Deceptive simplicity

"To improve at chess, you should in the first instance study the endgame" - Capablanca

Endgames are thought of to be relatively simple compared to the middlegame, because of the number of pieces that remain. However, this leads to a situation where there are more empty squares, and they are harder to contend with, while making decisions relative to the pieces. While a complex middlegame can be understood somewhat intuitively at least with regard to which pieces should be played and not played, we do not have the same liberty in the endgame with less pieces and more squares. I find Rook endgames and perhaps Queen endgames to be very complex in this regard to understand. Getting the key to the position is also far from easy!

Let us take a look at a position from a recent game of Mikey Adams which was shown to me by International Master Sambit Panda of Orissa.

Adams-Schroer Cambridge 2025

This was the position in the game after Black's 64th move. How complex can a position be with just Rook and one Pawn against a Rook? Well, that is where we might be in for a surprise! When trying to analyze this position it seemed to me that Black has checking distance with his rook, so if White pushes g4 that would lead to a draw as Black controls the important e4 square and stays in the vicinity of it.

So, my thoughts drifted towards taking the king forward with, 65. Kg6!? It was clear that if White cannot push the g-pawn up to g4 then the only winning try left was to get the rook to the g-file, ideally to g6 or g7 without letting Black's king near the first rank or f-file. A sample line would run 65... Rg8+ 66. Kf7 Rg3 67. Kf6! the clever idea being Re5 check followed by Rg5 and if Black King heads back to d6 we cut off one more file. I could do the same after Black's Rg8 too, and this seemed to finish the game off. But here was where an unpleasant surprise lay in store. Black stops the threat of Re5 with 67 .... Kd4!! and brings the king forward into the offensive. 68. Kf5 Kd3!! 69. Ra2 Ke3! 70.Rb2 Rg8 71.g4 Rf8+ 72. Ke5 Rg8 73. Rb2+ Kf2!

Black's King is cut off by one rank below White's passer, but the checking distance ensures the draw! What an active Black king this was exceedingly illuminating. I have not seen many positions in Rook endgames where the King attacks the pawns from the rear!

After trying in vain various King moves (as I knew 65.g4 would not work) I finally hit upon the winning move of 

65.Ra2!! in the initial position. Even after the move the win is far from easy and there are various interesting resources and nuances for both sides which look almost impossible to solve from the first position. But that apart, more one analyses these positions more ideas pop up, and this is perhaps why the generation of players of the yesteryear who had the privilege of adjournment analysis where way stronger in endgame than modern players who have to grovel with a rapid like time control throughout the game.

After analyzing the position deeply to the extent, I could I understood a few ideas which might help giving some direction in approaching the position.

1. White's winning try consists of getting the rook to the g-file and he needs to do this from ranks and sometimes use checking ideas to gain tempos.

2. Surprisingly Black's defensive idea is to in the first place not let this but also in the process take his king to the rear and de stabilize White's Rook on the second Rank if possible!

3. If White can cut Black off on the c-file then he is almost always winning even by just bringing the king to g1! The only case where this will not work is if Black is in a position to attack the Rook with Kb3! and White's king is not in a position 

4. White tries to combine getting to push the g-pawn up the board, with trying to further push Black further files away in cut off and also use the threat of reaching the g-file with his rook along the ranks. At some point Black's defense would crack under numerous threats.

There are several resourceful variations along the way but to tease the reader I will show some fascinating moments

This is a very interesting position, where the obvious move Rd2+ to further the cut off would actually throw away the win as it would allow Black's king to come up and attack our rook in some time! Readers can pause and think about a winning plan here- one which is far from obvious.

The key concept here is Zugzwang or rather passing the move to the opponent. If it were Black to move, he is running short of useful ideas. His king cannot come up to the 5th rank because of Ra5+ and Rg5! The rook gaining leverage via the ranks was one the key ideas of the first move Ra2! If Black's king drops back, he loses time and White can probably use it to cut him off further. If Black moves his Rook, then White's King zigzags and gains time for pushing g4 and combines this with pushing Black's King further along files, and if possible aslo get his own rook to the g-file. There are several nuances, but White wins here. Another interesting point is that White can lose a move by triangulating with his King or easier still by giving a couple of checks with his Rook and returning to the same post! 

Adams' game saw him going for 65. g4 which actually throws away the win, but only to be let off by his opponent in a few moves who went on to lose allowing a well-known rank to cut off. of course, time trouble was very much in place, so it cannot be judged too harshly.

Sandipan-Abhijeet Gupta 2004

While discussing the rook endgame with Grandmaster Sandipan, he brought his own game to my attention. I was playing in the same event but had missed it! This position occurred after White's 59th move. If Black plays Ke6 we reach a flank reversed mirror of the previous diagram! How should Black continue here and what is the assessment of this position?

The game went 60.... Ke5? and Black loses after 61. Rh5- b5

But the counterintuitive but purposeful 60...Kg4 or even surprisingly 60....Kg5!? was good enough a draw!

The Rook's path to b- file via rank is blocked and at the same time Black's king goes active in trying to destabilize the White's rook on the second rank.

Nakamura-Lenderman 2020

Finally, I came across this position from a Nakamura-Lenderman game from Jacob Aagard's wonderful book "Conceptual Rook Endgames" This is the position after White's 51st move. Black should clearly try for a defense based on the checking distance theme, which clearly, he would have been aware of. Yet it still does not guarantee choosing the right move in such endgames! 

While it might be too simplistic to say that going by the idea, Black had to play 51... Rh8! giving him enough squares to check opponent's king, it does not tell us why the move played in the game might not work, as Jacob mentions Alexander made an instructive mistake. He chose 51....Rg7+? figuring that he could retreat his rook to the 8th rank one move later, but as it happens White's king gains some shouldering space over his adversary and as the fascinating analysis show, this actually transposes to a variation of Adams and Sandipan's game. All in all, it shows the incredible complexity hidden amidst the seemingly simple Rook endgames with lot of empty spaces/squares. 

For the record the variation given by Jacob runs 51.... Rh8! 52.Kf5 Rf8+ 53. Kg6 Kd5!! Black's King steps out from a possible Re6+ and thereby hinders White's main plan of bringing the Rook to the g-file along the ranks and thereby apparently ensuring the draw!

All three endgames are worth careful study. It can be fascinating to improve our perception of coordination and space.

Link to a detailed analysis of games

I thoroughly enjoyed analyzing these endgames and I believe the readers would derive similar pleasure to doing the same. I conclude with a poem which resulted after analyzing these positions

கண்ணன் அன்ன கட்டக்களி ஆட்டம்

எண்ண எண்ண சொட்டும் அவன் மாயை

உண்ண உண்ண ஓயா சுவை நாளும்

அண்ணி என்னை ஆட்கொள் வரம்பு அற்றே

 kaNNan anna kattakkaLi aattam

eNNa eNNa chottum avan maayai

uNNa uNNa Oyaa chuvai naaLum

aNNi ennai aatkoL varambatrE

The game of chess is like Krishna, and the thoughts and possibilities that arise are like his Maya, the more we partake it the taste keeps growing forever and it overtakes me (chess lovers) without any boundaries!

Thanks to Sambit, Sandipan and Jacob Aagard