb. The results were as expected. However, in analysis after one of the games, we determined that we could have reached a Q+P vs Q ending, with Hleb having the extra pawn - on a rook file.

For some reason or another I had been looking at that ending a few months back, and I remembered that the "drawing zone" could be surprisingly large for the weaker side. I couldn't prove this, though, as (a) I'm not good at this kind of technical exercise, (b) I didn't actually remember anything I had (maybe) learned, and (c) Hleb outrates me by 400 points. But I kept repeating that "There's a drawing zone!" and Hleb would reply "Show me drawing zone!"

Now that I'm home and can look things up, I will show you drawing zone! Muller & Lamprecht's Fundamental Chess Endings shows a couple of examples.

Q + a5P vs Q

Q + a7P vs Q

In both cases, if Black has the move and his king is on one of the squares highlighted in green, it is a tablebase draw. But if you want the proof, you'll have to look it up elsewhere, or investigate the positions yourself!

I will add this, however: in the case of the a7 pawn, according to Muller & Lamprecht, "... h8 is a draw for specific tactical reasons, and does not constitute a drawing zone in the normal sense."

...

It seems that this ending had not merited much attention from GMs in the past. Reuben Fine in Basic Chess Endings (1941) simply states, "A RP draws just as easily as a KtP: once the Pawn gets to R7, the K to R8, perpetual check is unavoidable."

Paul Keres in Practical Chess Endings (1973, West Germany) Was slightly wordier in his exposition: "We shall not examine positions involving the RP, which only offer White slender winning chances. The reason for this is clear: White's king has difficulty escaping checks."

And of course, all of this reminded me of an anecdote from a tournament played over 40 years ago. In the 1975 Hoogovens Tournament in Wijk aan Zee, the following position was reached in the seventh round:

Jan Smejkal (2600) vs Walter Browne (2550)

 After 57 ... Kc6-b5

I'll pick up with Kavalek's notes, from R.H.M.'s tournament book:

Here the game was adjourned for the second time.Walter reportedly said that he did not know how Smejkal could advance his pawn. But what seemed hard for the American was easy for the Czech. I recommend that you be patient and watch Smejkal's winning technique. It looks rather convincing, even though according to some opinions Black could have had good drawing chances.

A week after this tournament was over the Yugoslav Grandmaster Ljubomir Ljubojevic gave a clock exhibition in a small Dutch town. After five hours one game had to be adjudicated. Each player had a queen on the board and Ljubo had one pawn: it was a KRP on the fifth rank. The result was a draw but somebody pointed out that the same position was won by Smejkal against Walter Browne. The master who called Ljubo's game a draw said, "But that's impossible! How did he win it?"

So, some practical difficulties exist!

Incidentally, Kavalek's book for R.H.M. on the 1975 Wijk aan Zee grandmasters tournament is my favorite tournament book of all time. Eleven of the sixteen players contributed some notes to the games, and Kavalek annotated a great many others. But Kavalek's notes before each game are what really make the book a winner, as it gives a good "slice of life" look at such a tournament. If you come across a copy somewhere, I'd recommend picking it up.

Fun with pawns!

The pawn looks unstoppable. Can you find the impossible defense? pic.twitter.com/RmcqY4IJpP

— Johan Salomon (@JohanSalomon) August 26, 2016

A very interesting position from the club tonight

A slow night at the club this evening, but at the end of it Jim McTigue and I played a game in 25 minutes with the usual five second delay. (Actually, we turned the clock off when we were both down to a couple of minutes and had a rather complicated endgame on the board, well before the position below.) The whole game was interesting, and I'm sorry I didn't record it. I may try to put more of it together tomorrow, but I'll probably forget most it by then.

That said, I did manage to get a couple of pictures of our hard fought endgame. Here's the most interesting position.

McTigue-Durham

Black to play

Can Black save the game?

I'll try to get back to this in a few days, or less.

Karsten Müller is at it again.

Karsten Müller presents a new endgame position for our edification.

Lissang, C.2295–Al Hadarani, H.2280

White to move. Can you find a line that draws?

Solution at ChessBase: Taking the shorter route.

Queen plus pawns versus Rook plus Bishop plus pawns, Part 2 [UPDATED]

Update is below the fold.

In the previous post I presented an endgame study. The solution is below the fold. First a digression on this type endgame. The original position was

It was White to play, though that's probably not critical for assessing the position. We thought this was probably a draw, as it looks like Black should be able to construct a fortress easily enough. But we decided to consult Garry's 2003 revised edition of Reuben Fine's Basic Chess Endings. The section "Queen vs. Rook and One Minor Piece" said the following:

Without Pawns the ending is a draw, though it is to be expected that there will be problem positions where one side or the other may win.

With Pawns, the Queen is equivalent to R+B+P. If the Pawns are even, the Queen will win (though not without difficulty); but R, B and two pawns are required to conquer the Q.

Where the pawns are even, the win is easier for the Q if they are not balanced. For then the superior side will be able to set up a passed Pawn and capture one of the opponent's pieces or tie him up so badly that some other part of the board will be defenseless.

He then follows it up with three examples, all of which have pawns on both sides of the board, or asymmetric pawns. Thus they were all useless for properly assessing our situation.

When I got home, I checked Muller & Lamprecht's Fundamental Chess Endings, but that book was silent on the issue, as were Dvoretsky's Endgame Manual and Paul Keres's Practical Chess Endings. Fine states that this is a win, and in 2003 Benko agreed with him. No one else says anything about it. (If anyone can consult Averbakh's endgame encyclopedia, or something from Informant, let me know.) So how to go about winning a position like the one above? I have no idea if the R+B side plays correctly. Feel free to give it a try, and add any research in the comment section below. Alternately, present it at the club. But this one is a bear.

I will also look in my database for similar positions, but not tonight as it's already passing 2am. Maybe Paul or Connor will do it for me!

Now for the solution to the study I gave in the previous post.

Queen plus pawns versus Rook plus Bishop plus pawns, Part 1

Garry and I played a few games tonight and the ending with Queen & three pawns versus Rook, Bishop & two pawns came up in a game. The Queen & pawns won easily enough. So we started looking at the ending withjust two pawns for the side with the Queen. We arrived at this position:

White to play

I thought I had a winning plan with 1 Qg8+ Ka7 2 Qc8. Was I correct that this was a winning line? If not, can White's play be improved?

I will give the answers in the following post.

Endgame technique, this is wondrous strange!

I'll end tonight by mentioning a couple of recent firsts for me in endgame play.

As some of you know, Connor especially, I am not a fan of technical endgames. I find them boring, and thus I haven't learned nearly as much as I should have. For example, the Lucena and Philidor positions are outside of my knowledge base. I defend this ignorance first by stating that I am just a Class B scrounger, and second by mentioning that R+P vs R endgames have almost NEVER come up in serious games for me. In fact, I can only find one R+P vs R endgame in my database of tournament games, and in that case it lasted for one whole move before a draw was declared. Not only that, but it was a rook pawn anyway, so there are less things in my heaven and earth than are dreamt of in Lucena and Philidor's philosphies.

(That said, after discussing the matter with Connor, the [radio edit] R+P vs R ending came up twice in three games with him. But I think those may be the only two times the ending has even come up even in skittles games for me.)

Nevertheless, in the last week I've had two nice moments wherein I demonstrated actual technique. First I had a game in which I had a R+g-pawn vs a R+f-, g-, & h-pawns. I managed to bamboozle my opponent into exchanging g-pawns, and then held the draw with R vs R+f-and-h-pawns. It was pure book, baby!

And then the other night I managed to win a Queen vs Rook ending, no pawns on the board. Sure, I did catch my opponent with badly placed pieces, but hey, I won the thing!

Maybe there's something to this technique jive after all.

Lost in translation

I stumbled upon a slightly interesting endgame position analyzing one of my games from the USATS. The position in question never even came close to appearing in the game, and the position I'm presenting below has been modified further for clarity's sake. (I'll be using the Nunn Convention for my analytical symbols.)

Analysis position

White to move & draw

The important line runs 1 Ke4! Obviously White's king has to get within the square of the g-pawn. Kf7 The clearest line. 2 Kd3! White remains within the square of the g-pawn. g3 3 Ke2! e4 Or 3...g2 4 Kf2! e4 5 Kxg2 Kf6!= 4 Kf1! 4 Ke1? throws the game away. Kf6 5 Kg2! Kxf5 6 Kxg3!=.

What makes this slightly interesting (but just slightly) is that, as in so many cases, a slight shift changes the assessment completely. If all the pieces were moved down one rank, Black wins easily.*

Analysis position

White to move now loses

Here 1 Ke3 Kf6 2 Kd2 takes the White king out of g-pawns square.

Endgame books are littered with an uncounted number of examples where a shift by a file or rank makes all the difference in the world. And these things really do come up in our own games fairly frequently.** Being able to spot such differences several moves ahead in your analysis, and assess them correctly, will earn useful half-points over time. 

* In the original analysis position, moving everything one rank up would have given White good winning chances in a Q+pawn endgame.

** Except for rook & pawn versus rook endgames. For all the emphasis placed upon them, those pretty much never come up, at least in my experience!

Not a proper problem

The situation below is simple enough: White to play and win. White has nine possible moves, several of which win. The rest do not.Your mission, should you choose to accept it, is to find the winning plan.

White to play: Find the winning plan

8/pp1k2b1/8/3PPK2/6P1/PP6/8/8 w - - 0 1

This isn't a proper problem, but it is a good exercise. I reached this position playing out a position from my game against Jim from last Saturday. My opponent, as usual these days, was Stockfish.

Now for a subtle, confounding hint: It appears that if Black's bishop were on h8 instead of g7, then the game is probably a draw. I'm still trying to work that out, as well as the whys and wherefores. I plan to post the answer next week sometime. Hopefully I'll understand the position a bit better then, else wise it'll be more of a variation dump. Also, I hope I'll be done with this game by then!

A Nasty Problem: Solution Time!

Last week I published the following nasty problem:

V. Halberstadt, 1938

White to play and win

4N3/3nP3/1k6/8/1K6/8/8/8 w - - 0 1

I had found the problem looking through Karsten Muller & Frank Lamprecht's Fundamental Chess Endings, where they used this a five star (extremely difficult) exercise.

(Solution below the fold.)

Games from the Clubhouse: Endgame Edition

A few months back Paul and I played a game that became a topic of lengthy debate afterwards, with Garry also contributing. At some point, the following position was reached, though I have no idea how we got here from the game.

Paul - Todd (analysis)

 

Black to move

We spent a lot of time looking at this particular position. I didn't try to recreate the analysis the night I put this in my database several months back, and I'm certainly not going to try to do so now. We eventually settled on 1 ... Qg5 as the best opening salvo for Black.

But then Paul ran it through Komodo 8 and "the cloud" for an entire night. The best move according to those worthies is 1 ... Qg7. White's problems are that Black's h-pawn is too fast, and if White removes the c5 pawn his own King becomes too exposed - and maybe even if White doesn't take the pawn. Evals ran from -1.50 to -1.80.

It seems like these kinds of very tricky/trappy complex endgames come up more often now in the games I see at my level than in the past. Perhaps it's merely a function of where & who I'm playing (and see play) these days, or maybe I'm just better at finding this stuff now. Still, I don't remember seeing so many complicated endgames in the past.

(Am I the only one that thinks of Aristophanes every time "the cloud" gets mentioned? Yes? Nevermind then.)