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

The Berlin Defense, often played in that era.

4. 0-0 Nxe4
5. d4 Nd6

Kramnik has had success with this ugly-looking move, I prefer 5...Be7

6. dxe5

Inferior to 6. BxN, but part of Didier's plan as we shall see.

6. ...NxB
7. a4

Winning back the piece.

7. ...d6
8. e6

A wild attacking idea from the fertile mind of David Janowski. Didier decides to give it a go here against Pillsbury. 8. axN is more usual and theoretically best, but the text is definitely playable.

8....fxe6

As GrahamClayton has helpfully pointed out, Pillsbury played the stronger 8...Bxe6 at Munich later that same year against Wolf. Pillsbury here repeats the line Showalter played against Janowski in their match. Though probably not as good as 8...Bxe6, the text is entirely playable and the suggestion that Pillsbury got a lost position by playing this variation is mistaken, as I will show.

9. axN Ne5
10 Nd4

An improvement by Didier on the previously played 10. Ng5.

10. ...Bd7

10...Be7 was much better.

11. Nc3

An improvement on the older 11. Re1.

11. ...Be7
12. f4 Ng4
13. Nd5 0-0
14. f5

A doubtful idea by Didier. Safest and best was 14. Qf3

14. ... Nf6

Marco notwithstanding, 14...Ne5 would not have been an improvement. With the text, Pillsbury is up a pawn with a solid albeit cramped game. By contrast, 14...Ne5 would have led to trouble after 15. Ra3 getting the Queen's Rook into the game. Marco's proposed 15. f6 would have been unsound here: 15. f6 Bxf6 16. NxB+ gxf6 17. Ra3 leads to nothing after 17...Kh8. 17. Nf5 would have been better in this line, but White would still have insufficient compensation for the two pawns sacrificed.

15. NxB+ QxN
16. Re1

Contrary to the claim that Didier achieved a won game through the opening, here was the position after Didier's 16. Re1

click for larger view

Didier has some pressure and come compensation for the sacrificed pawn, but Pillsbury clearly has the better chances and--in the long run--arguably a won game. But now Pillsbury loses his head.

16. ...Qd8

For reasons I cannot guess, none of the commentators mentions this poor and (for especially for Pillsbury) surprisingly passive move. 16...Ne4 was better and may well have led to victory for Pillsbury. After the text, Didier's attack is back in business.

17. Bg5 h6

Marco says that 17...c5 was better, but after 18. bxc6 e.p. Black gets nothing. Both the text and 17...a6 seem better than Marco's move.

18. Bh4 c5

According to Sergeant-Watts in their book on Pillsbury, "It is indeed difficult to suggest a move for Black here...Black seems at the end of his resources and must weaken his position by moving." Sergeant-Watts suggest 18...d5 as a possibility, and White at most achieves a slight edge after this move. Best for Black here seems to be 18...a6. But even after the text, Black has a fully playable game. Didier has compensation for his sacrificed pawn, but nothing approaching a winning or even significantly better game.

19. bxc6 e.p.

19. Ne6 appears good at first sight, but Rosenthal is correct that after 19...fxN 20. fxe6 Qc8 (or 20...Qc7 or 20...Qe8) Black is better. Didier's move is thus best.

19. ...bxc6
20. Ra6

"Very well played. The move should win the game..." (Rosenthal). Excellently played." (Marco).

While 20. Ra6 is OK for White, it does not lead to anything approaching a "win" for Didier and arguably 20. Qd3 was better.

20. ...Qc8

Pillsbury avoids the disaster that would have ensued after 20...c5: 21 Rxd6! Qc7 (21...cxN may be slightly better, but insufficient to hold the game) 22. Nb3 wins for White. Marco's proposed 22. RxN here would blow the win after 22...cxN. Marco's line only works because he has Black playing the foolhardy 22...gxR which leads to a forced mate for White after 23. Qg4+ Kh7 24. Re7 (even here Marco's analysis is sloppy, since his 24. Bxf6 lets Black escape after 24...Rg8 25. Qh5 Bc6 and only works because he has Black playing the awful 25...Be8 and thus losing to 26. Re4!).

The best analysis here seems to be by Sergeant-Watts, who proposed 20...Re8 as best. Pillsbury's move (20...Qc8), however, seems almost as good. He certainly was not lost or even in any real trouble---yet.

21. BxN

"!!" (Rosenthal). The text seems fine (and very pretty, since the exchange sacrifice is sufficient for equality). But 21. Ra3 seems more logical.

21...gxB??

This move is terrible and leaves Pillsbury open to a winning King-side attack. Contrary to all the commentators, best and fine was 21...QxR. After 22. Qg4 (22. Re7 was equally good) g6, Pillsbury would have been fine after either: (i) 23. Qh4 Rae8 (and not Marco's 23...Rfe8 which loses to 24. Ne6 or Tinsley's 23...Kh7 which loses to 24. fxg6+) or (ii) 23. Qh3 Rae8 or 23...h5 (and not Rosenthal's 23...Rfe8 which loses to 24. Rf1! h5 25. Qg3 [and not Rosenthal's dreadful 25. g4 Re4 26. gxh5 gxf5 27. Nxf5 Kh7 which not only is not a win for White as Rosenthal erroneously claims but actually loses]).

22. Ra3

This is sufficient to win but (though not mentioned by any of the commentators) drastically inferior to the crushing 23. Qg4+.

22. Re8
23. Rf1

Tinsley states that it was "necessary and best to play 23. Qg4+" here. But this is one move too late and would blow Didier's win. Best was 23. Ne6 (if 23...fxN 24. Qh5; and if 23...BxN 24. Rg3+), though the text is also probably sufficient to win as well.

23. Re4

Fine defense by Pillsbury in a tough position. It is essential to prevent White's Queen from getting to h4 (a point Pillsbury seems to forget three moves later).

After Pillsbury's 23...Re4, the position was as follows:

click for larger view

24. Qh5

Rosenthal in the Tournament Book calls this "a very well played move," but in fact it probably throws away Didier's winning chances. Correct was 24. Ne2 getting the Knight into the fray, a move Didier plays on his next turn.

Didier's move is pretty since, as Rosenthal shows in the Tournament Book, the Knight sacrifice cannot be accepted. If 24...RxN? 25. Rg3+ Kf8 26. Qxh6+ (26. Re1 is even more devastating, though Rosenthal's move is certainly decisive) Ke7 27. Re1+ (or--even better--27. Qe3+).

24. ...Qf8

Pillsbury carefully avoids the fatal 24...RxN trap Didier had set for him.

25. Ne2 d5
26. Ng3 Re5???

How could Pillsbury have played this lemon? And how could all the commentators miss the fact that this was a blunder? Pillsbury (thanks to Didier's celebrated but erroneous 24th move) could hold the game here with 26...Rc4. It was essential to prevent White from playing 27. Qh4. But now---

27. Qh4

And now, Pillsbury is busted.

27. ...Qc5+. This was indeed desperation, as Sergeant-Watts noted. While Pillsbury is lost here, his best chance (objectively) was 27...Qae8. In reality, however, this theoretically poor move leads to victory for Pillsbury.

28. Kh1 Qc4

Rosenthal gives this move "!!" since it allows the following atrocious blunder by Didier.

29. Raf5 ????

How many question marks does this blunder deserve? Didier deserved a better fate, and--as Sergeant/Watts note--Pillsbury got a "lucky point" here. As is obvious, both 29. Qxf6 and 29. Qxh6 would have won the game for Didier. Instead, he hangs his Queen

29. ...QxQ

Ouch!