1. e4 e5
2. Nf3 Nc6
3. Bb5 Nf6

The Berlin Defense was popular at Munich 1900.

4. 0-0 Be7

4...Nxe4 is more usual and better. 4,,,Bc5 is also a frequent choice.

The text is also OK. Pillsbury had faced this line in a game he won against Steinitz at Vienna 1898.

5. Nc3

5. Re1 is the main alternative. Both moves yield a slight edge to White.

5... d6
6. e4 Bd7

6...exd4 was also a reasonable alternative.

7. BxN

The Tournament Book claimed that 7. Re1 was best, but that is hardly much of an improvement--if any. 7. d5 was the best alternative to the text. All of these moves leave White with a small advantage.

7... BxB
8. Qd3 exd4
9. Nxd4 Bd7
10. b3

The same move Pillsbury played against Steinitz at Vienna 1898 and that he was to play against Napier in the same position at Buffalo 1901 (another game Pillsbury won).

10. Bf4 or 10. h3 are good alternatives.

10... 0-0

11. Bb2

11. Bf4 was also OK, but having played b3 White might as well continue with Bb2.

11... Re8

Steinitz played 11...c6 here. Napier tried 11...Ng4. The text, which was also Gligoric's choice, seems as good as anything.

12. Rae1

12. Rad1 was another reasonable idea.

12... Bf8

I prefer 12...c6 to cover d5, but the text was also Gligoric's choice.

13. f4

13. Nf5 looks stronger and more in Pillsbury's style. But Parma played the text in the above-cited game against Gligoric.

The position was now:

click for larger view

Parma and Gligoric agreed to a draw in this position. Pillsbury played on and achieved a winning position within just a few moves.

So who was right--Parma to take a draw or Pillsbury to play for a win?

The answer is: they were both correct. Fritz rates the game as (0.00). Stockfish rates it as (0.02). So, objectively, a draw--especially against a formidable opponent such as Gligoric, is a reasonable result.

But this was exactly the kind of position in which Pillsbury excelled. Given that Pillsbury was playing against Bardeleben who had been sick at Munich 1900 and was having perhaps the worst tournament of his life, his decision to press for a win was surely correct.

Through his first dozen moves, Bardeleben had played well and kept Pillsbury from obtaining any meaningful advantage. But beginning on Black's 13th move, the game swung decisively in Pillsbury's favor.

13... Qe7

There was no reason to put the Black Queen on the same file as Pillsbury's e1 Rook. Bardeleben should have limited the squares on which Pillsbury's Knights could invade with 13...c6 or 13...a6 or 13...h6.

14. h3

Pillsbury might have tried 14. e5 here. But I like Pillsbury's move, which combined defense (preventing ng4 or Bg4) with offense (an eventual King-side storm).

The position was now:

click for larger view

14... c5?!

While it was usually best not to allow Pillsbury to dictate play, and while the idea of countering a wing attack by a center pawn thrust is often best policy, Bardeleben's temerity here was over-the-board. His move opened up d5 for invasion by Pillsbury. Predictably, Pillsbury accepted the invitation.

15. Nf3 Bc6?

Bardeleben's 14th move was premature, but his 15th was just bad. He should either have gotten his Queen off the d-file with 15...Qd8 (though this would have conceded that his 13th move was poor) or--better still--restricted Pillsbury's attacking options with 15...h6. After the text, Pillsbury never gave Bardeleben a chance to recover.

16. Nd5!

I'm not sure that even the greatest defensive players in history could have held this position against Pillsbury:

click for larger view

16... BxN

There is nothing better for Black.

17. exB

Bardeleben now had to find a move in the following position:

click for larger view

17... Qd7

Sergeant-Watts "SW", in their book on Pillsbury, discussed at length why 17...Qd8 would have lost after 18. Ng5 g6 (forced) 19. f5! RxR 20. RxR. While I agree with S-W's conclusion, and while some of their variations are pretty, their analysis was flawed in several respects.

The only moves for Black that S-W consider in the above variation are (A) 20...Nh5 and (B) 20...Bg7. The best try in fact would be 20...Bh6, though Black would be in bad trouble after 21. h4. But this is better than the moves SW analyze:

(A) If 20...Nh5 21. fxg6 [21. Nxf7! immediately is even better in light of 21...KxN 22. Fxg6+ hxhg 23. Rf1+] hxg6 (21...fxg6, though leaving Black in big trouble after 22. Qe3, is better than SW's 21...hxg6) 22. Nxf7! KxN 23. Rf1+ Kg8 (this move by SW is suicide. 23...Nf6 is the only way to hold on for even a while) 24. Qxg6+ Ng7 25. Rf7 and mates shortly.

(B) if 20...Bg7 21. fxg6 hxg6 (21...fxg6 would be a little better, though White still wins after 22. Qf1) then 23. Qf3! wins (but not SW's awful 22. Rf1? after which Black survives with 22...Qe7).

Returning to the actual game, the position after Bardleben'w 17...Qd7 was:

click for larger view

In the position with which I ended my last post, Pillsbury recognized and carried out a fascinating--and decisive--plan. He reduced the game to an ending in which he was effectively a piece ahead because of hopelessly blocked Black Bishop.

18. BxN

18. Ng5 would probably also have won, but having identified a winning plan Pillsbury was not to be distracted.

18... gxB
19. Nh4

Karpov would perhaps have been delighted to tie up Black with 19. c4, but Pillsbury's winning plan was probably just as effective. White could probably also win here with 19. RxR and then 20. c4.

19... b5

Bardeleben apparently saw what was coming, and decided to take counter-measures with a Queen-side demonstration. This plan did not have a great chance of success, especially against Pillsbury, but some form of counter-attack was Bardeleben's only conceivable chance.

20. Qf5

20. RxR followed by 21. Nf5 and 20. Nf5 immediately both seem to win quickly. Objectively, these moves were better than 20. Qf5. But it is hard to fault Pillsbury for a plan that--though not the fastest route to victory--certainly seems sufficient.

20... QxQ

The ending held little hope for Bardeleben, but everything else was worse.

21. NxQ

This left the following endgame for the contestants to contemplate:

click for larger view

In a classic understatement, Sergeant-Watts stated that: "After the exchange of Queens the Black Bishop remains shut in." In fact, Bardeleben's Bishop was not just bad, it was utterly useless.

21... Reb8

21...Rec8 was perhaps slightly better, though still leaving Black hopelessly constricted.

22. h4

"!"--Tournament Book.

Pillsbury is playing to advance his h-pawn to h5 and, if allowed, to h6. 22. Rf3 was seemingly a faster way to win.

22... a5

Continuing with his Queen-side build-up. But 22...Rc8 or 22...c4 were more forceful and better ways to pursue this plan.

23. h5!

23. Rf3 was also good. But Pillsbury throughout this game was relentless in pursuing his plan.

23... h6

"!"--Tournament Book and Sergeant Watts.

"h6 must be stopped" (Sergeant-Watts)."

If Black is willing to endure a White pawn on h6, he might try 23...a4 or 23...Rc8 to continue his own counter-attack. An unpleasant choice of poisons for Bardeleben.

24. Rf3

Reading himself to pound on the g-file.

24... a4!

One must admire Bardeleben's spunk in his difficult position, which now was:

click for larger view

25. Kh2

He could just have played 25. Rg3+, but Pillsbury was in no hurry and decided to eliminate any risks from counter-play.

25... axb3

The plan to attack on the a-file was almost certainly futile, but passive defense with something like 25...Rc8 would not have been much better.

26. axb3 Ra2

This incursion with the Black Rook was less dangerous than it may appear on first glance.

27. Re2

Pillsbury could have equally well preceded this move with 27. Rg3+

The position was now:

click for larger view

The finish came quickly after 27. Re2

27... c4

Bardeleben had clearly lost patience. He might have tried to hang on with 27...Raa8 or 27...Rd8 or perhaps 27...b4.

Giving Pillsbury an opening was fatal, as might have been predicted.

28. bxc4

There were many ways for White to win here: e.g., 28. Rfe3, 24. Rg3+; 28. b4; etc. Pillsbury's method was simple.

28... bxc4
29. Rc3

This left Bardeleben in the following unpleasant situation:

click for larger view

29...Ra5 30. Rd2 Rb4 31. Rd4 Rc5 32. Ne3 loses the c-pawn and in the long run the game, but this was the "best" I can find for Bardeleben here.

But, perhaps tired from the pressure Pillsbury had been applying, Bardeleben blundered here:

29... Rb4?

"?"--(Tournament Book)(Sergeant-Watts).

"A blunder, but the game was already lost." (Sergeant-Watts).

30. Re8

30. Rg3+ also forces mate.

After the text (30. Re8), the sorry position for Bardeleben--with 31. Rg3+ coming-- was:

click for larger view

30... Rbb2

There are varying accounts of what Bardeleben actually played here. The Tournament Book give the move as 30...Rb7. Sergeant-Watts give the move as 30...Rbb2. The difference may derive from the change from descriptive to algebraic notation (i.e., "30...RKt7" being construed as 30...Rb7]. The discrepancy is not of much importance, since White mates in 3 on either move [Bardeleben could have dragged the game out for a few more moves with 30...Kh7 or 30...Rb8, both of which are nonsense moves].

31. Rg3+

1-0