Historical riddle: Benko vs Kortschnoi

by Karsten Müller
9/16/2020 – Karsten Müller is the Sherlock Holmes of unsolved endgame riddles, and he passionately searches for the truth in complicated endgame positions. This time, he wants to know whether the famous endgame Benko vs Kortschnoi, Curacao 1962, was indeed won for White. And he needs your help to find the solution. | Photos: Dutch National Archive

ChessBase 16 - Mega package Edition 2021 ChessBase 16 - Mega package Edition 2021

Your key to fresh ideas, precise analyses and targeted training!
Everyone uses ChessBase, from the World Champion to the amateur next door. 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 old riddle Benko - Kortschnoi

Hte Hungarian-American Grandmaster Pal Benko (July 15, 1928 – August 26, 2019) was a theoretician, an author, and an outstanding composer of chess problems and endgame studies. Of course, he also was an excellent endgame player and the following masterpiece has been often praised and analysed. However, endgame specialist Dr. Karsten Müller thinks that this fascinating duel knight vs bishop deserves to be studied and investigated again, and he invites the readers to do so.

The five Soviet participants (Paul Keres, Efim Geller, Mikhail Tal, Tigran Petrosian and Viktor Kortschnoi, front row, from left to right) and the Soviet delegation arrive in Curacao.

The duel bishop vs knight can be very complicated. Usually, the quick and wide-ranging bishop wants dynamic positions while the flexible but slow knight prefers static positions in which it can maneuver.

In the diagram position below Benko's knight is in control, and the only question is whether White can win or not. However, this question is not that easy to answer. So, dig deep!


The task is: was the endgame always won for White, and, if not, where did White and Black make mistakes? Use the comment section below to share your ideas, insights and analyses!

Magical Chess Endgames

In over 4 hours in front of the camera, Karsten Müller presents to you sensations from the world of endgames - partly reaching far beyond standard techniques and rules of thumb - and rounds off with some cases of with own examples.


Karsten Müller, born 1970, has a world-wide reputation as one of the greatest endgame experts. He has, together with Frank Lamprecht, written a book on the subject: “Fundamental Chess Endgames” in addition to other contributions such as his column on the website ChessCafe as well as in ChessBase Magazine. Müller's ChessBase-DVDs about endgames in Fritztrainer-Format are bestsellers. The PhD in mathematics lives in Hamburg, where he has also been hunting down points for the HSK in the Bundesliga for many years.


Rules for reader comments


Not registered yet? Register

Karsten Müller Karsten Müller 9/28/2020 01:56
On the German website the solution has just been published:
Karsten Müller Karsten Müller 9/25/2020 02:11
Good point! 57.Nh7? only spoils it due to the threefold repetition. Otherwise the position would still be winning.
malfa malfa 9/25/2020 12:00
Charles, personally I never played a second adjournment, most probably because you are right: if necessary, it would take place after move 72, not 56, as otherwise the second play session would last no more than two hours, which does not make much sense. So Benko presumably decided on 57.Nh7? over the board, after reaching the time control with a couple of repetitions. It must be noted that in itself this move does not spoil the victory, since on the next move he could have played 58.Nf6+ in order to follow with 59.Ne8, going back to the winning path, but the problem is that in this case Black would have been able to claim a draw by repetition, since 58...Kc6 would have produced for the third time the position which occurred after Black's moves 54 and 56. So it is only in conjunction with those repetitions before the time control that 57.Nh7 spoils the win.
Karsten Müller Karsten Müller 9/25/2020 09:49
malfa: Good question. I will send a few e mails to friends...
CharlesSullivan CharlesSullivan 9/25/2020 02:34
malfa: Benko does not mention a second adjournment. Someone please correct me if I am wrong, but in those days, after the first time control of 40 moves in 2 1/2 hours for each player (5 hours combined), there was usually an adjournment. After resumption of play, time control would be 1 hour for every 16 moves with a second adjournment possible after another 5 hours of combined play. I suspect that both players reached move 72 with time to spare and, since the end result was obvious (?), were able to complete the game without a second adjournment.
malfa malfa 9/24/2020 08:01
Well, very interesting input by Zoran and Charles,, thanks! First of all, regarding the first adjournment I think that Benko deserves more credit than Timman, since even if timetrouble addicted would normally play more than 40 moves in order to ensure that they passed the time control, four more is a bit too much. Anyway my first remark is that apparently Korchnoi's faulty pawn advance to h3 was conceived during home analysis, unless over the board he suddenly smelled some rat and hastily changed his defensive plan. But even more surprisingly, from Benko's notes we can infer that by playing 57.Nh7? he went wrong immediately after resuming from the second adjournment! His words let me think that at that stage he felt the win in his pocket and due to this sense of false security he did not pay enough attention to the conversion procedure. Rather bad homework from both, therefore, which may also be an indication of how tired the players were in these last rounds...
Karsten Müller Karsten Müller 9/24/2020 06:18
From Zoran Petronijevic:
Dear Karsten,

In his book "My life, games and compositions" Benko had annotated his game against Kortschnoj, and only place where he mentioned time was after 56th move: "White safely made the time control because Black couldn't do anything but repeat moves (this made it easy for me to flick out a bunch of solid moves without any risk whatsoever)."

Best wishes,

CharlesSullivan CharlesSullivan 9/24/2020 04:18
Benko also says that "White safely made the time control" at move 56. I suspect that Benko was, as usual, not managing his time very well. Benko provides no information about his opponent's time management, but Korchnoi also tended to "over think." Curiously, I believe that Benko and Korchnoi were both very good blitz players.
Karsten Müller Karsten Müller 9/24/2020 12:35
From Zoran Petronijevic:
Dear Karsten,

About adjourned move, we have two data. According Timman, sealed move is 44.Nc8 (Timman, Curacao 1962). However, according Benko, sealed move was 41...Kc6 - Kortschnoj offered draw and Benko rejected. We do not have data about time.

Best wishes,

Karsten Müller Karsten Müller 9/24/2020 09:25
Yes 71...Qb7+!? is indeed much more tenacious. White would then have to win the queen endgame with the pawn f4. I do not have more information on the clock situation and ask the readers to help here.
malfa malfa 9/23/2020 06:23
Karsten, thanks for Your comments. One last point: the queen endgame is forcibly won according to the TBs, but had Black played 71...Qb7+! instead of 71...Qa2+ it appears that he would have forced White to give back one pawn, since in this case his could not protect h2 and simultaneously escape to the continues checks. Then the resulting Q+P vs Q would have been still lost for Black, but at least he could have rendered the conversion task much more difficult for his opponent than in the game. It would be very interesting to know the clock situation throughout the whole endgame, which I suppose was played after an adjournment. Maybe was 42.Kd4 the sealed move? Anyway, if I am not wrong this was round 25 of the 28 played in this gruelling tournament, so Korchnoi had already said farewell long time before to his initial form and to his hopes of becoming the Challenger.
Karsten Müller Karsten Müller 9/21/2020 10:30
malfa: Very good! In both cases you are right. 57.Ne8! wins, while after 57.Nh7? it is drawn. And most often White triangulates with his king with Kc4-d4 to bring Black into zugzwang. Indeed quite deep!
malfa malfa 9/19/2020 12:12
Oh, I forgot to mention another try I considered, maybe more promising: 57.Ne8!? instead of 57.Nh7, when 57...Kd5 58.Nd6 produces some sort of Zugwang, which forces Black to worsen his setup, e.g. 58...Bb1 59.Nc4! and 59...Ke4?? is now temporarily impossibile because of the knight fork on d2, so 59...Kc6 must be played, when 60.Kd4 Ba2 61.Nd6 Bb3 62.Nf7 eventually forces the black king to retreat on d7, when again White should be able to win by infiltrating his king on the kingside as he could have done after 47.Ng6.
malfa malfa 9/19/2020 10:07
Karsten, that a win after 47.Ng8 escapes to modern engines is no wonder to me, since it is the winning concept which eludes my human analysis: the best I can think of is some way to go back to the winning path I found after 47.Ng6, for example via some triangulation attempt to displace the black pieces from their optimal squares, like 49.Kc4!? instead of 49.Kd4, but I am not sure that it forces the win, since in this case it is not clear to me which should be the best defensive strategy by Black.
Karsten Müller Karsten Müller 9/19/2020 09:19
albitex: Yes all correct. But there are more mistakes...
Karsten Müller Karsten Müller 9/19/2020 09:17
malfa: The solution will be published soon. This is so deep that even several engines have problems to see the win after the game move 47.Ng8?!. Such systematic wins are sometimes easier for humans than for the modern engines...
albitex albitex 9/19/2020 04:54
45...Kb5?! (45...Kd7 = ex: 46. Ng6 Ke8 47. Nxh4 Bb7 =)
albitex albitex 9/19/2020 04:43
46... h3? (46... Bb7 =)

66...Kd3? (66...Kd4 =)
malfa malfa 9/18/2020 07:46
@Zanoni and Kartsen

That the exchange of the 'b' pawns is losing for Black is not a surprise for me, since the resulting endgame strongly resembles to a from the Sultan Khan-Tartakower 1931 match, game 5, which Grigorjev, the legendary endgame expert, analyzed to a forced win in a book I read as a youngster. I wonder if Timman borrowed from there his idea of "systematic win". The fact that, when White reaches a piece placement like W. Kf6/Ng6 vs. B. Ke8/Ba2,
he wins with Nh8!-f7-g5 again highlights how fatal was to push Black's h-pawn all the way to h3.

I am still puzzled to learn that 47.Ng8 was also winning, I need to give it a deeper look!
Karsten Müller Karsten Müller 9/18/2020 05:03
Yes indeed right! A typical systematic win as Timman put it.
Zanoni Zanoni 9/18/2020 04:58
Thanks for the hint. Now I see the King's march to g7 followed by Nf7-h8-g6+ and so on.
Karsten Müller Karsten Müller 9/18/2020 04:29
Timman is right and you are also right as your 59.Kh4 wins. Please go deeper into this line!
Zanoni Zanoni 9/18/2020 02:51
I have a question.

Timman once wrote that 50...Kxb4 51.Nxb6 Kb5 52.Nd7 Kc6 53.Nf8 Bd5 54.Nh7 Kd7 55.Ke3 leads to a systematic win by picking up the h-pawn.

The pawn ending is a draw however, provided the black King is on e7 so that he can move it to f7 when it occurs, for instance 55...Ke7 56.Ng5 Bg2 57.Kf2 Kd7 58.Kg3 Ke7 59.Nxh3 Bxh3 60.Kxh3 Kf7 with an easy draw.

Of course the white King can instead walk to g6 beginning with 59.Kh4, but I could not find a win in this line either.
Karsten Müller Karsten Müller 9/18/2020 02:08
Yes well done! That is right. And so the game move 46...h3? was a losing error. But not all secrets are uncovered yet as the game move 47.Ng8 wins as well...
malfa malfa 9/18/2020 12:36
I see. After your 62...Bf1 I counted on 63.Nb3 Bc4 64.Nc5 but after 64...Ba2 65.Rd3 Bc4+ 66.Kc3 Bf1 it apparently leads nowhere.

Therefore if White has blown it this happened earlier. Since I still maintain that when he advsnced his rook pawn as far as to h3 Black was careless in hs turn, is it possible that 47.Ng6 instead of 47.Ng8 was the way to punish him? The idea is to somehow forbid the black king from getting to d5, e.g. 47...Kc6 48.Kd4 Kb5 49.Nf8 Bd5 50.Kc3 Kc6 51.Nh7! Bg2 52.Kd4, when the black bishop, as I suggested previously, cannot simultaneosly cover both the pawns on e6 and h3, so the passive 52...Kd7 becomes necessary, but in this case after 53.Ng5 Bf1 the retreat 54.Ke3 Bg2 55.Kf2 now ensures that either one of those pawns will fall or the white king will manage to infiltrate via h4, without Black being in time with his queenside counterplay, e.g. 55...Ke7 56.Kg3 Bf1 57.Kh4! (57.Nxh3? Bxh3 58.Kxh3 Kf7 59.Kh4 Kg6 is a draw, since the black king is in time to close the way) Bg2 58.Kh5 and while there is still work to be done the feeling is that the black pieces are too passive to avoid the unavoidable.
Karsten Müller Karsten Müller 9/18/2020 09:44
But the position is drawn in any case. After your 61.Nf3 Black plays 61...Bg2 62.Nd4 and now 62...Bf1!=. I have also written this to the ChessBase Let's Check database.
malfa malfa 9/18/2020 01:24
After 60...Bf1 another critical point strikes me: exchanging the b4 pawn for the one on e6 allows Black too much counterplay, since with passed pawns on opposite sites there is pretty good chance that the bishop even proves itself stronger than the knight, so what about seizing a golden opportunity to centralize the latter? I mean 61.Nf3 instead of 61.Kf2, when 61...Kc4?? loses to the fork on d2. If instead Black goes on with 61...Bg2 as in the game, than after 62.Nd4 Kc4 63.Nxe6 Kxb4 White's king is much better placed than in the game and he looks winning, since with both pieces into play he will easily keep Black's passed pawn under control, while preserving his two trumps: his own passed pawn on the e-file and the weakness on h3.
Karsten Müller Karsten Müller 9/17/2020 01:34
But Black is lost in any case. 51...Kc6!? does not save him.
tip4success tip4success 9/17/2020 01:09
I don't like 51...Ba2, it puts the bishop diagonal off the H1-A8 diagonal and black soon after is "almost" forced to play b5 to keep the white king from invading via the white squares. Some kind of king retreat seems safer, i.e. 51...Kc6!?
Karsten Müller Karsten Müller 9/17/2020 10:05
Please send it to info et chessbase.com
Scorpion29 Scorpion29 9/17/2020 09:38
Karsten, I seemed to have found the losing with Kd3?, as already pointed out by the others. How do I send you the annotated PGN of the same? Thank you!
Karsten Müller Karsten Müller 9/17/2020 09:23
brian8871: Yes you are right with the point that 66...Kd4 draws and 66...Kd3? loses. But there are more mistakes earlier...
Karsten Müller Karsten Müller 9/17/2020 09:11
You have a point and your move 44...Bg4 indeed draws, but 44...Bg2 also does not lose. The mistakes come later.
Poiuy Trewq Poiuy Trewq 9/17/2020 05:19
I believe Black should have been more defensive minded, and protected his "rear guard" more effectively. After exploring the positions (without use of a computer program) I agree with Malfa that Black would have done far better with an active Bishop, and so my choice for the fatal error is 44 ... Bg2. Black would still have a chance to control the game with 44 ... Bg4, e.g. 45 Ne7+ Kb5 46 Kc3 Bh5(!), limiting the Knight's ability to fork pawns and force a pawn race.
brian8871 brian8871 9/16/2020 10:18
66...Kd3 appears to be a mistake. After 66...Kd4 67. Nxg2 hxg2 68. Kxg2, Black wins. The best White can do is hold a draw.
Karsten Müller Karsten Müller 9/16/2020 06:36
You have a point and your move 42...Bg4 indeed draws, but 42...h5 also does not lose. The mistakes come later.
malfa malfa 9/16/2020 05:02
My gut feeling is that the Black's pawn advsnce led to nowhere else than a long-term weakening of his pawn structure, since e6 and h3 became impossibile to defend simultaneously for the bishop. Besides until it is kept on h6 that pawn contributes to keeping the kingside closed. Therefore I would have rather given priority to giving this piece more defensive scope, as it is initially out of play, so 42...Bg4 followed by 43...Bh5 if necessary. It is also possible that White's initial move was imprecise as well, and that the immediate 42.Nd6 was better.