Tchigorin's play against Pillsbury's Petroff Defense was not always the best (his 7. c4 was inferior to Tahl's 7. Nc3), but after Pillsbury's 7...c6 (7...BxN would have been much better) and 12. Be6 (12...b6 was best), Tchigorin emerged with much the better position.

Tchigorin lost most of his edge with 16. c4 (16. Be5 or 16. Rae1 were both much better).

Pillsbury's 16...Be5 involved a clever tactical trick (17. BxB would have lost a pawn to 17...Qxd4+). Needless to say, Tchigorin didn't fall for this.

I agree with Sergeant and Watts that Tchigorin should probably have kept his Bishop on the board (given his more broken pawn structure) with 19. Bf4, but Tchigorin had King's side threats after his 19. Qg3 and was at least equal.

Tchigorin's 21. Kh1 set a neat trap for Pillsbury. The position was as follows:

click for larger view

Had Pillsbury here tried to grab Tchigorin's c-pawn with 21...Qxc4, he would have gotten crushed after 22. f6 g6 23. Qg5.

Pillsbury didn't fall for that one, and played the careful 21...Kh8.

The game was now approximately even. Tchigorin's next move, however, has come in for much unfair criticism.

Tchigorin played 22. Rad1. Sergeant and Watts (in their collection of games by Pillsbury) give this move a ? and state that it loses a pawn and wastes a move. The Tournament Book calls the move "premature" was says that 22. Rac1 was best. Fritz prefers 22. f6.

In fact, all of the proposed moves are fine, as was Tchigorin's 22. Rad1. This move definitely had bite to it, and Tchigorin's problems came from his failure to follow up properly.

Pillsbury took the bait with 22...Qxc4. Now, either 23. Rfe1 or 23. f6 look good and seem to lead to equal--and fascinating--chances. Tchigorin's King-side threats seem to have been worth the pawn sacrificed.

But--if the score of the game is to be believed--Tchigorin here played 23. Rde1. I do not understand this move. It not only loses time (moving the Rook he just posted on d1), it allows Pillsbury to grab another pawn with 23...Qxa2.

However--and once again if the score is to be believed--Pillsbury did not take the pawn and instead played 23...Rad8. Now, Tchigorin could have been very much back in the game with 24. f6. Instead, he played 24. Rf4, once again allowing Pillsbury to snatch the a2 pawn with 24...Qxa2. And once again, Pillsbury left the pawn untouched and played 24...Rd4.

The position was now as follows:

click for larger view

As Sergeant and Watts note, Tchigorin could here have simply played 25. RxR QxR 26. e6 with about equal chances.

Tchigorin, however, chose to set a nasty trap with 25. f6. Had Pillsbury played 25...gxf6 he would have been mated after 26. exf6 Rg8 27. QxR+!!

Pillsbury saw the danger and played 25...Rg8 (25...g6 would also have been good).

click for larger view

Tchigorin would have had approximately even chances with 27. Qb3. But Tchigorin instead set a lovely trap with 27. e6?!?! Had Pillsbury played 27... gxf6 he would have been blown away with 28. QxR+!! KxQ 29. e7!

Indeed, after 27. e6, the only saving move for Pillsbury was 27...fxe6. Unfortunately for Tchigorin, Pillsbury found this resource, which happens to win the game.

The combination Tchigorin had cooked up with 27. e6 was so gorgeous, I regret having to label it the losing move!

Even after Pillsbury's theoretically "winning" 27... fxe6, the thrills were far from over.

Tchigorin could have gone quietly with the nominally "best" 28. fxg7+. This, though, was hardly Tchigorin's style. He played 28. f7?!. Pillsbury carefully played 28...Rf8. Here, a lesser player might have played 29. Rxe6 and lost to 29...g6. Fritz recommends 29. h3, a move that would have solved Tchigorin's back-rank mate problems but set no difficulties for Pillsbury.

Tchigorin, however, was neither a lesser player nor a computer. He was a fiend of a tactician and was still playing to win. He therefore responded 28. Rf1. Tne position was now as follows:

click for larger view

Tchigorin's 29. Rf1 may seem harmless, but in fact he now threatened 30. Qb8!! winning.

Wow!

Pillsbury was sufficiently disconcerted by this threat that he played the second-best 29...Qd8. In fact, Pillsbury had an amazing retort of his own that all the commentators have missed: 29...Qc4!

After Pillsbury's 29...Qd8, Tchigorin remained on the attack with 30. Qe3. Pillsbury here should have played 30...Qe7. Instead, he tried to ratchet up some threats of his own with 30...e5.

Pillsbury's move contained yet another tactical trick. Had Tchigorin played 31. Qxe5 Pillsbury would have had a simple win with 31...Rxf7 (as Sergeant and Watts pointed out more than a century ago).

Once again, Tchigorin saw the threat, and played the accurate 31. Qxa7.

After accurate and tactically astute play by both sides through Pillsbury's 36...Qd6, the position was as follows:

click for larger view

This position is a theoretical win for Pillsbury. But the win is hardly easy, and Pillsbury would have had his hands full in the resulting Rook ending had Tchigorin played 37. Qc5.

But here Tchigorin fell from grace with the careless 37. g3 and after Pillsbury's 37...Rf2+ the game was suddenly over (37...Qd4 would have been even more brutal).