But below the leading trio was a group of five vying for 4th place:

Marco (8-3)
Burn (7-4)
Berger (6.5-4.5)
Janowski (6-5)
Wolf (6-5)

With Marco and Berger drawing their games in this round and Wolf losing to Pillsbury, the Janowski-Burn game provided an opportunity for a pick-up for both of these contestants.

Although the game ultimately ended in a draw, it could not have disappointed any of the onlookers. While not a flawless gem, it featured slashing gambits and attacks from both sides.

As <zaman216> and <Straclonoor> had correctly noted on this site, Janowski missed a win on move 39. As I will seek to show, there were other opportunities missed, but lots of exciting play. I greatly enjoyed playing over this game and exploring the tactical opportunities it provided to both players.

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

The Berlin Defense which was very much in fashion at Munich 1900. As Vladimir Kramnik demonstrated in his match against Gary Kasparov, the Berlin can present real headaches for White in trying to obtain an opening advantage.

4. 0-0 Nxe4
5. d4 Nd6

a fairly normal position in the Berlin was thus reached:

click for larger view

The "normal" and seemingly best move here is the one played by Kasparov against Kramnik in all the Berlin games in their match: 6. BxN (followed by 7. dxe5).

But Janowski had other ideas:

6. Ba4?!

Not a move generally to be recommended; among other things it seems to lose a pawn. The move was, however, played four times (!) by Showalter in his 1897 match against Pillsbury (two wins apiece).

6. dxe5 NxB 7. a4 winning back the Knight is what MCO-13 calls an "amusing" alternative.

6... exd4

This is best, and picks up a pawn and gives Black the better chances. 6...e4 is the main alternative.

7. c3?!

"!"--(Tournament Book).

Janowski makes this a true ganbit opening. He was clearly playing for a win. 7. Nxd4, the more solid choice here, still leaves White a pawn down. All in all, Janowski's 7. c3 seems most in the spirit of the opening.

The position was now:

click for larger view

7... Be7!

The text of 7...b5 seem best. 7...dxc3 is risky, as Pillsbury learned in a game he lost to Showalter in his 1897 match: 7...dxc3 8. Nxc3 (or 8. Re1+) Be7 9. Nd5 (9. Re1 is equally good) 0-0 (Pillsbury could equally well have played 9...b5) 10. Re1 Bf6? (10...b5 was much better) 11. Bf4! with a big edge for White (e.g., the following wild line: 11...Bxb2 12. Nxc7! Qxc7 13. BxN Qd8 14. Rb1 and White--though two pawns down--has excellent chances.

Burn's choice was far better, especially in over-the-board play.

8. cxd4 b5

"!"--(Tournament Book).

8...0-0 was perhaps safer, though the text seems to give Black an active defense to White's gambit.

9. Bb3

9. Bc2 was more accurate.

9... Na5

9...0-0 or 9...Bb7 were less committal and probably better choices.

10. Bc2 Bb7

The position was now:

click for larger view

11. Ne5

The developing move 11. Nc3 might be theoretically better, but Jaowski preferred this Pillsbury-like attacking motif.

11... 0-0
12. Nc3 Nac4

With his hanging b-pawn, Burn might have considered 12...a6. To forestall the coming double attack on his h7, Burn might have played 12...g6. Instead, Burn--normally a cautious positional player--went for an uppercut. Burn's aggressive defense, even if not always sound, made this game an exciting encounter.

13. b3

It is hard to understand why Janowski did not just grab the pawn with 13. Nxb5 (with Qd3 to come, e.g., 13. Nxb5 NxN(e5) 14. dxN NxN 15. Qd3 g6 16. QxN and White has won back his pawn with at least even chances).

13. a4 was another way for White to exploit the weak Black b5 pawn.

13... Nb6

Best.

The position was now:

click for larger view

14. Nxb5

14. Re1 was a possible alternative.

14... NxN
15. Qd3

The mating threat on h7 regains the piece.

15... f5?!

Continuing his ultra-aggressive defense, but creating needless weaknesses on his King-side. 15...g6 was best and would leave Burn well-situated.

16. 16. Qxb5

This left:

click for larger view

16... a5

"!"--(Tournament Book)

16...Qe8 was another good way to play for a counter-attack.

17. Qe2

17. Rd1 was perhaps more accurate, but the text was also fine.

17... Bd5

17...g6 and 17...Bd6 were also good choices here for Burn.

18. Re1 Bb4
19. Bd2 Qh4

Burn could also have played 19...Qf6 or 19...BxB, but he had apparently decided to play for a King-side attack. Not a bad idea again an attacking player such as Janowski.

20. BxB axB

The position was now:

click for larger view

21. Qd2 d6

21...Rae8 was perhaps better, since if then 22. Qxb4 d6 23. Nf3 (Not 23. Nd3? Qg4!) BxN 24. gxB and White's busted pawn structure renders his extra pawn meaningless).

22. Nd3

The position was now:

click for larger view

In the diagrammed position with which I ended my last post, Burn had four plausible options, at least three of which were reasonable:

A) 22...Qxd4, which lessons the tension and yields at most a small edge to White after 23. Qxb4 Qf6 (better than trading Queens) 19. Nf4.

B) 22...Rf6, a solid way to build up toward a King-side attack but that would likely lead to equalizing exchanges; e.g., 22...Rf6 23. g6 Qxd4 24. Qxb4 QxQ 25. NxQ.

C) the wild sacrifice 22...Bxg2?! 23. KxB Qg4+ 24. Kf1 Qh3+ 25. Ke2 Qg4+ 26. f3 Rae8+ 27. Kf2 Qxd4+ 28. Kg2 Nd5 with compensation for the sacrificed piece.

Burn, however, chose a fourth and manifestly inferior (though, like C, a hyper-aggressive) line:

22... f4?!
23. f3!

The refutation of Burn;s 22nd move, the position now being:

click for larger view

23... Rf5!

The only way to proceed after his doubtful 22...f4?! It does, however, result in loss of a pawn.

24. Qf2

Another excellent move by Janowski that beats back Burn's clumsy attempt at a King-side attack.

24... Qh6

24...QxQ+ 25. KxQ Rh5 26. Nxb4 would not be pleasant for Black.

25. Nxb4

Picking up a pawn through the discovered attack on the Black Rook on f5.

This left:

click for larger view

25... Rh5!

Throwing caution to the wind and going all out in his attack on Janowski's King. The whole idea was probably not sound, but Burn was too far down this rabbit-hole to back off at this point.

26. h3 Rf8?

An inaccuracy that should have cost Burn the game. He had to follow up with 26...Be6 here.

27. a4!

Black is busted. White threatens 28. a5, winning a piece and advancing his monster passed pawn down the board.

27... Be6

Either this or the unattractive 27...Qg5 was forced.

28. a5?

As is so often the case, a threat is stronger than its execution. Janowski should have played 28. Qe2!, threatening the b6 Black Bishop, creating an escape square for the King, and keeping a5 in reserve. A likely continuation was 28. Qe2 Nd5 29. QxB+ QxQ 30. RxQ NxN 31. Rc1 leaving White in an ending a pawn up he almost certainly should win.

The text (28. a5) left the position as follows:

click for larger view

Burn was still alive, and with 28...Nd7 he would retain counter-chances. Instead, he went for a knock-out blow:

28... Bxh3?!

Unsound theoretically, but perhaps motivated by the fact that Janowski was far better at attack than defense. The text would have been fatal against a strong defender like Lasker.

29. axN

29. gxB? Rxh3 would give Burn a winning attack. The text is best and was the refutation of Burn's wild attack.

29... cxb6

This leaves Black a piece down and gives White a chance to organize a defense, but 29...Bxg2?! 30. QxB Rg5 31. Be4 RxQ+ 32. KxQ would have left Janowski with a Rook and two minor pieces for a Queen and a winning position.

30. Bd3?

This let Burn off the hook. Janowski should have played 30. Be4 or 30. Qe2. With a piece to the good, he should almost certainly have won. Now, however, things got crazy, the position being:

click for larger view

Down a pawn and with Janowski's Rooks each controlling an important file, Burn decided to attack rather than try to defend with 30...Kh8. With Janowski's Queen sitting on f2, Burn was able to pick up Janowski's Queen, but at a significant cost:

30... Bxg2!
31. QxB

Not 31. KxB? Rh2+ 32. Kf1 RxQ+ 33. KxR Qh2+ 34. Kf1 Qh3+ 35. Kf2 (or 35. Ke2 Qg2+ immediately) Qg3+ 36. Ke2 Qg2+ 37. Kd1 Qxf3+ and Black wins.

31... Rg5

Picking up Janowski's Queen with a skewer.

32. Ra2

32. Be4 would have given Janowski better winning chances, though the threatening position of Burn's Queen would give him significant counterplay and reasonable chances of holding the game.

32... RxQ+
33. RxR

This left:

click for larger view

Janowski has emerged with a Rook and two minor pieces for Burn's Queen and two pawns. Thus material was nominally equal. But Janowski's Rooks control the e and g files, and his chances must surely be reckoned as better here.

33... Qf6?

This attack on the d-pawn was easily countered, and the game was now probably lost for Burn. He should have played 33...Qh4.

34. Nc6?

Missing his chance. With 34. Re4 (or 34. Bc4+ and then 35. Re4), Janowski would threaten Nd5 and Nxf4. I see no good way for Burn to have defended the Black position after that. The threat of Ne7+ entailed by 34. Nc6 should have been easily parried by Burn.

34... d5?

Suicide. He had to play 34...Kh8 (or perhaps counterattack with 34...Ra8). After the text, Black is helpless--or should have been.

35. Ne7+! Kf7

35...Kh8 36. Bxh7 is also hopeless for Black.

After 35...Kf7, the position was:

click for larger view

White to move and win.

36. Re5?

Janowski was still winning after this second-best move, but 36. Bxh7! (threatening 37. Bg6 checkmate) would have left Burn without resource, e.g., 36...Qxd4+ (taking the pawn Janowski's 36th move protected--for no good reason as will be seen in this variation) 37. Kf1 Rh8 38. Nc6 Qc3 39. Ne5+ winning Burn's Rook and the game.

36... g6

Burn has life again. He could also have played 36...Ra8 with at least some counter-attack. Burn would still be lost in either case, but unlike the situation after 36. Bxh7!, he can play on for a while (and in practice, he managed to draw).

37. Nxd5

This left:

click for larger view

37... Qd8?

Very weak. With 37...Qh4, Burn could have offered resistance for at least a while.

38. Re7+

38. Rh2 was even stronger, but the text should have been more than sufficient to win.

38... Kg8

Burn is busted. He might have played 38...QxR 39. NxQ KxN leaving him with Rook and four pawns against Rook Bishop and three pawns. But his chances in such an endgame were probably slim to none. Thus, the text was probably his best practical chance. Indeed, it enabled him to draw the game, but only because of missteps by Janowski.

After 38...Kg8, the position was:

click for larger view

There are at least three crushing moves for White here that should each wrap up the game quickly:

A) 39. Bxg6! This suggestion by zaman216 is my favorite: 39. Bxg6! hxg6 (hopeless, but Black has nothing better) 40. Rg6+ Kh8 41. Rh6+ Kg8 42. Rhh7 Rf7 43. RexR QxN 44. Rfg7+ Kf8 45. Ra7 and wins.

B) 39. Rge2 (Straclonoor's move) which also wins easily, e.g., b5 (39...Kh8 40. Rh2) 40. Nf6+ RxN (forced to stop mate in one) 41. Re8+ leading to an ending in which White has Rook, Bishop and three pawns against Rook and four pawns.

C) 39. Rh2 Rf7 (to stop Rhxh7) 40. RxR KxR 41. Rh7+ Ke6 42. Be4 leaving White with Rook, Bishop, and Knight against Black's Queen.

Janowski, however, did not play any of the foregoing crushing moves, but instead played:

39. Bc4

This did not necessarily blow the win, the position now being:

click for larger view

Janowski probably still had a theoretically won game, but his task was now harder.

39... Kh8
40. Rh2! h5

Forced.

41. Re6

"?"--(Tournament Book)

41. Rhe2 was best. (Tournament Book).

While the Tournament Book's suggested move was clearly better than Janowski's effort (as was 41. Re5), the game was probably still a win for White.

41... Rf5

"!"--(Tournament Book).

42. Rhe2

This left:

click for larger view

Burn could have made Janowski's task very difficult with 42...Qh4. Instead, he erred badly with:

42... Kh7?

43. Rg2! was now overwhelming. 43. Re8 was also sufficient to win. But here Janowski blundered--and threw away his win with:

43. Re7+?

Burn now had an easy escape:

43... Kh6!
44. Re8 Qh4

Forced, but sufficient to save the game, the position now being:

click for larger view

Janowski'a win was now gone:

45. Rg2 b5
46. Rh8+ Kg7

The position was now:

click for larger view

After 46...Kg7, Janowski (White) had two pieces hanging and Burn's Queen had become a major threat to Janowski's King. Janowski had only one way to save himself from immediate defeat. It was a brilliant resource, and Janowski found it:

47. Rg8+!

The position was now:

click for larger view

Had Burn taken Janowski's Rook, he would have been lost: 47...KxR 48. Ne7+ Kf8 (48...Kg7 and 48...Kh8 would each run into a devastating Knight fork; and 48...Kh7 would lose to 49. Bg8+) 49. Nxg6+.

But Burn would have been fine after 47...Kh7! But instead Burn blundered with:

47... Kf7?

"!"--(Tournament Book).

The Tournament Book notwithstanding, 47...Kf7 was a blunder, the position now being:

click for larger view

Janowski now once again had a win, this time with 48. Nxf4+ (e.g., 48...bxB 49. Rg8+! (49...KxR 50. Nxg6+; or 49...Kg7 50. RxR Qf1+ 51. Kh2 Qh4+ 52. Nh3).

But--yet again--Janowski erred with:

48. R2xg6? bxB

This left:

click for larger view

Now Janowski had to settle for a draw by perpetual check:

49. R8g7+ Kf8
50. Rg8+ Kf7
51. R8g7+ Kf8
52. Rg8+