Riddle: Korchnoi-Karpov 1978

by ChessBase
11/27/2025 – As chess players occasionally do, Alex Fishbein, the U.S. grandmaster, was recently looking at a classic game from the past: Viktor Kortschnoj's victory over Anatoly Karpov in the 21st game of their 1978 match. Something was wrong. Charles Sullivan had done some intense research on this volatile encounter and asks for your assessment on the results he presents.
Your personal chess trainer. Your toughest opponent. Your strongest ally.
FRITZ 20 is more than just a chess engine – it is a training revolution for ambitious players and professionals. Whether you are taking your first steps into the world of serious chess training, or already playing at tournament level, FRITZ 20 will help you train more efficiently, intelligently and individually than ever before. 

The annotations in Mark Dvoretsky's famous Analytical Manual were not convincing.  And not just one or two variations seemed "off" – Fishbein had questions about everything. It turns out recent excavations have uncovered some fascinating points about this game from 47 years ago. GM Fishbein suggested to GM Karsten Müller that the game would make a good endgame puzzle, so here we go.

I've compiled the most important variations for this game from some of the most famous annotators, starting with the 22nd move.  You can play over these moves in the score below.  Our ultimate goal will be to pinpoint Black's final critical error. 

A) Black's 24th move: As Kasparov says, "this move was generally condemned."  Timman defended the move (24...Qe5), but Kasparov, in 2021, identifies it as the "likely" game-losing error.  Kasparov thought 24...Be7 would hold and Dvoretsky thought 24...Bf8 offered drawing chances.  What moves hold the draw for Black?

B) Black's 30th move: GM  Alex Fishbein thought that 30...axb5 was not only playable, but drawish.  Is he right?

C) White's 31st move: Conventional wisdom was that Larsen's suggested 31.b6 probably would be a winning endgame. Fishbein thinks it is a draw.  What do you think?

D) White's 35th move: Dvoretsky says that 35.Ra7+ was a winning move.  Do you agree?

E) Black's 38th move: Early commentators thought 38...Rb4 gave good drawing chances, but both Kasparov2021 and Fishbein think White still wins. Do you agree?

F) Black's 40th move: How do you evaluate Müller's variation?

G) White's 45th move: After 45.Rd8 Ke7 46.Rd7+ Ke8 47.e4 Rb5, do you prefer Timman's 48.Rd5 variation or Kasparov2021's 48.exf5 recommendation?

H) Black's 46th move: Does Black have a chance of saving the game after 46...Rb5?

So what was Black's fatal error?  Was it Black's 24th move or Black's 46th move, or something in between?

Sources consulted were:

  • Viktor Kortschnoj, My Best Games (Russian edition, Loan from Zoran Petronijevic)
  • Jan Timman, The Art of Chess Analysis (Everyman Chess paperback reprint of 2003, first published in 1980 by RHM)
  • Mark Dvoretsky, Dvoretsky's Analytical Manual (Russell Enterprises, 2008)
  • Gary Kasparov [Kasparov2006], My Great Predecessors, Vol. 5 (Everyman Chess 2006)
  • Gary Kasparov [Kasparov2021], My Great Predecessors, Vol. 5 (Russian edition 2021, Loan from Wolfram Schön)
  • Karsten Müller, Master Class No. 15, Viktor Korchnoi (ChessBase 2022)
  • Alex Fishbein, emails (October, 2025)

At the beginning of this game, Karpov led 4 victories to 1 (6 victories needed). Here is the game with all annotations available. YClick on the notation to get a replayer with engine support.

About the author

Charles Sullivan, who is 76, grew up in the San Francisco Bay Area. While at the University of California, Berkeley, he was captain of the golf team, marched with the Cal Band for one year, played trumpet in the concert band, and was tear-gassed by helicopter during the People's Park demonstrations in 1969 (it was a weird time). He taught elementary school for three years (his hardest job ever) before switching to computer programming until retirement. A favourite memory is playing chess while sitting on the steps (which was allowed in those days) of the Parthenon in Athens in 1973. Today he lives with his wife in Davis (near Sacramento).

<< Javokhir Sindarov wins FIDE World Cup
London Chess Classic: Theoretical debates >>
Reports about chess: tournaments, championships, portraits, interviews, World Championships, product launches and more.
Discussion and Feedback Submit your feedback to the editors

See also

FIDE World Cup, Finals - Live!

11/26/2025 – Today marks the final day of the World Cup in Goa, which began on October 31. But the key question of who will play in the Candidates Tournament in March and April 2026 has already been settled: Wei Yi, Javokhir Sindarov and Andrey Esipenko are in. Now it’s “only” a matter of deciding who wins the tournament. That will be determined today in the tiebreak between Wei Yi and Sindarov. The winner takes home the title and 120,000 USD, the runner-up receives 85,000 USD. | Follow the action live with expert commentary starting at 10.30 CET (5.30 ET, 15.00 IST)
More...
 85

See also

Congratulations - Ljubomir Ljubojevic turns 75!

11/4/2025 – Ljubomir Ljubojevic turned 75 on Sunday, 2 November. At the height of his career, he was one of the world's top players, ranking third behind Karpov and Kasparov in the 1980s. Thanks to his tactical attacking style, he was able to defeat almost any opponent. He retired from tournament chess in 2010. | Photo: Dagobert Kohlmeyer
More...
 11

Video

60m How to fight Queen's Pawn Openigns

After 1.d4 d5 many players with White avoid the great amount of theory in the Slav, Semi-Slav, QGA and Orthodox Queen's Gambit and do not therefore play 2.c4. This is not very ambitious, but the painful experience of many chess players has been that the Colle System, the Trompowsky Attack, the Torre Attack and the London System are nevertheless extremely dangerous. Black has to be prepared for each of these openings and IM Valeri Lilov offers you some help with his six instructive videos, in which he demonstrates for each single opening a relevant plan for Black. In addition to the openings mentioned, the Bulgarian trainer also delves into the Catalan, the Blackmar-Diemer Gambit and the Richter-Veresov Opening.

Discuss

Rules for reader comments

 
 

Not registered yet? Register

Karsten Müller Karsten Müller 12/2/2025 03:43
From Alex Fishbein:
“Fishbein first looked at some of these questions in the American Chess Magazine in 2022, but returned to this ending in 2025. He was not then familiar with Karolyi’s work which had been published in 2024.”
CharlesSullivan CharlesSullivan 12/2/2025 05:10
albitex: (a) Regarding best winning method after 45.Rd8 Ke7 46.Rd7+ Ke8 47.e4 Rb5 48.Rd5 fxe4+ 49.Kxe4 Bb8 do you see a way for Black to draw in Timman's variation that continued 50.Kd4 Ba7 51.Kc4 Rb6 52.Ne4 Bb8 53.Rb5 Re6 54.Nc5 Re2 55.Na6 Ba7 56.b8=Q+ Bxb8 57.Rxb8+ Kf7 58.h3 Rh2 59.h4 which Timman (and Stockfish) evaluate as a win?
(b) Regarding White's win after 46...Rb5. In Dvoretsky's several variations, do you see any improvements for Black that might lead to a draw?
(c) Regarding Black's 24th move, the computers do think that Kasparov's 24...Be7 reaches equality, but the variation you give for 24...Bf8 also would hold (according to the computers).
Karsten Müller Karsten Müller 12/2/2025 02:23
albitex: Many thanks for many points!
1) "Reader albitex also gave the hypothetical 38...Be5 39.b7 Rb5 40.Ra6+ Kf7 41.Rb6 Rxb6 42.Nxb6 Ke6". Yes your 43.Kh3!! wins.
2) "CharlesSullivan, Yes, 36...Ke6 is better, however it is not enough, White always wins (if he plays like a computer with tablebases..)" But 36...Ke6! should hold according to Renette/Karolyi. Can you give more details?
albitex albitex 12/1/2025 11:06
Point A)
Kasparov is obviously right (and how could he be otherwise?) 24...Be7 maintained the balance, but the question I can't answer is: why does Dvoretsky prefer 24...Bf8 to the spontaneous 24...Be7?
Let's see what happens after 24...Be7:
After 24...Be7 the plausible moves for White are 25. b3, 25. Qb3, 25. Rd4.

- 24... Be7 25. b3 Rc6 26. Qxe4 Qxe4 27. Nxe4 Bxa3 28. Rd8+ Bf8 =

- 24... Be7 25. Rd4 Rxd4 26. exd4 g6 27. Qxe4 Qxe4 28. Nxe4 =

- 24... Be7 25. Qb3 Qc8 26. g3 h6 =

After 24...Bf8 the only variation that differs from those of Be7 is:
24... Bf8 25. Rd8 Rc8 26. Rxc8 Qxc8 27. Qxe4 +/=
Here, even if the situation is still balanced, you don't have to be Kasparov to understand that the positions after 24...Be7 are slightly preferable.
Why does Dvoretsky prefer 24...Bf8? I don't know...
albitex albitex 12/1/2025 10:25
Point H)
For me, after 46... Kb5, Black can draw; the advantage is only theoretical.
46... Rb5 47. g4 fxg4+ 48. hxg4 Bd6 49. Ke2 h6 50. Kd3 Rb4 51. e4 +/=
In this position, the engines give White a clear advantage, but they don't tell us how to turn this advantage into a victory. In my opinion, Black can draw if he doesn't push pawns
albitex albitex 12/1/2025 09:32
Point G)
I prefer Kasparov's variation 48. exf5, which draws like 48. Kd5, but eliminating pieces makes it easier to avoid innacurancy.
After 48. exf5, we end up with a R + 3 pawns Vs R + pawn ending, which despite White's two extra pawns, is still a draw:
45. Rd8 Ke7 46. Rd7+ Ke8 47. e4 Rb5 48. exf5 (Kasparov variant) Rxc5 49. Rxg7 h6 50. f6 Kf8 51. Re7 Bb8 52. Rd7
(52. Rh7 Rf5+ 53. Kg4 Rxf6 54. Rh8+ Kg7 55. Rxb8 =)
52... Rf5+ 53. Ke4 Rxf6 54. Rd8+Kg7 55. Rxb8 Rb6 =
*
For 48. Rd5 variant:
48. Rd5 fxe4+ 49. Kxe4 Bb8 =
I tried several variations from here, but I never managed to win with White.
albitex albitex 12/1/2025 07:59
CharlesSullivan, Yes, 36...Ke6 is better, however it is not enough, White always wins (if he plays like a computer with tablebases..)
albitex albitex 12/1/2025 07:50
CharlesSullivan 43. Kh3 win
CharlesSullivan CharlesSullivan 11/29/2025 10:25
Reader albitex also gave the hypothetical 38...Be5 39.b7 Rb5 40.Ra6+ Kf7 41.Rb6 Rxb6 42.Nxb6 Ke6. Question: How should White continue [43.Nc4 or 43.h4 or 43.Kh3]? (And aren't you just a little bit surprised?)
CharlesSullivan CharlesSullivan 11/29/2025 05:24
If 39...Bd6, Kortschnoj's analysis indicates a win after 40.Ra6. Do you think so?
Also, one analyst gives 48...Ke7 49.Rd7+ Kf6 50.Rd5 Bc7 and "Karpov may well have held the draw if he had maintained a passive stance." Any thoughts?
CharlesSullivan CharlesSullivan 11/28/2025 11:57
"Bonus" followup quiz (try to solve without the computer):
What is the correct move after 36...Ke6! 37.Ra7 (...Kf6, ...Be7, or ...Bf8)?
What is the correct move after 36...Ke6! 37.b6 (...Bd6, ...Be7, ...g6, or ...Ke6)?
What is the correct move after 36...Ke6! 37.h4 g6 38.Rh8 (...Ra1, ...h5, ...Bd6, or ...Be7)?
CharlesSullivan CharlesSullivan 11/28/2025 08:41
As albitex has discovered, 36...Ke6 is the best try. However, 37.Ra6+ Bd6 38.b6 g6 39.h4 Rb4 40.h5 gxh5 41.b7 Ke5 42.Nb6 Bc7 43.Nd7+ Kd5 44.Rh6 Rxb7 45.Rxh7 Ra7 46.Nf6+ Ke6 47.Nxh5 wins for White (according to Stockfish 17.1 anyway). The followup to 36...Ke6 is very tricky -- I suspect that a human (without computer) playing Black would probably lose 90% of the time against the computer. So where can Black save the game in this variation?
Karsten Müller Karsten Müller 11/28/2025 05:21
albitex: Yes Kasparov is indeed right, that 38...Rb4 does not save Black. And neither does 38...Be5. And yes: 36...Ke6 seems to hold. And there are more questions open...
albitex albitex 11/28/2025 02:14
Point E)
I agree with Kasparov 38... Rb4 is not enough for a draw and neither 38... Be5 preferred by the engines, it seems that White has objectively won at this point (38... Be5!? 39. b7 Rb5 40. Ra6+ Kf7 41. Rb6 Rxb6 42. Nxb6 Ke6 +-).
Rather, I believe that Black made a inaccuracy two moves earlier; perhaps this is the crucial moment of the critical error that compromises the game for Black:
Karpov, instead of playing 36...Bd6 he had a better alternative > 36...Ke6
(36... Ke6 37. Ra6+ Bd6 38. b6 g6 39. h4 Rb4 40. h5 gxh5 41. b7 Ke5 42. Nb6 Bc7 43. Nd7+ Kd5 44. Rh6 Rxb7 45. Rxh7 Ra7 46. Nf6+ Ke6 47. Nxh5 =/+)
Maybe Black can get a draw here.
Karsten Müller Karsten Müller 11/27/2025 05:51
One recent and very important source is missing, which will be included in the solution:
Hans Renette and Tibor Karolyi, Korchnoi Year by Year, Volume II (Elk and Ruby Publishing 2024, Loan from Petronijevic)
1