Rook Endings (4)

Two more practical examples of rook and pawn against rook from games played at Richmond Junior Club.

In this position the good news for Black is that his king is in front of the pawn and the white king is subject to mating threats on the side of the board. The bad news is that his rook is badly placed, and that it’s White’s move. (If it was Black to move he could win by moving his rook in a westerly direction.)

His plan should be to get his rook round the back to threaten mate, while White will need to counter this by moving his rook away to check the black king from the other side.

White now has two moves to draw: Ra6 and Rb6. He needs to meet mate threats with horizontal checks, and has to be as far away as possible from the enemy monarch.

But instead he played 55. Re6, presumably with the idea of keeping the black king on the f-file. Now any westerly rook move is winning for Black. He chose 55… Re1, having observed correctly that the pawn ending would be winning. White went back behind the pawn: 56. Rf6, and now, out of Black’s 17 legal moves, 11 are winning and 4 are drawing. The quickest winning moves are Re7 and Re8, both mating in 21 moves according to the tablebases. He actually chose one of the drawing moves: 56… Re2, missing the winning plan of threatening mate on the h-file. Now White again has time to draw by moving his rook over to the far side of the board (note that this is one of many positions in these endings where you want your rook on the side rather than behind the passed pawn). This time, Ra6, Rb6 and Rc6 all draw, but in principle he should move as far away as possible. Instead, stuck with the mistaken idea that rooks always belong behind passed pawns, he played 57. Kh3.

Now Black has four winning moves: Re8, Re7, Re5 and Re3 (but Re4 is only a draw). Still not thinking about potential checks on the h-file he chose perhaps the least obvious of these, 57… Re3. White played 58. Kh2 when Black has a choice of 14 moves, of which 8 win and 5 draw. As you would expect by now, the quickest wins are Re8 and Re7. Instead he went for one of the drawing options: 58… Ke2.

Now White has 16 possible moves, but only one of them draws: Kg3, hitting the f-pawn. After his actual choice, 59. Kg1, though, Black can again win by moving his rook in a northerly direction, again planning a check from behind. Instead he gave up and pushed the pawn: 59… f2+. White was happy to capture the pawn: 60. Rxf2+, and a draw was agreed.

If you’re down to the last few minutes on the clock, or, as is likely these days, playing on an increment, it’s all too easy to think inflexibly, as both players did in this example. Black seemed to be thinking purely about how to push his f-pawn, while White was just trying to prevent this. Neither player was thinking about how to check the enemy king.

Our final example starts off by being about getting your king in front of the pawn, but when Balck fails to do this it’s just about calculation. Will White calculate accurately? We’ll see.

Black has to make his 52nd move. He has 15 moves to choose from, three of which lose his rook, although one of them, Rg2, still draws (rook against pawn is another interesting subject). There are 10 winning moves and two other moves that draw: Rg4 and the move he chose, 52… f3.

Now it seems very natural and obvious to push your pawn, and you’ve probably been taught that passed pawns should be pushed, but when you possess the only remaining pawn on the board you often want your king in front of the pawn. This is the case here.

White found the only move to draw: 53. Kd4, correctly rushing back with his king. His rook is well placed on the h-file here, preventing the black king from travelling to g2 via h3. Black pushed the pawn again: 53… f2, for the moment preventing the white king’s approach. White again found the only drawing move: 54. Rf7. (Rg7+ would have led to king and queen against king and rook, which would have been another story entirely.) Black naturally replied by defending the pawn with 54… Rg2.

On his 55th move White has no less than 21 choices (the maximum number of 8 king moves and 13 rook moves, one short of the maximum, for those of you who care about this sort of thing). Nine of them draw and the other twelve lose. The most obvious draw is the simple Ke3 just winning the pawn and demonstrating to black that he pushed his pawn too quickly. However he was seduced by the skewer 55. Rg7+, no doubt playing too fast to notice that after he won the rook Black would promote.

Now Black has six king moves, but the only one to win is Kf6, when he’ll reach the tricky ending king and queen against king and rook. It’s mate in 28 according to the tablebases, but would he have been able to win? We’ll never know because instead he played 55… Kh4.

White’s now drawing again if he finds 56. Rf7, getting back behind the passed pawn and preparing to meet 56… Kg3 with 57. Ke3, when Black can make no progress. His actual choice of 56. Rh7+ was too slow, though, because now after 56… Kg3, which Black played, his king will have time to reach g1 via h2. The game continued 57. Rg7+ Kh2 58. Rh2+ Kg1 and Black won by promoting his pawn.

Richard James

This entry was posted in Annotated Games, Articles, Children's Chess, Endgames, Improver (950-1400), Intermediate (1350-1750), Richard James, Strong/County (1700-2000) on by .

About Richard James

Richard James is a professional chess teacher and writer living in Twickenham, and working mostly with younger children and beginners. He was the co-founder of Richmond Junior Chess Club in 1975 and its director until 2005. He is the webmaster of chessKIDS academy ( or and, most recently, the author of Chess for Kids and The Right Way to Teach Chess to Kids, both published by Right Way Books. Richard is currently the Curriculum Consultant for Chess in Schools and Communities ( as well as teaching chess in local schools and doing private tuition. He has been a member of Richmond & Twickenham Chess Club since 1966 and currently has an ECF grade of 177.