An interesting and instructive rook and pawn ending

My favourite chess vlogger (not a word I use often, in fact, the first time I have used it) is GM Alex Yermolinsky. His series on ICC 'What every Russian schoolboy knows' has been consistently very good and often superb.

Note the word has: Alex wrote to me recently to say that he has parted company with ICC. I hope the website will continue to make the series available for readers through their archive. I heartily recommend them, particularly the endgame ones.

A few weeks ago, Alex had recorded one of his best videos, looking at certain aspects of R+3P v R+2P, starting with some aspects of a recent Aronian-Carlsen game 0-1 but mainly concentrating on a similar ending Can-Laurusas from the 2014 Tromsø Olympiad. (the video entitled ”they're all good chess players”, 9/28/14)

As white, how would you proceed?

As black, how would you defend?

What 'should' the result be?

Can won the game, which Alex covers in detail. His video is replete in insightful comments and in tactics: some cute mating tactics and potassium cyanide make the video a joy to watch.

I found one comment particularly illuminating, when he explained how easily the above pawn structure can arise, for instance from 4v3 endings where the side with three pawns has a pawn on f4 (white). For instance, in this position, which I have made up, how might white try to make progress? I have taken the rooks off, to concentrate just on the pawns, but imagine each side had a sensibly placed rook on the board.

Before watching Alex's video, I might stupidly as white have played e3-e4, especially if my king were at d3: but swapping off my extra offside pawn is, I now realise, entirely the wrong thing to do. Maybe I wouldn't have played e3-e4, but if I didn't, I think I would have played aimlessly, just hoping something came up. Now, in such positions, I would try g3-g4! getting the pawn structure in the diagram at the start. And as black, I might well play h7-h5 here to prevent white's pawn break.

The over-riding lesson from Alex's lesson is that white's plan comprises (i) fiddle around, or more professionally, 'do not hurry': black's defensive task is dour, and making it more painful by not hurrying can cause a mistake. We have all been there; (ii) try to create two connected e and f passed pawns, by playing h5, and hoping black will exchange on e5; (iii) if black doesn't exchange, try to make the Pg6 into a weakness, but (iv) there are some tactical reasons for not rushing to exchange h5*g6.

Later, in his analysis, we get positions like the following:

Black to play

If 1…Kh5 2 e8(Q)! and 3 Rh7 mate. In several of Alex's lines this mating trick occurs: and it is significant that after white plays h4-h5 he is happy to keep the pawn there, and is in no rush to swap on g6. This tension and delay has profound tactical implications.

Purdy's potassium cyanide also occurs, out of the blue.

Alex ends by suggesting there was a narrow line to draw, but also invites his audience to write to him with further suggestions since he felt there might be further aspects. I looked at it, and wrote to Alex suggesting a possible winning line with a manoeuvre relying on different again tactics.

Alas, Alex won't be commenting on my analysis, so I may not know the ultimate conclusion on the ending. I have therefore posted my present analysis here for my readers. I would suggest ICC members watch Alex's video first but they might choose just to look at my analysis. It would be great if readers could post comments and confirm, improve or destroy my analysis.

My conclusions?

i) I am not a strong enough player to know what I am talking about. As a child, I should have listened to my trainer (shame is, I never had one) and studied the endgame. Only now, as a past my sell by date player, do I appreciate endgames;

ii) It is close to a draw. In one of Alex's lines, black can probably get to a R+fp+hp v R ending. One awful shame of Alex leaving ICC is that he mentions on his video that he might cover this ending is future 'Russian schoolboy'; one of his lines results in a QvR ending but the engines show a long line of checks whereby the LPDO R drops off, before black can create a fortress;

iii) It just might be a win. My analysis gives one attempt (see my suggested 49 Rc7+ and a subsequent Rc4 rather than Alex's 49 f4) against Alex's suggestion of a best defence. I'm not confident it is a win, since I know that all rook endings are drawn, and if I have more time one day I will look at it afresh, but before then a reader might be able to comment. I hope so.

Posted with Blogsy