Lasker-Chigorin: A legendary clash of two bishops vs two knights

by Wolfram Schön
10/12/2025 – Already in the second round of the famous Hastings tournament of 1895, two of the favourites faced each other in a tough and complex battle. After just thirteen moves, the game had reached a semi-endgame with two rooks and two minor pieces each. Lasker held the pair of bishops, but it was Chigorin with the pair of knights who emerged victorious in the end. The game made history, and its course has been interpreted very differently over time. We invite our readers to join us in the search for new insights.

Winning starts with what you know
The new version 18 offers completely new possibilities for chess training and analysis: playing style analysis, search for strategic themes, access to 6 billion Lichess games, player preparation by matching Lichess games, download Chess.com games with built-in API, built-in cloud engine and much more.

In May 1894, Emanuel Lasker somewhat surprisingly defeated the 58-year-old Austro-American William Steinitz in a 19-game match played in various cities across North America, by a clear score of 10:5. Thus, the title of World Champion passed to the then 25-year-old German. But experts were quick to express doubts about his qualifications, considering him a relatively inexperienced newcomer.

Lasker was willing to accept Steinitz's challenge for a rematch, but only at a later date. In the meantime, a tournament was to determine the true hierarchy in the world of chess. Five years after the Congress in Manchester, another major chess event was to take place in England in the summer of 1895, for which the organisers in the popular seaside resort of Hastings were able to raise an attractive prize fund.

From 5th August to 3rd September, 22 players from all over the world should meet in a single round-robin tournament. As was customary back then, the time limit was two hours for 30 moves, followed by one hour per 15 moves. Games were played regularly on five days of the week, with the sixth day reserved for any unfinished games. Each day consisted of up to eight hours of play, split into two sessions: one from 1 pm to 5 pm in the afternoon and one from 7 pm onwards. So the players had a month of hard work ahead of them.

Among the big favourites were Steinitz and Lasker, with the latter seemingly having recovered well from the typhoid fever he had suffered in autumn 1894. But so were Dr Siegbert Tarrasch from Nuremberg, who had won several strong tournaments in a row, and Mikhail Chigorin from St. Petersburg, who had lost to Steinitz twice in world championship matches, most recently in 1892 by a narrow margin. Representatives of the younger generation, such as David Janowski from Paris, Harry N. Pillsbury from Boston and Carl Schlechter from Vienna, were considered to have only an outside chance at best.

In the first round, Lasker defeated Vienna's Georg Marco in a rather erroneous game. Meanwhile, Chigorin gained the upper hand over Pillsbury by taking a wild course that even included the early sacrifice of a rook.

The following day saw the first ever meeting between two major protagonists of that era. The New York Sun, 7th August 1895, reported: "Chigorin had a much harder fight against Lasker. This contest was hotly, though evenly, conducted in the opening and middle game, neither side being able to make the least progress toward establishing a winning position. Toward the end Lasker was altogether outplayed, and he had to resign after fifty-six moves." Over the past 130 years, this game has been commented on and analysed many times. This includes contemporary chess magazines, books on middlegame strategy and more recent game collections.

Sources and abbreviations used (in chronological order):

Hof = La Stratégie (15. August 1895), p. 239, notes by L. Hoffer
DWZ = Deutsches Wochenschach (18. August 1895), p. 279
DSZ = Deutsche Schachzeitung (August 1895), p. 237
BCM = British Chess Magazine (September 1895), p. 397, notes by W. Wayte
Stei = Horace F. Cheshire: "The Hastings Chess Tournament" (London 1896), p. 30, notes by W. Steinitz
Sch = Emil Schalopp: "Das internationale Schachturnier zu Hastings" (Leipzig 1896), p. 71
Kob = Alexander Koblenz: "Lehrbuch der Schachstrategie 1" (Berlin 1980), p. 224
Sergei Soloviov: "Emanuel Lasker 1, Games 1889-1903" (Sofia 1998), p. 181
Garry Kasparov: "My Great Predecessors, Part 1" (London 2003), p. 102
KM = Linder, I. & Linder, V.: "Emanuel Lasker" (Milford 2010), p. 71, notes by K. Müller
GK = Garry Kasparov: "Moi velikie predshestvenniki, Tom 1" (2nd ed., Moskva 2020), p. 118

In the presentation of the game, I have already included a few selected excerpts from the various comments. These are intended to provide guidance on the different strategic themes, but also to point out some of the numerous controversial moments.

Click on the notation to get a replay board with engine.

There are regular disputes about the correct notation of historical games. In this case, it is about the exact moment when Lasker resigned. Of the six contemporary sources, two (Hof and DSZ) state that it was after 55...Rag1; three (DWS, BCM and Stei) after 56...Ke8; and one (Sch) after 57.Bg5 R6xg5. We stick to the majority and the daily press (New York Sun), according to which the game ended after 56 moves.

Contemporary commentators were therefore of the opinion that Lasker had gained an advantage from the opening, but that a balanced position had emerged after 19...f5 at the latest. As the game progressed, Chigorin handled his pair of knights (significantly) more skilfully than Lasker did his pair of bishops. Some expressed their appreciation, like "the whole game is played with great strategic finesse" (Hoffer) or "a masterly game, worthy of two great players" (BCM). Others treated Lasker very critically, such as Steinitz and Levenfish.
For a long time, Steinitz was the only one who at least questioned some of Chigorin's decisions. Koblenz, for example, summarised: "However, if the position is blocked, the pair of bishops proves to be a fiction. The bishops do not always represent an advantage. This game illustrates what has been said." From the 1990s onwards, the overall assessment began to change. According to Soloviov, followed by Kasparov in 2003, Lasker owned (significant) advantages during most of the game.

GK's concluding commentary from 2020 is quite nuanced and summarises all the findings to date: "In this ultra-tense, far from faultless duel, it is not so much variants that are important as the text of the game itself: Chigorin was after all playing with the world champion, and, as always, he upheld his principles! In the given instance, this was a struggle of two knights against two bishops in a semi-closed position and the blockade of a pawn centre. And on the whole, although Lasker twice in the course of the game gained the better chances, he did not demonstrate a superiority in a very complex game rich in tactics."

Much to the misfortune of human chess players, recent decades have shown that, thanks to increasingly powerful chess engines, the detection of small tactical details can nullify even the most profound strategic considerations. Therefore, it is not surprising that, in 2020, GK changed many of his lines and evaluations from 2003, overturning some of them entirely. However, he refrains from making clear assessments in some places, and at least one very important moment has not yet been identified at all.

So, it is time to get down to work and take a closer look at some of the game's turning points. In an analysis, it is often very helpful to tackle the simple questions first and then work your way up to the more complex problems. In chess, this usually means starting the analysis of a game from the end.

Question 1

Almost all commentators consider White to be hopelessly lost here. Lasker played 55.Rxd3 and had to resign two moves later.

Instead 55.Bc7 Ra2+ 56.Kf1 Rgg2 57.Rxd3 Rxh2 58.Kg1 is given by Steinitz as well as Kasparov as the best defence. While the 1st world champion says "and it is not so clear that Black can win", the 13th states "the struggle is prolonged for a long time".

Can you confirm GK's assessment that Black is winning?

Question 2

At the time, the consensus was that "White's game cannot be saved", as reflected in Hoffer's comment on Lasker's 48.Rb5.

As evidence, for example DSZ cites: "48.Rxc4 fails due to 48...Nxd4+ etc., and even after 48.Bc2 exd4 49.Bxd4 Nxb4 50.cxb4 Rh5, White is lost." Both alternatives are also discussed briefly by GK: "Perhaps more persistent is 48.Rxc4!? Nd6 49.Rxc6 Rxc6 50.dxe5 Rxe5 51.Ba2 or 48.Bc2!? exd4 49.Rxc4 Nd6 50.Rxc6 Rxc6 51.cxd4, relying on the strength of the bishops." Of course, we want to be more precise.

Can White avoid losing with 48.Rxc4 and/or 48.Bc2?

Question 3

This is the position in the game that was talked about the most, with everyone condemning Lasker's 47.Rd2.

Among the classical commentators, some recommended 47.Ba2 (such as Steinitz), others 47.d5 (such as Chigorin), but most favour 47.Bc2. Later, Levenfish added 47.dxe5 to the mix. Kasparov considered all of these suggestions and concluded: Best is 47.Bc2! Nc6 48.Rbb1, which "still left White the better chances."

Can you work out if GK's approach will result in a winning position?

Question 4

So far, Chigorin's knight retreat 37...Nf6 has gone almost uncommented. Instead, a viable option to consider is 37...Nc7 38.Bxf4 e5. GK offers a minimal insertion: "37...Ne3!?". Before him, Karsten recommended: The pair of bishops "should be halved by 37...Ne3 38.Bxe3 fxe3 39.Kxe3 Ra5 with better drawing chances than in the game."

Can any of the two ideas bring salvation for Black?

Question 5

This is truly uncharted territory. No one has dared to say anything about Chigorin's 36...gxf4 yet. There were some critics of the move before, 35...g5, from Steinitz to Kasparov, as this "ultimately opens the position for White's bishops", as Karsten puts it. However, in the given situation, engines are quick to flash out either 36...h6 or 36...g4 as improvements.

What are your thoughts on these suggestions? Do you think any of them will really make a difference?

Further considerations

It almost seems as if there are opportunities for improvement, or at least respectable alternatives, in many of the moves in the game. Like, instead of 50.Bh4, there are two candidates available: 50.Bd4 and 50.Rdd5. Or, instead of Lasker's 42.Raa1, the central pawn push 42.e5 did not escape the eye of the commentators, from Chigorin to Kasparov. Was Steinitz right in saying that 33...g5 is better than Chigorin's realisation of the same plan with 35...g5?

Of course, there are also plenty of questions about the early stages of the game. Was Lasker really already winning after the queen exchange on move 13, as Levenfish and Kasparov suggest? Did Chigorin play flawlessly from 18...c4 to 32...Nf7, as all commentators, including Kasparov, claim?

Whether it is the impression of a particular moment or analytical ‘proof’ that this or that move wins or not. Please feel free to explore what interests you most and share your insights with us!


Wolfram Schön, born 5.6.1967, is an International Master 1987 and a grandmaster in correspondence chess since 2006. Greatest success: third place in the Correspondence Chess World Championship 2003-05. Wolfram is a programmer and an asset manager by profession.
Discussion and Feedback Submit your feedback to the editors


Discuss

Rules for reader comments

 
 

Not registered yet? Register

Karsten Müller Karsten Müller 10/17/2025 08:18
albitex: Many thanks for your hard work and for pointing out my move numbering mistake! Now we agree on Q1, Q2 and Q3. What about Q4 and Q5?
albitex albitex 10/16/2025 05:09
45. Bb1 what? Instead of 45. Bd1?
You're probably referring to 54. Bb1 instead of 54. Bd1 of my variant.
In that case, you're right. The maneuver of moving the white King to a3 to open the position and gain a pawn proves successful:

47. Bc2 Nc6 48. Rbb1 Ra6 49. h4 Rd7 50. h5 Rd6 51. d5 Na5 52. Rh1 Rh6 53. Rb4 Ke8 54. Bd1
(54. Bb1! Nd6 55. Ba2 Kd7 56. Kd2! Ke8 57. Kc2 Kd7 58. Kb2 Ke8 59. Ka3! Kf7 60. Bb1 Nab7 61. Bc2 Nc5 62. Rbb1 Ra5 63. Rh2 (63. Kb4? Rxa4+ 64. Bxa4 Nd3+ 65. Ka3 Nxf2) 63... Ne8
(63... Nd7? 64. Rg1 Nc5 65. Rg5 Nxa4 66. Rhg2 Nc5+ 67. Kb2 Nd7 68. Rg7+ Ke8 69. Rg8+ Nf8 70. R2g5 Nf7 71. Rf5 +-)
64. Kb4 Na6+ 65. Kxc4 Nd6+ 66. Kb3 Nc5+ 67. Bxc5 Rxc5 68. Rg1 +-)

Here we can see the negative impact of blocking the Rook on a6: as white's King approaches the a3-square, Black has no counterplay; both of his Rooks are blocked, and he can't move his Knights because are holding the defense.
Ultimately, the human assumption that trapping the Rook on a6 wasn't a good idea turns out to be true.
MickyMaus90 MickyMaus90 10/16/2025 12:25
albitex: Many thanks from me as well, you are giving us a lot of input!
Q3: Yes, Karsten showed the way with 54.Bb1 to your latest suggestion. I have elaborated a bit on it to produce a typical line:
47.Bc2 Nc6 48.Rbb1 Ra6 49.h4 Rd7 50.h5 Rd6 51.d5! Na5 52.Rb4! Rh6 53.Rh1Ke8 54.Bb1 (Karsten) Kd7 55.Ba2 Nd6 56.Kd2 Ke8 57.Kc2 Kd7 58.Kb2 Ke8 59.Ka3 (now White can start the regrouping of his Ba2 and Rb4 from the queenside) Kd7 60.Bb1 Nf7 61.Bc2 Ng5 62.Bd1 Nf7 63.Rb1 Nb7 64.Bc2 Ke8 65.Rbg1 Kf8 66.Kb4 Nbd6 67.Bd1 Ra5 68.Be2+- (the c4-pawn will be gone soon)
Before that my mainline began with 48...Rd7 49.h4 Nd6 50.Rg1 Rg7 51.Rxg7. White's strategy is the same to squeeze out Black, but with one pair of rooks less on the board. So, probably 48...Ra6 will become the new mainline.
Q2: I am sorry, but here I have to pick up a point from yesterday. You claimed after 48.Rxc4, there is no objective way for White to hold. Do you have more info about this?
I still think that GK's 48.Rxc4 Nd6 49.Rxc6 Rxc6 50.dxe5 Rxe5 51.Ba2! is the best and Black can't win here.
Karsten Müller Karsten Müller 10/16/2025 03:37
albitex: Many thanks for your hard work! I suggest 45.Bb1 (instead of your Bd1) 45...Kd7 46.Ba2 Nd6 and now march with White's king to a3 to free White's forces.
albitex albitex 10/15/2025 10:33
Sorry: Let's see a possible development of 49. h4 (not h5 us I write)
albitex albitex 10/15/2025 10:31
Karsten Muller I disagree. The move 49. h4 doesn't seem decisive to me.
My Stockfish gives me +0.9 at 55 depth of analysis.
Let's see a possible development of 49. h5:
47. Bc2 Nc6 48. Rbb1 Ra6 49. h4 Rd7 50. h5 Rd6 51. d5 Na5 52. Rh1 Rh6 53. Rb4 Ke8 54. Bd1 Nd6 55. Kd2 Nab7 56. Be2 Nc5 57. Bxc5 bxc5 58. Rb8+ Ke7 59. Bd1 Rb6 60. Ra8 +=
I don't see how White can gain a decisive advantage here.
Karsten Müller Karsten Müller 10/15/2025 05:16
albitex: I agree completely. I would also never consider 48... Ra6 in a game. And your point "The engines, however, often seem to elude these principles." is deep and good. Principles can help, but chess is a concrete game after all.
But my question still remains: What do you think about 49.h4! ?
albitex albitex 10/15/2025 04:10
Allow me to comment on the difference between human thought and engine analysis:
with reference to my variants on question 3) the engines suggest the move 48... Ra6 to prevent d5, considering the a5-square the only good escape route for the Knight-c6 and judging the alternative 48... Ng5 inferior.
I in a real game (like I believe most low/medium level players) therefore without unlimited time for analysis, would never have considered 48... Ra6: trapping the Rook and wasting a time to freeing it and place a Knight on the edge of the chessboard would be an option I would have refuse regardless. In a real game I would have analyzed only 48... Ncd8, 48... Ne7, 48... Ng5.
With the limited time of a real game, we are forced to prune the tree of variations, and the strategic principles we have studied make you discard possibilities. The engines, however, often seem to elude these principles.
Karsten Müller Karsten Müller 10/15/2025 03:57
albitex: You work really hard. Many thanks! Wolfram Schön gives after 47. Bc2 Nc6 48. Rbb1 Ra6 the preparatory 49.h4! to win space on the kingside first. SF 17 gives more than +2 already. What do you think?
albitex albitex 10/15/2025 03:03
Question 3)
The move 47. Bc2 suggested by Kasparov is obviously good, but I have not been able to find a clear way for White to win:
47. Bc2 Nc6 48. Rbb1 Ra6 {This move seems mandatory, White threatens d5 and for the engines the only good escape square for the Knight is a5} 49. dxe5 Nfxe5 50. Rd6 Ra5 51. Rb5 Rg7 52. Rf6+ Rf7 53. Re6 Re7
(53... Rg7 54. Rxa5 bxa5 55. Rf6+ Rf7 56. Rf5 Ke8 57. h4 h6 +=)
54. Rh6 Ke8 55. Rxa5 bxa5 56. h4 Rf7 57. Bd4 +=)
*
White has an advantage, but it's not decisive. I can't find a way to win; it seems Black can hold out.
What surprises me is that the bishop pair should be an advantage, but in this game, that advantage isn't apparent.
Karsten Müller Karsten Müller 10/15/2025 02:10
albitex: Many thanks for your great work! So finally we all agree on Q1 and Q2. What about the other questions?
albitex albitex 10/15/2025 01:53
Karsten Müller after 63... Rg7 of the Wolfram Variation: Black exchanges Rooks and his two passed pawns to the board edges leave no escape for White:
55. Bc7 Ra2+ 56. Kf1 Rgg2 57. Rxd3 Rxh2 58. Kg1 Rhe2 59. Bd6+ Ke8 60. Rb1 Rg2+ 61. Kh1 Rh2+ 62. Kg1 Rag2+ 63. Kf1 Rg7
Now White has two plausible moves:
- 64. Rbd1
- 64. Re1
in both cases Black forces the exchange of Rooks and White succumbs to black's passed pawns at board edges:
- 64. Rbd1 Rh1+ 65. Kf2 Rxd1 66. Rxd1 Rd7 67. c4
(67. Rd5 Ne7 68. Rd4 Ng6 69. Rd5 h5 70. e5 Kf7 71. Rb5 Ke6 72. c4 Nh4 73. c5 bxc5 74. Rxc5 Rb7 75. Rc6 Rb6 76. Rc3 Rb2+ 77. Kg1 a5 -+)
67... a5 68. Rd5 Ne7 69. Rd2 Ng6 70. Rd5 a4 71. Ke2 h5 -+)
*
- 64. Re1 Rh1+ 65. Ke2 Rxe1+ 66. Kxe1 Rd7 67. Ke2
(67. Rd5 h5 68. c4 Ne7 69. Rd2 a5 70. Kf2 a4 -+)
67... a5 68. Bxf4 Rxd3 69. Kxd3 b5 -+)
albitex albitex 10/15/2025 11:20
Wolfram Schön: You're right, after 58... Ke7 59. Rxf4 h5, Black wins. I've spent hours looking for improvements, but I've found nothing. Even after 48. Rxc4, I can't find an objective way for White to hold.
Kasparov was right; here, with correct play, White has already lost; Black's pawn structure is superior. But, then again, who can doubt it? Kasparov is Kasparov...
Karsten Müller Karsten Müller 10/15/2025 09:29
arzi: Q2: 50...Ra1 is stronger than 50...Re7? as given by Wolfram Schön. Then Black wins.
MickyMaus90 MickyMaus90 10/14/2025 12:46
albitex: Q2: You are right, after 48.Rb5 Rxa4 49.dxe5 Nfxe5, Lasker's 50.Bh4 is not the best.
50.Rdd5 was recommended by the Bulgarian team led by S.Soloviov, in 1998. But after 50...Ra1!, your 51.Bh4 is more critical than the old 51.Be1.
My line is the same as yours for a long time: 50.Rdd5 Ra1 51.Bh4 Rg7 52.Bf6 Rg2+ 53.Kf1 Rxh2 54.Bxe5 Nxe5 55.Rxe5 Rxb1+ 56.Rxb1 Rh1+ 57.Ke2 Rxb1 58.Rf5+. But now not 58...Ke8, but 58..Ke7! 59.Rxf4 h5! What do you think?
Your comment about Chigorin's 53...Nd3+ being dubious is also correct. KM and GK proposed 53...Rh6 as an improvement. But I agree, 53...a6! is probably the most easy win.
arzi arzi 10/14/2025 12:41
Question 2: I agree with albitex that Lasker`s 48.Rb5 is the best move and the game should be ended a draw because once again black king has no chance to protect black pawns in Queen side of the board. All the pawns will be changed off the board and only King and rook are left in both side.

48. Rb5 Rxa4 49. dxe5 Nfxe5 50. Rdd5 Re7 51. Bd4 Ra1 52. Bxe5 Nxe5 53. Rxe5 Rxe5 54. Rxe5 Rxb1 55 Rf5 Ke8 56. Rxf4 a5 57 Rh4 b5 58. Rxh7 b4 59. cxb4 axb4 60 Kd2

Now white has 3 pawns and black 2 pawns. With the help of his h -pawn (h7) white is able to take both black pawns.
MickyMaus90 MickyMaus90 10/14/2025 12:27
arzi: Your analysis is completely correct.
After 55.Bc7 Ra2+ 56.Kf1 the move 56...d2 was already commented on before as a sideline. Steinitz: "57.Rb1 and Rb1-d1 next." Kasparov: "57.Rb1 Na5 58.Bxf4". Therefore, your line confirms that 56...d2 is a mistake.
Yes, 56...Rgg2 is the right way. My mainline is exactly the same as yours (with one transposition of moves) up to 66...Rd7! The point is, I think, that the black rooks on the 2nd rank don't win on their own (blind swines!). In fact, in the end, Black's two outside passed pawns on the a- and h-file are the decisive factor.
(MickyMaus90 = Wolfram Schön)
Karsten Müller Karsten Müller 10/14/2025 11:58
albitex: Question 1: You really work hard and long! Many thanks! Wolfram Schön gives: 58.Kg1 Rhe2 59.Bd6+ Ke8 60..Rb1 Rg2+ 61.Kh1 Rh2+ 62.Kg1 Rag2+ 63.Kf1 Rg7 ... -+
What do you think?
albitex albitex 10/14/2025 11:38
Regarding question 1), I can't add anything more to Kasparov's opinion. Yes, Black has a theoretical advantage, but after two hours of analysis with Stockfish, I haven't been able to find a demonstrable way for Black to win.

Question 2) In my opinion 48. Rxc4 or 48. Bc2 are no better than 48. Rb5. Lasker played the better move, 48. Rb5 is a good attempt to free the Bishops, however Lasker made a mistake after the 50th move, he played Bh4 too early:
In the game played:
48. Rb5 Rxa4 49. dxe5 Nfxe5 50. Bh4?
This move is a mistake, played too early, as it allows Black to immediately play Rg7 which then prevents Bf6.
Now Black could take a clear lead:
50...Rg7 51. Kf2 Rg6 52. Rdd5 Ra1 53. Bd8 Nd3+?! (53... a6! 54. Rb2 b5 -+)

But if Lasker had instead preceded 50. Rdd5 to Bh4, thus preventing Black from immediately playing Rg7 and thus being able to play his bishop on f6, we would have ended up with a R + 3 pawn Vs R + 4 pawns endgame, a drawn endgame:
48. Rb5 Rxa4 49. dxe5 Nfxe5
(50. Rdd5 Ra1 51. Bh4 Rg7 52. Bf6 Rg2+ 53. Kf1 Rxh2 54. Bxe5 Nxe5 55. Rxe5 Rxb1+ 56. Rxb1 Rh1+ 57. Ke2 Rxb1 58. Rf5+ Ke8 59. Rxf4 =)

In conclusion: according to Stockfish, it's completely untrue that the game couldn't have been saved.
In my oersonal opinion, Lasker played the best move with 48. Rb5, and according to Stockfish, he could have drawn; he only missed one tempo.
Karsten Müller Karsten Müller 10/14/2025 08:26
arzi: Good start! Indeed your first line should be correct and so 55 Bc7 Ra2 56. Kf1 d2?= a mistake. So your second line is the solution to question 1 and Black wins. It was given by Steinitz and Kasparov until 58.Kh1 -+. So nothing new under the sun...
Karsten Müller Karsten Müller 10/14/2025 02:54
albitex: Good point of course. Usually the knights wants statical control to have time for their manovers and to open roads for them. The bishops want dynamics and open diagonals and want to restrict the knights. Here Black has the initiative and the bishops have problems to get really activated.
albitex albitex 10/14/2025 02:25
I ask myself: why in this situation the pair of Bishops, which should be an advantage, is of little use and the Knights are stronger?
arzi arzi 10/13/2025 07:23
Question 1: 55 Bc7 Ra2 56. Kf1 d2 57. Rb1 Na5 58. Bxf4 Nc3 59. Rd1 Rf6 60 Bxd2 Rxf4 61. Kg1 Nxd2 62. R5xd2 Rxd2 63. Rxd2 Rxc3 64. Rd7 a5 65. Rxh7 a4 66. Ra7 b5 67. Kf2 a3 68. Ke1 b4 69. Kd2 Rh3 70 Kc1 Re3 71. Kb1 Rxe4 72. h4 Rxh4 73. Ka2 Rf4 74. Kb3 or Rb7 draw because black king can not go over 7th line.

55 Bc7 Ra2 56. Kf1 Rgg2 57.Rxd3 Rxh2 58.Kg1 Rhe2 59.Rb1 Rg2 60. Kh1 Rh2 61. Kg1 Reg2 62. Kf1 Rg7 63. Bd6 Ke8 64. Rbd1 Rh1 65. Kf2 Rxd1 66. Rxd1 Rd7 67. Rd5 Ne7 68. Rd4 h5 Black might be winning here.
1