The Longest Game of the Week

A while ago I used to run an endgame group at the Melbourne Chess Club. This group proved quite popular and had a number of regulars attending. I lead this group for about 2 years and in this time I learned a lot about endgames, and about players abilities and appreciation of this phase of the game.

The Melbourne Chess Cub Endgame Group was taken over by FM Bill Jordan after work commitments meant I could no longer run it, but I think Bill will not be running it this year, and I'm not sure if this means that the Endgame Group has stopped. If it has it will be a shame, as all club players would greatly benefit from regular endgame study. Putting an hour or 2 towards endgame study each week does such a lot for your chess:

- improves confidence
- builds technical skills
- increases resilience and resourcefulness
- develops planning skills, and prepares for transition from middle games
- builds awareness of the strengths and weaknesses of each piece
- helps with calculation skills

Add to this that virtually no club players are looking regularly at endgames, and it is easy to see that becoming any where near proficient at this part of the game will reap benefits, where players can maximise their results, winning games that should be won, holding difficult positions, and even turning around positions from lost, to level, to won.

One of the things I used to do was look at the longest game of the week in The Week in Chess (TWIC) magazine. TWIC is a chess news website which publishes a pgn of the games played that week. Using a database to sort and view the games, it is possible to keep up to date with all games played, and published, each week. I would sort the games by number of moves, and look at the longest, or at least the final part of the longest. Each week there are games played in interesting and important endgame positions to understand. Regularly there would be massive games that ended in pawnless endgames, or queen and pawn endgames. Take into account that the players may be tired from already having exerted 4-6 hours worth of effort into the game, and players can be forgiven for not playing perfectly. But it means the student of endgames has great material to learn from, if they're willing to put the effort in.

For example, the latest TWIC was issue 1111 published just yesterday (it is published every Monday in the UK, or Tuesday morning here in Australia). When sorting by number of moves, I find the longest was a 147 epic between 2 sub 2000 players from an open tournament in Italy. The player who was white won a piece early on, then some trades happen, and the leading side had some trouble because of opposite coloured bishops. Finally, the following position arose after 93 moves.

Black tried his best practical chance and played 93..Bxd6 forcing white to prove that he could win the dreaded KNB v K endgame. It was certainly worth a try, and black was probably heading towards this endgame for the past 30 moves. I've even used this myself, testing players to finish you off, when they are tired, and maybe only have 30 seconds per move to work things out. A couple of wrong turns by white here, and the game is a 50 move draw!

So that is something already learned. It's a good idea to work out which endings give you good chances of success, or some chances of survival, etc. Looking at endgames without pawns can be a great exercise, and even better if you have someone to play through them.

In the above game, after about 17 moves the black king had been forced to the edge of the board.

Now perhaps this drive could have been achieved quicker, but never mind, there are about 33 moves left to win. The first thing to understand is that black has wisely chosen to send his king to a corner that white's bishop doesn't control. White cannot checkmate by force in the a1 corner, the black king must be driven to the h1 corner, or the a8 corner. White understood this, but still couldn't manage to do it! From the diagram above play continued 111..Kc1 112 Ba2! black's king is driven across the board 112..Kd1

Now white started the correct procedure called the 'W' manouvre. White's knight moves in a W formation from c2-d4-e2-f4-g2, each time it hits the second rank, it aims at squares of the opposite colour to the bishop on the back rank helping with the drive of the enemy king. 113.Nf4! Ke1 This is not what you would necessarily expect. Normally, the king will try to sneak back to the safe corner, not voluntarily march towards the dangerous corner! However, by doing something unexpected white wasn't sure what to do, and floundered around for a while and after 124 moves, the same position as above was reached again! Obviously flustered and under pressure to complete the mate in the required move (and probably with an audience watching) the game continued 126.Nf4 Ke1 127.Kd3! Learning to use pieces in cooperation is a major strength of working on endgames. Here, white has to use all 3 pieces to achieve the win. 127..Kd1 128.Ne2!

White is now half way through the W manouvre  and as can be seen, the black king is being forced across towards h1.  After the moves 128..Ke1 129.Bd5 [129. Ke3 was better] 129..Kd1 130.Bb3+ Ke1 131.Ke3 Kf1, the following position was reached.

Unfortunately, this is where white's technique ran out. Following the W manouvre by moving the knight to f4 was mate in 9 with perfect play, and would have gained a victory within the allotted 50 moves. The technique is 132.Nf4! Ke1 133 Ng2+ [Completing the W manouvre] 133..Kf1 134.Kf3

Black's king is now stuck in the mating corner. 134..Kg1 135.Kg3 Kf1 136.Bc4+ [Shutting the door on escape] 136..Kg1

Ok, here's where a knowledge of the checkmate pattern comes in handy. Imagine white could put his knight anywhere on the board. The square he would choose is h3 with check, the king would have to move to the corner, and then the white bishop would checkmate on the long diagonal. So the way to achieve that is: 137.Nf4 [heading to h3] 137.Kh1 Bd3 [losing a move along the diagonal which controls black's f1 escape square, forcing the black king back to move to g1] 138.Kg1 Nh3+ 139.Kh1 140.Be4#

Here's the checkmate pattern, and white's bishop could be anywhere along the long diagonal from f3-a8.

Unfortunately, the player in the game went astray and didn't get back on track. The fact is that it isn't an easy technique without practice, and if the technique is learned and then not used for a long while, then it may be that the practical application of the technique is forgotten, or half remembered, or remembered but miscalculated or misinterpreted on the board.

I don't think it would do club players any harm to practice this endgame a few times until they have it off comfortably. Perhaps set up the initial position in this post, and play it a few times, even against a computer. Try to defend as well as win, and you will learn from your opponent. Once you think you can do it, give yourself 30 seconds per move to do it and see how it goes! Once that is not a challenge, you have probably mastered this endgame, and can be confident of converting it if you have to play it in a rated game. And don't forget to set your opponent's the challenge of converting this endgame. Perhaps they won't know how to do it, and you'll save a half point!