Riddle: Is active defense always best?

by Karsten Müller
4/24/2024 – It is a famous game, played between Siegbert Tarrasch and Akiba Rubinstein in 1911. It ended in a draw, and is given as a model example for an active defence in rook endings in the books of Levenfish/Smyslov and Averbakh. However, matters are not that easy. Can you find the mistakes in the game play strategy?

ChessBase 17 - Mega package - Edition 2024 ChessBase 17 - Mega package - Edition 2024

It is the program of choice for anyone who loves the game and wants to know more about it. Start your personal success story with ChessBase and enjoy the game even more.


The San Sebastián tournament of 1911 was one of the strongest ever, despite the absence of World Champion Emanuel Lasker. Fifteen masters were invited who had won at least two 4th prizes in international tournaments. Somewhat surprisingly, the youngest player, 22-year-old European debutant José R. Capablanca, won the competition with 9½/14.

Drawing of the participants of the San Sebastián tournament of 1911 in the Kronen Zeitung. Rubinstein is fifth from the left (behind the chess clocks), Tarrasch seated on the right.

The favourite, Akiba Rubinstein from Łódź, 30 years old and co-winner of St Petersburg 1909, shared 2nd prize (with Milan Vidmar) with 9/14. He started modestly with 5 draws, but later in round 13 he defeated the Cuban in their head-to-head encounter to make it a close finish.

On the other hand, Dr Siegbert Tarrasch from Nuremberg, a doctor by profession, had achieved his most glorious triumphs some years earlier, such as in Monte Carlo in 1903. In 1908, he suffered a clear 3-8 defeat in a World Championship match against Lasker. But at the age of 48, he was still considered one of the best players. He came into this game with one win and four draws, and eventually shared 5th to 7th place with 7½/14.

Over the course of their lifelong direct encounters, Rubinstein would do very well against Tarrasch. Out of 20 games, he won 8, drew 12 and lost none. However, in their first-ever meeting here in round 6, Rubinstein got into trouble in a double rook endgame and lost a pawn on the queenside.

Click on the fan button to start an engine, and on the maximize or the layout button (second from the right) to optimize the replay window. You can also download the PGN (disk button below the notation) and analyse the endgame with the chess engine of your choice. 

This famous endgame is analysed by R. Fine in his classical book "Basic Chess Endings" from 1941. Later, Levenfish/Smyslov in "The theory of rook endings" (1957), Y. Averbakh in "Turmendspiele 2" (1984), Donaldson/Minev in "The life and games of Akiva Rubinstein 1" (2006), J. Pinter in "1000 Rook Endings" (2007) and A. Panchenko in "Theory and practice of chess endings 2" (2009) approved the original annotations.

Based on Rubinstein's successful play to save his critical position, it is cited in all sources as a prime example of an active defence. Send us your assessment and analysis of this endgame. Wolfram Schön has analysed it in depth and reached amazing new conclusions – which we will publish in a week.

So what do you think? Please share your ideas, thoughts and analyses in the comments!

Previous riddles

Karsten Müller is considered to be one of the greatest endgame experts in the world. His books on the endgame - among them "Fundamentals of Chess Endings", co-authored with Frank Lamprecht, that helped to improve Magnus Carlsen's endgame knowledge - and his endgame columns for the ChessCafe website and the ChessBase Magazine helped to establish and to confirm this reputation. Karsten's Fritztrainer DVDs on the endgame are bestsellers. The mathematician with a PhD lives in Hamburg, and for more than 25 years he has been scoring points for the Hamburger Schachklub (HSK) in the Bundesliga.
Discussion and Feedback Submit your feedback to the editors


Rules for reader comments


Not registered yet? Register

Karsten Müller Karsten Müller 4/29/2024 02:50
For the solution see
Karsten Müller Karsten Müller 4/28/2024 09:23
njg149chess: You have a point, but matters are complicated:
1) 45.a4 f4 46.Ke1! indeed wins. But 46.a5? f3 is only drawn. Ke1 is a deep point.
2) 46.a4 indeed wins.
3) 47.a4? would be losing due to 47...f4 but of course the game continuation 46.Rb5? was too slow.
njg149Chess njg149Chess 4/27/2024 01:51
I don't understand why on move 35 and after, Tarrasch didn't just push the pawns -- starting with a4! Instead, by move 40 he gave up his OUTSIDE CONNECTED PASSED PAWNS while his king was perfectly positioned in front of the enemy pawns should they try to advance. What he got in return are two isolated pawns and a poorly placed king -- all through unforced errors. I teach my grade school students to push passed pawns in the endgame, especially outside passed pawns, and especially when they are connected. This is all rudimentary. The rook could not stop the a-pawn without letting the b-pawn advance, and black is helpless.
Karsten Müller Karsten Müller 4/24/2024 04:14
PhishMaster: Well done! Many thanks! And indeed in this case active defense is wrong...
PhishMaster PhishMaster 4/24/2024 03:14
My take is that active defense is probably the right way to go most of the time. Sometimes, as in this game it is wrong, HOWEVER, and that is a big "however", it creates pressure and practical problems to overcome. In this game, that pressure was rewarded.

At my club, frankly, I out-rate the second tier by about 200 points. Not always in the endgame, but I win a lot of games just by putting positional pressure on them, and they crack, or they make more weakening moves worsening their positions further. Tal, and I am no Tal, put pressure on in attacking positions, and was famed for his sacs being incorrect. Sure, they were incorrect if you feed them to a computer, but that pressure in a game between humans was unbearable for most mortals. Pressure is a critical aspect in chess, and that is what an active defense does for you.

When deciding what way to go, you have to weigh if your activity is great enough to create REAL problems against if you just sit there, can your opponent improve the position enough to eventually break your position down? A lot of that will depend on where you are in the game, and how many permanent weaknesses you have. I think in the middlegame, passive defense is far less likely to hold than in the endgame. In the endgame, you often have only one weakness, maybe two weaknesses, so you can hunker down.

Here is my analysis saved to the cloud. Some of the lines with a4 are not that obvious, so it is not surprising that both players did not see them, despite how great they were.

Karsten Müller Karsten Müller 4/24/2024 02:38
Hammerchewer: Well done! But one problem remains: the position after 33.Ra6 is drawn...
Karsten Müller Karsten Müller 4/24/2024 02:34
michelw: It was not part of the riddle, but both moves 33.Ra6 and 33.Ke2 are OK and it is a matter of taste. I like 33.Ra6 more as it makes a bit more pressure...
arzi arzi 4/24/2024 02:02
Riddle: Is active defense always the best?
Of course not. It depends on the game situation, i.e. the position of the pieces on the board, and sometimes also on your and your opponent's use of time.

michelw "I would rather play 33. Ke2 in orde..."

It doesn`t work because then after black pawn to f4 and black king f5 black rook use open g-line. Black can sacrifice b and e pawns for of powerful fpawn.
michelw michelw 4/24/2024 01:29
I am at work right now, so I don't have access to any books. My first impression is that 33. Ra6 is wrong. I would rather play 33. Ke2 in order to avoid Black occupying the second rank with a Rook
Hammerchewer Hammerchewer 4/24/2024 12:04
I checked the old books and it seems that Fine, Smyslov etc. only considered 35 a4 f4 36 a5? (but Ke1 would transpose to the winning line below) when after 36...f3 37 Ke1 Re2+ White must play 38 Kf1 leading to repetition since Kd1 loses. Tarrasch correctly played 35 Ke1 gaining a tempo but after 35...Rc2 his 36 Rb5 (which Smyslov awarded a !) was actually the move that spoiled the win. Instead 36 a4! f4 37 a5 f3 38 a6 Re2 39 Kd1! is possible, though some of the winning lines are difficult and long. The main point seems to be 39...e3 40 h4+ (or 40 fxe3 first) but I look forward to seeing the full solution. In some lines Black obtains a new Q but it doesn't save him.