An end to endings

 Learning endgames is hard. Just like exam revision, a passive read-through of the textbook won't cut it. In fact, lately it's revision methods that I've been recycling. Armed with an A4 notepad, a pen and an endgame encyclopaedia,* I have looked at some positions of interest, and summarised them in my own hand. It's easiest to remember your own words, but above all I aim for concision - that way I can identify what's truly important. Finally, I conclude in prose, not a 15-move variation. I've done this for a few endgames (trying to do everything in a few weeks would be pointless) and perhaps an example will be of interest: the idiot's guide to Rook and 4 vs Rook and 3 with all the pawns on one side. After that, the reader will finally be free from my endgame wittering!

 My chosen endgame isn't noted for high-octane thrills, but is of great practical importance. It can easily arise from a minority attack (Exchange QGD, Caro-Kann, etc.) and is well-known as a draw where the defender (henceforth always black, as per convention) must do some work - so both sides may aim for it. All told, budding Arkellites must know this endgame! My 3½/4 score** belies my total ignorance, but now I understand it better, and hopefully I can pass that on.

The Basics: With all pawns on one side, there is no passed pawn - obviously the main winning plan is to create one. But with the defender's King nearby, and Rooks on the board, this may not be enough - all else being equal, it is a theoretical draw. Material and positional advantage is needed for victory. But the attacker has lots of ideas, and passive defence is not sufficient. There is a lot of play left in the position.

Foundational Knowledge (3 vs 2): I can't cover pawn endings and KRP vs KR here, but they must be learnt: how else can you know what to aim for? But a few words on the related endgame of 3 vs 2. With f/g/h-pawns vs g/h, this is rather dead save for extreme King positions; more interesting is e/f/g vs f/h. In this case it is sometimes possible to win, for example:

 The weakness of f7 was decisive, excluding black's King and finally leading to zugzwang. But this was an extreme case, and even then white only just won. The simplest defence is ...f6 to avoid the cramp, and then ...h5 to simplify into e/f vs f, usually a safe draw. The attacker needs an active King and a fixed target at f7 - so in conclusion, simplifying into 3 vs 2 is only worthwhile in especially good circumstances.

The Main Defensive Idea: The limited 3 vs 2 winning chances lie in e/f/g vs f/h, so in 4 vs 3 (e/f/g/h vs f/g/h) the defender should prevent the exchange of his g-pawn for the attacker's h-pawn. This can be done by ...h5, a key resource - but it's still possible to go wrong, even for the best:

 Nevertheless, getting in ...h5 is a big achievement, and the defender should hold after that with accurate play.

Attacking Possibilities: If the defender aims for ...h5, then the attacker aims to stop this - g4 is the way. If he can do this and secure a positional grip, there are two main plans - create a passed e-pawn or transpose into a winning 3 vs 2 position:

Summary and Final Thoughts: I already noted the value of a verbal summary - I quote from my notepad:

• (3/2) With f/gh/h vs g/h, the attacker usually cannot win unless the defender's King is cut off, and his own King can infiltrate.
• The attacker's chances improve with an e-pawn, but even so he must be able to get his King to e5/e4 and fix the f-pawn on its starting square.
• So the best defensive plan with e/f/g vs f/h is ...f6 followed by ...h5, then holding e/f vs f. The attacking plan is to break through with e6, shelter the King on f6 and promote the f-pawn.
• (4/3) The attacker needs material and positional advantage to win - his plan is to either break through with the e-pawn or transpose into a favourable 3 vs 2.
• The best defensive plan is ...h5, to stop hxg leading to e/f/g vs f/h. In reply, the attacker's best idea is marching the King via h3 and aiming for the e6 breakthrough.

 This brief introduction to 4 vs 3 is meant to be cursory, but it's still a big step forward for me, and enough to be of practical benefit. I noticed a couple of things: one, the cliché is true and players of my strength are often terrible at endgames (see **); two, this ending may not be as simple a hold as is often made out, especially if the attacker is resourceful. So it's always worth a try...

 This series is finally done, although my endgame work is not - such a huge issue cannot be resolved in six weeks. My very last suggestion, for anyone working on their chess in general, is to be patient: my most intense study often coinides with my worst form, and it never leads to instant jumps. The mind seems to take its time to assimilate things, but the results will come in the end. In this case, I may get some psychological benefit right away - I don't need to go into endgames expecting the sky to fall down at any moment. For those who aren't interested in such things, I salute your patience - recent weeks may have been tedious. But now it's almost time to send in the clowns! We're heading back to Downend for the second week of the Summer Quickplay, after my summary post of the second half of 2014-15.


* Endgame knowledge is concise enough to fit into one volume. I've stitched this post together from a few sources, but the modern reference works of choice are Fundamental Chess Endings by Müller and Lamprecht, and Dvoretsky's Endgame Manual.

** I won a game in the SCCU U-18 Jamboree in 2011, but I had no idea what I was doing until my opponent hung a pawn. Two years later I beat Cassie Graham in a crucial rapidplay game - but I was still clueless until the disastrous 43. f3??. This blog's uncredited star Alex Hardwick snapped the streak by drawing easily soon after; in analysis I picked up the basic idea of ...h5 being a Good Thing, but it didn't stick. Finally Robert Radford self-destructed like Graham with 47. h3??. Apart from Alex, the standard of play was terrible! But everyone knew it should be a draw...