I recently got Huebner’s 2008 book on this match. As anyone who flipped through an <Informator> from the 1980s will remember, Huebner’s annotations are formidably long and dense, but worth the effort.

So, here are his annotations game #9 from the book, supplemented by Andy Soltis’ <Why Lasker Matters> (2005) and J.G. Cunningham’s <The Games of the Steinitz-Lasker Championship Match> (Leeds 1894), which compiles contemporary annotations. Huebner himself made use of a wide array of contemporary and subsequent publications, including Soltis and Cunningham. I have omitted his footnotes.

Huebner’s comments are in brackets. The translation is mine, making use of wordreference.com, Google Translate, and Webster’s New World German Dictionary. I quoted German I was unsure about (there was a lot of it). My own comments are in plain text.

I doubt I will translate his notes for all of the games — it’s just too much work. But I strongly recommend his book, if you have any German at all.

1.e4 e5 2.Nf3 Nc6 3.Bb5 d6

Soltis points out that Steinitz recommended 3….d6 in <The Modern Chess Instructor> (1889) because 3….Nf6 (or 3….a6 4.Ba4 Nf6) because they prevented Black from playing what he regarded as the best line: 4.c3 f5!?. <On that remarkable basis, 3….d6 became Steinitz’s main defense in the Lopez.> The funny thing is, there are no examples of Steinitz playing either side of 1.e4 e5 2.Nf3 Nc6 3.Bb5 d6 4.c3 f5 in the database.

4.Nc3 a6 5.Bc4

<This is quite feeble. Stronger is 5.Bxc6+ bxc6 6.d4 with sharp play.>

5….Be6 6.Bxe6 fxe6 7.d4 exd4 8.Nxd4 Nxd4 9.Qxd4 Ne7

click for larger view

10.Bg5

<This furthers Black’s development. I prefer <Besser gefallt mir> 10.Be3 Nc6 11.Qc4 Qd7 12.f4 0-0-0 13.0-0-0 (after 13.0-0 Black gets counterplay with 13….g5) and g2-g4 with a slight advantage for White. >

10….Nc6.

<“Practically forced; White threatened 11.Bxe7 with a very strong position,” said Lasker. This is an error in judgment <Urteilsfehler>; in the resulting position Black’s bishop is stronger than White’s knight. Nevertheless Steinitz’s move is naturally the best.>

11.Bxd8

<A. Soltis recommends 11.Qd2 Be7 12.Be3 <schlagt 11.Dd2 Le7 12.Le3 vor> and thinks that White then has a slight advantage; I believe, however, that Black after 12….Bf6 is a tiny bit more comfortable <Schwarz nach 12….Lf6 eine Winzigkeit bequemer steht>. Equality naturally results from 11.Qe3 Be7 12.Bxe7 Qxe7 13.f4. The text moves gives Black a slightly better ending. >

11….Nxd4 12.0-0-0

click for larger view

12....Nb5 <(?)>

<This move is artificial and has incurred universal disapproval <allgemeines Missfallen errecht>. The following other possibilities come into consideration:

12….c5 13.Bg5 Be7 is Chigorin’s suggestion; 14.Be3 Bf6 15.f4 follows. Rhe1 with Nb1 and c3 if occasion arises <gegebenfalls> leaves White a little more comfortable. 12….Nb3+ was recommended by Steinitz; after 13.axb3 Rxd8 14.f4 the position is about equal. The simplest is the strongest: 12….Rxd8 13.Rxd4 g6 (this is a little better than 13….Be7, which was given by A. Soltis and L. Pachman). Black’s bishop is superior to White’s knight; after completing his development Black continues with …c6 and …d5 with slightly better play.

The text move leads to equality.> It’s not entirely clear how chess masters viewed knights and bishops in 1894, but if the Cunningham book is any indication they definitely believed that having a knight against bishop was an advantage in the ending. This may explain why Steinitz didn't play …Rxd8. As for why he preferred …Nb5 to …Nb3+, it seems he thought he could make use of the a-file.

13.Nxb5 axb5

<Black can no longer get an advantage with 13….Rxd8 <Mit 13….Td8 kann Schwarz keinen Vorteil mehr erreichen>; White answers with 14.Nd4, and now:

14….Kf7 15.f4 g6 16.f5 gxf5 17.exf5 exf5 (Black should not let the opposing knight establish itself permanently on e6) 18.Nxf5 and White has a slight advantage; the black king is unsafe. 14….Kd7 15.f4 g6 16.Nf3. White continues with g3 and h4; he is slightly more comfortable. Whether the knight is on f3 or c3 makes a big difference! 14….e5 leads to equal play.

After the text move the black pawns are isolated <Nach dem Partiezug kommt eine Verinselung der schwarzen Bauernstruktur zustande>; the little groups at b7 and b5, d6 and e6, g7 and h7 constitute two more pawn islands than the opponent has. At the moment this is of slight importance; the game remains in balance.>

14.Bxc7

click for larger view

14….Rxa2 <(?)>

15.Bb6

<White must meet the threat of 15….Ra6. Naturally 15.Bxd6 fails to 15….Ra1+ 16.Kd2 Rxd1+ followed by …Bxd6, as well as 15…..Bxd6 16.Kb1 Ra6.>

15….Be7

<“Weak,” said Steinitz, who thought that 15….b4 would have made White’s game difficult. That is of course quite wrong <Davon kann naturlich keine Rede sein>. White has a variety of good alternatives.

I. 16.Bd4

16….Kf7 17.Kd2

Aa) 17….e5 18.Be3 Rxb2 19.Rb1 with advantage for White.

Ab) 17….Be7 18.Ra1 Rha8 19.Rxa2 Rxa2 20.b3, and White stands better.

B) 16….Be7 17.Kb1 Ra6 18.Bxg7 Rg8 19.Bd4 Rxg2 20.Rdg1 Rg6 21.f4 with some advantage for White.

II. 16.Kd2 Rxb2 17.Bd4 Ra2 18.Ra1 b3 19.Rxa2 bxa2 20.Ra1 and White has the advantage.

The move played is no worse than 15….b4.>

16.c3

click for larger view

16….Kf7 <(?)>

<After this Black’s disadvantage starts becoming noticeable; Steinitz considered this the losing move.

I. He proposes 16….Kd7 17.Bc5 Kc6 18.Ba3 (no better is 18.Kb1 Rha8 19.Ba3 R8xa3 20.bxa3 Rxf2) 18….Ra1+ 19.Kd2 Rxd1 with at least equal play, but again he is not objective in his valuation of his position <aber wieder lasst er es an Objektivitat im Bezug auf den Wert der eigenen Stellung fehlen> . After 20.Rxd1 the threat of e4-e5 hangs over his head; after 20….e5, however, Black has fae better prospects for salvation than later in the game.

II. Lasker proposes 16….Ra4 17.Rhe1 b4; after 18.Kc2 bxc3 19.bxc3 White has a slight advantage, because his pawn structure is favorable, and he may develop pressure against b7.

III. The most precise is perhaps 16….b4. After 17.cxb4 Bf6 18.Rhe1 Ke7 19.Bd4 Bxd4 20.Rxd4 Ra1+ 21.Kd2 Rxe1 22.Kxe1 Ra8 White’s extra pawn is without significance.>

17.Kc2 Rha8 18.Kb3 R2a4 19.f3 R8a6 20.Bd4 g6 <(?)>.

21.Rd3 Ke8

<Soltis advises 21….d5, but I cannot agree. After 22.exd5 exd5 23.Be5 will be lost without compensation, and Black remains with the worse position; his situation is hopeless. Again 21….g5 came into consideration.>

22.Rhd1

click for larger view

22….e5 <(?)>

<No one approved of this move, although some commentators thought it was unavoidable in the long run. I quote some examples:

I. “On 22….Kd7 23.f4 would follow, and g6 and h7 would be targets,” wrote Lasker.

II. Steinitz recommended 22….g5.

III. “If 22….Kd7, then 23.Bc5! Rc4 (on 23….Kc7 there follows 24. Bb4 with the threat 25.e5) 24.Bxd6 Rxd6 25.Rxd6+ Bxd6 26.e5 with a winning position. In place of the text move, it was better to play 22….Rc4 and then …Ke8-d7” indicates Chigorin.

IV. “Obviously a very bad move; after this the black pawn position remains very weak. He had two other plausible possibilities at his disposal:

A) 22….Kd7 23.Bc5 Rc6 24.Bxd6 Rxd6 25.Rxd6+ Bxd6 26.e5 Ra6 27.f4 Kc7 28.exd6+ Rxd6 29.Rd4! with a better endgame. If Black exchanges rooks, he is lost: 29….Rxd4 30.cxd4 Kc6 31.Kb4 Kb6 32.g3 h6 33.h3 h5 34.h4 Kc6 (or 34….Ka6 35.Kc5 Ka5 36.Kd6 and wins) 35.Ka5 Kd5 36.Kxb5 Kxd4 37.Kb6 Kc4 38.Kxb7 Kb3 39.Kc6 and wins. In the rook endgame following 29….Ra6, White has a considerable advantage: 29….Ra6 30.Re4 Kd7 31.Re5 Ra5 (or 31….Rb6 32.Kb4) 32.g3 winning a pawn.

B) 22….Rc6 23.Bf2 Kd7 (23….Raa6 24.Kb4!) 24.Bg3 Kc7 25.Rd4 Rxd4 26.Rxd4 Rc5 27.Kb4 with a big advantage.

[bad FEN: 8/1pk1b2p/3pp1p1/1pr5/1K1RP3/2P2PB1/1P3PP/8]

The text move weakens the d-pawn for all time, and also assists White’s attack on the b-pawns.” So declare R. Fine and F. Reinfeld.

V A. Soltis makes a selection from these examples. He asserts that 22….Rc4, 22….g5, and 22….Rc6 are better than the text move. He takes continuation B from Reinfeld’s book, followed by 27….Kc6 and quotes Benko’s judgment that, although Black stands worse, it isn’t by much.

I would like to add some comments to these lines <Einige Hinweise zu diesen Angaben mochte ich beisteuern>.

Zu I. After 22….Kd7 23.f4 there follows 23….Ke8 with the equalizing threat of …e5. White has lost the greater part of his advantage now that the e4 pawn has come under attack. The move 23.Bc5 is stronger than 23.f4.

Zu III A: in place of 27.f4, White can play 27.Rd4; after 27….Kc7 28.exd6 Rxd6 29.Kb4 Ra6 30.Re4 he wins without difficulty.

Zu III B: The final judgment given in Section IV seems to me to be correct, but 27.cxd4 is much stronger than 27.Rxd4. After 26.cxd4 Rc4 27.d5 e5 28.h4, Black cannot be saved. Also 25….Rxd4 is a mistake; after 25….Rca6 it is not readily apparent how White can make progress.

Black must attempt to hold the position in balance <die Stellung unverandert zu halten>. For this there are two methods; one is given above, the other is the continuation 22….g5 23.Bf2 Rc4 24.Bg3 Rcc6, and 25.Kb4 accomplishes nothing on account of the response 25….Rc4+ 26.Kxb5 Rc5+ 27.Kb4 Rb6+ 28.Ka4 Rxb2 29.Bxd6 Rc6 30.Bb4 Rxg2, etc. I think that one cannot yet speak of a winning position for White. <Ich meine, dass man noch nicht von einer Gewinnstellung fur Weiss reden kann.>

23.Be3 Kd7 24.Bc5 Ra1

click for larger view

25.R1d2

<If White allows the exchange of rooks, the black king can settle on c6. Then it is not easy for White to attack the Black queenside pawns nor open up the position; Black holds on. >

25…Ke6

<After 25….g5 26.Rd5 Black must bring the king to e6; the position after 26….Kc7 27.Bb4 Rb6 28.Kc2 is completely passive.>

26.Ba3 g5 27.Rd5 Rb6 28.Kb4

<Now Black gets counterplay against White’s kingside pawns. It is logical <Es liegt nahe> to eliminate this possibility with 28.h3. After 28….h5 29.Kb4 g4 30.hxg4 hxg4 31.Ka5 White stands to win:

I. 31….Ra6+ 32.Kxb5 Rf1 33.Rd1 Rf2 34.R5d2 g3 (after the exchange of rooks White’s rook would invade decisively on the h-file) 35.Kc4 Bg5 36.Rxf2 gxf2 37.Kd3 followed by 38.Ke2, and White wins.

II. 31….Bd8 32.Rxb5:

click for larger view

A) 32….Ra6+ 33.Kb4 g3 34.Kc4 (34.Rxb7 fails to 34….R6xa3) 34….Bg5 35.Rdd5 Bf4 (on 35….Rg1 follows 36.Bxd6) 36.Rxb7 Rg1 37.Kb5, and White wins.

B) 32….Rxb5+ 33.Kxb5 Bc2 34.Re2 gxf3 (34….g3 35.Rd2 Rf1 36.Kc4 is hopeless for Black) 35.gxf3 Rf1 36.Re3 d5 37.exd5+ Kxd5 38.c4+ Ke6 39.Bc5 and White will win eventually.

Nevertheless, after 28.h3 Black has still another defense, consisting of 28….Rb1. Now it is difficult for White to make progress:

I. 30.Rxb5 fails to 30….d5+ 31.Ka4 Rxb5 32.Kxb5 Bxa3.

II. 30.Ka5 Ra6+, and 31….Kxb5 would be answered with 31….Rxa3.

III. 30.Rd1 Rxd1 31.Rxd1, and now:

A) 31….d5+ 32.Ka5 Ra6+ 33.Kxb5 Bxa3 34.bxa3 Rxa3 35.exd5+ Kd6 36.Kb4 (36.c4 Rb3+ 37.Ka4 Rb2 is not effective; on 38.c5+ there follows 38….Kxc5 39.d6 Ra2+ and …Ra8) 36….Ra2. White’s advantage is probably not sufficient for the win.

B) 31….Kd7 32.Rd5 Kc6. I do not see how White can make progress.

The text move gives Black the choice of two defensive possibilities.>

28….g4

click for larger view

29.Ka5

<Some analyses of this move: <Hierzu gibt es einige Anmerkungen.>

“Perhaps it was better to first accept the offered pawn: 29.fxg4 Re1 30.Ka5 Bd8 31.Rxb5+ Ra6+ 32.Kb4 Rxe4+ 33.Kb3, and White retains an extra pawn,” wrote Lasker in <Common Sense in Chess> in 1896. In the German edition from 1924 this variation is deleted without a replacement.

He was right to do so. After 29.fxg4 Re1 30.Ka5 Bd8 31.Rxb5 Rc6+ 32.Ka4 Bg5 33.Rd3 Rxe4+ White’s advantage is very small. On 34.Rb4 follows 34….Rxb4+ and 35….d5, and the black e-pawn becomes dangerous.

A. Soltis gives the continuation 29.Rxb5 Rxb5+ 30.Kxb5 gxf3 31.gxf3 Rf1 32.Kb6 Rxf3 33.Kxb7 Re3 34.Kc6 and White wins, but fails to reach a correct assessment of the overall position <ohne zu einer Gesamtbewertung der Stellung vorzudringen>. Already W. Pollock in 1894 recognized that 29.Rxb5 is best answered by 29….d5+, and after 30.Ka4 Ra6+ 31.Kb3 Bxa3 32.exd5+ Kd6 33.bxa3 R1xa3+ 34.Kc2 (on 34.Kc4 follows 34….R6a4+ 35.Rb4 b5+ 36.Kxb5 Ra5+ and Black gives perpetual check) 34.…Ra2+ 35.Rb2 (35.Kd3 Rxd2+ 36.Kxd2 Ra2+ and 37….Rxg2 is also not winning) 35….Rxb2+ 36.Kxb2 gxf3 37.gxf3 b5 38.Kb3 Ra4 and Black holds the draw without difficulty.

The text move is the most promising.>

29….Ra6+ 30.Kxb5 h5<?>

Soltis considered this the losing move, and it gets Huebner’s only unconditional question mark of the game.

<Here also Lasker wrote initially (in 1894) “Black can continue 30….Rh1 31.fxg4 Re1 32.g3 Rxe4 33.c4”

click for larger view

“But also in this case the White kingside pawns remain dangerous.” In his subsequent published analysis [<Common Sense in Chess> (1896) he limited himself to reprinting the variation without any evaluation, and 28 years later [in the German edition of <Common Sense in Chess> (1924) he offered the following additional considerations: “that would make h2 [h7?] weak and subject to attack.”> <und nach weiteren 28 Jahren Uberlegungzeit aussert er sich so: “Dadurch wird h2 unbeweglich gemacht und soll Angriffsziel werden.”>

<A. Soltis considers White’s advantage at the end of Lasker’s analysis to be slight, and in my opinion he is correct. After 33….Re3 34.Bb4 Rb3 35.R5d3 Rxd3 36.Rxd3 Bf6 Black’s e-pawn would be strong; the white king is somewhat out of play on b5.

Also after 30….Rh1 31.fxg4 the continuation 31….Rxh2 comes into consideration.

I. 32.g3 Rh6. While loses the g-pawn and his advantage is very slight.

II. R5d3 Rh4

A) 33.Rh3 Rxg4 34.Rh6+ Kd7 35.Re2 (also after 35.Rxh7 Rxe4 Black has little to fear) 35….Rg7 and Black holds.

B) 33.Rg3 Bd8. Black can bring the black bishop into play on b6, now that White has given up the thread on d6; it is unclear whether one can speak of a winning advantage for White.

After the text move White can exchange the active Black rook, when he clearly stands to win <steht er klar auf Gewinn>.>

31.Rd1

31….Rxd1 32.Rxd1 gxf3 33.gxf3

click for larger view

33….Ra8 <(?)>

<After 33….Bf6 34.Rg1 d5 White has still has a hard task in securing the point. Now the rest is simple. >

34.Kb6 Rg8 35.Kxb7 Rg2 36.h4 Rh2 37.Kc6 Bxh4

<After 37….Rxh4 38.Bxd6 Rh2 39.Bxe7 Kxe7 40.Rb1 Rc2 41.Kd5 Kc6 42.c4 White wins without difficulty.>

38.Rxd6+ Kf7 39.Kd5

<Simpler is 39.c4.>

39….Bf6

click for larger view

<Lasker mentions the variation 39….Rd2+ 40.Kxe5 Rxd6 41.Kxd6 Bg3+ 42.f4 h4 43.Bc5 h3 44.Bg1 and White wins. A little bit tougher than the text move is 39….Bg3, but after 40.Rd7+ Kg6 (after 40….Kf6 41.Be7+ Kg6 42.b4 the b-pawn is swift) 41.Ke6 h4 42.Bf8 Kg5 43.b4 Rc2 (42….h3 44.Rh7 is hopeless for Black) 44.Be7+ Kf4 45.Rd3 h3 46.Bf6 h2 47.Bxe5+ Kg5 48.Rd1 the pawn mass must ultimately decide <muss sich schliesslich die Masse durchsetzen>. >

40.Rd7+ Kg6 41.Ke6 h4 42.Rd1 h3 43.Rg1+ Rg2 44.Rxg2+ hxg2 45.Bc5 Bd8 46.b4 Kg5 47.Kd7

The last of the white king’s many exploits in this game, driving the black bishop off the d8-a5 diagonal.

47….Bf6 48.b5 Kf4 49.b6 Black resigns.

A magnificent game by Lasker.