American Postal Chess Tournaments
Click to download all 1999 Games
Jonathan Voth, Games Editor
507 N. Arthur St., Apt I-106
Kennewick, WA 99336
Grandmaster Jack Peters judged Burke-Johnson in 92RF-3 as the 1998 game of the year.
This is a repeat for John Burke, bringing him into the small group of repeat winners:
George Thompson, Rudy Vance, Ed Frumkin, and Jon Edwards. Kinda makes me wonder
who the first three-pete will be!
[Peters: One of the less-investigated answers to the Benko Gambit.
4 ... Na6
[Peters: Inviting White to grab pawns by 5 cxb5 Nb4 6 Qxc5 as 6 ... Nfxd5 7 a3?! Bb7! prepares the rejoinder 8 ... Rc8. Another reasonable method for Black is 4 ... bxc4 5 e4 e6.]
5 a3 Qa5+!?
[Burke: Black is looking for tactics. However, if this maneuver is unsuccessful it could be a costly waste of time.]
6 Nd2 bxc4 7 e4 Nb4?!
[Peters: Black cannot win the d-pawn by 7 ... e6 8 Bxc4 exd5 9 exd5 Nc7 because 10 Ne2! anticipates 10 ... Ncxd5?? 11 Bxd5 Nxd5 12 Qe4+ or 10 ... Bb7 11 Nc3 Ncxd5 12 Nxd5 Bxd5 13 Bxd5 Nxd5 14 Qe4+. But 10 ... Bd6!? 11 Nc3 Be5 12 0-0 Bxc3 13 bxc3 0-0 creates a genuine threat to White's pawn.]
[Peters: White must speculate, as 8 Qxc4?! Ba6 9 Qc3 fares poorly. After 9 ... Nd3+ 10 Bxd3 Qxc3 11 bxc3 Bxd3 12 f3 e6 13 c4 exd5 14 cxd5 Rb8! (preventing Bb2) 15 Nh3 g6 16 Nf4 Ba6 Black's pair of bishops and passed c-pawn assure him of an edge.
8 ... Qxa1 9 Bxc4
Position after 9. Bxc4
9 ... cxb4?!
[Peters: Black must find an improvement here to justify his seventh move. Unfortunately,
9 ... Ba6 10 bxc5 gives White a mighty center for a tiny material investment, while 9 ... e5 10 Ngf3 d6 11 bxc5 dxc5 12 Nxe5 does not help Black's development. Maybe 9 ... e6 10 Ngf3 exd5 11 exd5 Bb7 is the best solution.
10 Nb3 Qa4 11 e5 Ng8
[Peters: Despite his great advantage in material, Black is in serious trouble.
After this humble retreat, he will never manage to untangle his kingside pieces.
However, 11 ... Ng4 gets driven back by 12 Qe2! h5 (not 12 ... d6?? 13 Bb5+) 13 h3. Perhaps he should return material by 11 ... Ba6 12 Bxa6 Qxa6 13 exf6 gxf6 so that his rooks can operate on the c- and g-files.
12 Qd3 a6 13 Nf3 e6?!
[Peters: Black's last chance for normal development is 13 ... d6 14 e6 fxe6 15 dxe6 Nf6 16 0-0 Bb7 with the threat of 17 ... d5.]
14 0-0 Bb7 15 d6
[Peters: Although White has no immediate threats, he has more than enough compensation for the exchange and pawn.]
15 ... Qc6
[Burke: The queen is reunited with her troops at last but at the cost of a bottled-up kingside.]
16 Nbd4 Qe4
[Peters: Or 16 ... Qc8 17 Bd2, followed by Rc1.]
17 Qd2 a5
[Burke: The white pieces must be stopped from infiltrating.]
18 Re1 Qg4 19 h3
[Burke: The queen is pushed far from the weakened queenside.]
19 ... Qg6
[Peters: Another Black piece huddles pitifully on the kingside. Naturally, White searches for an entry on the other wing.]
20 Qe3! Rc8
Position after 20 ... Rc8
[Peters: When it's time to attack, White does not shrink from offering sacrifices!
He plans 21 ... Rxc4 22 Qb6 Bc8 23 Nc7+ Rxc7 24 Qxc7 Be7 (what else?) 25 Qxc8+ Bd8 26 Be3 Nh6 27 Bb6 0-0 28 Bxd8 recouping his material with interest. Not much
better is 22 ... f6 23 Qxb7 Nh6 as White breaks through with 24 Nc7+ Kf7 25 Nxe6! Kg8 (or 25 ... Kxe6 26 exf6+ Kxf6 27 Bg5+ Kf7 28 Qd5+) 26 Nxf8 Kxf8 27 e6 dxe6 28 Qe7+ Kg8
21 ... Qc2 22 b3 Ne7!
[Peters: It's too late to buy White off with a mere piece.]
23 Qb6! Bxf3 24 Nc7+!
[Peters: Imagining the smothered mate after 24 ... Kd8? 25 Nxe6+ Ke8 26 Qd8+ Rxd8 27 Nc7#]
24 ... Rxc7
[Burke: The only move 24 ... Kd8? 25 Nxe6+ Ke8 26 Qd8+ Rxd8 27 Nc7 is the always pleasing smothered mate.]
25 dxc7 f6 26 gxf3
[Peters: White trails by only a pawn, and his passer will claim at least a piece. A few more accurate moves clinch victory.]
26 ... Qc3 27 Rd1! Kf7
[Peters: Black rejects the pretty finishes 27 ... Qxf3 28 Rxd7! Kxd7 29 Bxe6+ Ke8 30 c8Q+ Nxc8 31 Qb5+ Ke7 32 Qd7# and 27 ... Qxe5 28 Rxd7! Kxd7 29 Bb5+ Nc6?! 30 Qxc6+ Kc8 31 Ba6#]
[Peters: Not falling for 28 Rxd7? Qxc1+ 29 Kh2 when 29 ... Qf4+ 30 Kg2 Qg5+ 31 Kf1 Qc1+ 32 Ke2 Qc2+ bothers White with checks.
28 ... g5
[Peters: White dispatches 28 ... Qxe5 simply by 29 Qxe5 fxe5 30 Rxd7 Kf6 31 Rd8. Cover your eyes for the rest of the massacre.]
29 Qxd7 Kg6 30 Qxe6 Qxe5 31 Qxe5 fxe5 32 Rd6+ Kh5
[Burke: Ouch! But if the king goes to g7 then Bb2 wins the house.]
33 Bf7+ Kh4 34 Rc6!
[Peters: Threatening mainly 35 Rc4+ Kxh3 36 Bxg5 and 37 Rh4#.
34 ... g4 35 Rc4 Rg8
[Peters: Black has been soundly thumped, and it's past time to resign.]
36 Bxg8 e4 37 Rxe4 h5 38 hxg4 Nd5 39 g5+ Nf4 40 Bxf4 Bd6 41 Be3+ 1-0
1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Be2
[Hitselberger: Fred and I had played six or seven times prior to this game going back to the mid-80's. I chose 6 Be2 because I wanted to play a safe Karpovian move against a strong opponent.]
6 ... e5 7 Nb3 Be7 8 0-0 0-0 9 Kh1
[Peters: This popular treatment of the Najdorf Sicilian tries to keep open the options of f2-f3 and f2-f4. White has also tested 9 a4, 9 f4, 9 Be3, and 9 Bg5.]
9 ... Qc7 10 g4!?
[Peters: As played twice by British GM Michael Adams. Both of his opponents replied
10 Be6 11 g5 Nfd7, which seems best.
10 ... h6?!
[Peters: Helping White open a file leading to Black's king.]
11 g5 hxg5 12 Bxg5 Be6 13 Rg1 Rd8
[Hitselberger: My model game here was Lau-Roeder, Muenster 1997. Roeder continued here 13 ... Nc6 14 Bh6 Ne8 15 Nd5 Bxd5 16 exd5 Nd8 17 Bd3 f5 18 Qh5 e4 19 Bxg7! Nxg7 20 Rxg7+! Kxg7 21 Rg1+ Kf6 22 f4 +-]
14 Bh6 g6
[Peters: Perhaps 14 ... Bf8 is more solid, but White already has a promising attack.
[Peters: The simplest way to prevent Black's freeing thrust, ... d6-d5.]
15 ... Nxd5
[Hitselberger: 15 ... Bxd5 16 exd5 Nbd7 may be a better try here because it preserves the knight defender on f6.]
16 exd5 Bf5
Position after 16. ... Bf5
[Hitselberger: The text is possible because 17 ... e4 is met strongly by 18 Nd4! For example, 17 ... e4 18 Nd4! exd3 19 Nxf5 dxc2 20 Qd4 f6 21 Rxg6+ -+]
17 ... Bxd3?
[Hitselberger: In retrospect, this may be the losing move because it brings the
white queen into the fray. The try 17 ... Qd7! preserves a kingside defender. However, it is not "natural" because the black knight
wants to develop to d7.
18 Qxd3 Bf6?
[Peters: Black must have underestimated the following sacrifice. Like it or not,
he had to try 18 ... Kh7. Then 19 Qh3?! Qc8 may survive, but 19 Bg5 continues the attack.
Position after 18. ... Bf6?
19 Rxg6+! fxg6 20 Qxg6+ Bg7 21 Rg1 Rd7
[Peters: Black lasts longer with 21 ... Qf7!? 22 Qxg7+ Qxg7 23 Rxg7+ Kh8 but 24 Rxb7 costs him more material. For example, 24 ... a5 25 Nd2 Na6 26 Ne4 Nc5 loses to 27 Bg7+ Kh7 (or 27 ... Kg8 28 Nf6#) 28 Nxc5 Rdb8 29 Bxe5+ Rxb7 30 Nxb7 dxe5 31 d6.
[Peters: The last White piece joins the battle! The knight will land on g5 or f6,
with decisive effect.
22 ... Nc6 23 Ne4 Rf7 24 Ng5
[Peters: White threatens both 25 Nxf7 and 25 Qh7+ Kf8 26 Ne6+.
1 e4 c6 2 d4 d5 3 f3 e6 4 Nc3 Nf6 5 Bg5 h6 6 Bh4 Qb6 7 Rb1?
[Better is 7 a3, protecting the pawn on b2 as now 7 ... Qxb2? 8 Na4 picks off the Black queen.]
7 ... c5?
[Too aggressive. I think maybe 7 ... Be7 is better. Now I have to fight very hard to hold my game together.]
8 Bxf6 gxf6 9 Bb5+ Nc6 10 exd5 cxd4 11 dxc6 bxc6 12 Bxc6+ Qxc6 13 Qxd4 Bd7
[Forced. Otherwise, White can exchange queens with Qh4.]
14 Nge2 Be7 15 Qg4?
[White needs to protect his king. I suspect the queen move is an attempt to go pawn hunting.]
15 ... 0-0-0 16 Qg7 Rdf8 17 0-0
[Now Black has a target.]
17 ... Qc5+ 18 Kh1 Rhg8 19 Qxh6 Rh8 20 Qd2 Rxh2+ 21 Kxh2 Rh8+ 22 Kg3 f5
[22 ... Rg8+ draws, but I play chess for fun. If I lose the game and learn in the process, then I don't see it as losing.]
23 Qd4 Qc7+ 24 f4
24 ... Rg8+ 25 Kh2 Bc5
[This bishop on c5 is a monster!]
26 Qf6 Bc6 27 g3 Be7 28 Qd4 f6 29 Ng1
[This knight proves to be a poor "fig leaf" for the naked White king.]
29 ... Rh8+ 30 Nh3 Bd8!
[The bishop is defending and attacking on both sides of the board. It defends f6, allows the Black queen to enter the attack, it can go to the strong post b6 after the queen moves, and the rook can remove the "fig leaf" on h3.]
31 Ne2 Rxh3+ 32 Kg1
[32 Kxh3 Qh7#]
32 ... Qh7 33 c4 Rh2
[Threatens mate on g2.]
34 Rfd1 Bb6
[Now the bishop will defend d8 and pin the white queen.]
35 c5 Qh3 0-1
1 d4 e6 2 Nf3 c5
[Not being too loyal to any particular opening, I try to get into something other than what I think my opponent wants. It appears Tom does the same thing. We considered or avoided the French, Dutch, Sicilian, and Franco-Sicilian. We finally get into a hybrid Colle.]
3 e3 cxd4 4 exd4 Nf6 5 Bd3 d5 6 Nbd2
[Played to prevent a ... Ne4.]
6 ... Bd6 7 0-0 Nc6 8 c3
[Played to prevent ... Nb4.]
8 ... h6
[This feels antipositional but appears to be okay.]
9 Re1 0-0 10 Ne5
[This was the only plan I could come up with.]
10 ... Bxe5 11 dxe5 Nd7 12 f4
[I wanted to avoid exchanges with the natural move 12 Nf3 Qc7 13 Bf4 f6 14 Qe2 fxe5 15 Bxe5 Ndxe5 16 Nxe5 Nxe5 17 Qxe5 Qxe5 18 Rxe5.]
12 ... Nc5 13 Bc2
[This move keeps my bishop, but 13 Nf3 looks like a good alternative.]
13 ... b6
[This is a good plan to activate the bishop.]
14 Nf3 Ba6 15 Nd4 Rc8 16 a4 Nxd4 17 cxd4 Qc7!
[A nice move catching me by surprise. Perhaps 17 Qxd4 should have been played.]
18 Bd2 Nd3
[My 18th was played to prevent this.]
19 Bxd3 Bxd3 20 Bb4 Bc2 21 Qd2 Rfd8 22 Rec1 Qc4 23 Ra3 1-0
1 e4 e5 2 Nf3 Nc6 3 d4 exd4 4 Nxd4 Nf6 5 Nxc6 bxc6 6 e5 Qe7 7 Qe2 Nd5 8 c4 Nb6 9 Nd2 Bb7 10 b3 0-0-0 11 Bb2 Qe6 12 0-0-0
[I felt like I had a slight edge here, but had no solid plan. Basically I was going to try and keep his knight hemmed in as long as possible while I tried for some kingside play.]
12 ... Be7 13 f4 f6 14 Nf3 Qf5
[I thought this move wasted time, maybe 14 ... fxe5 was better.]
15 g3 c5 16 Bg2 Rhe8 17 Rhe1 fxe5 18 Nxe5 g5
[I thought for sure he would play 18 ... Bxg2 here. I had already started eyeing the knight sack and was glad he didn't.]
19 Bxb7+ Kxb7 20 Nc6
[I spent hours going over this and still wasn't sure if it was sound, but I was quite sure I got a lot of play, so I said "what the heck!"]
20 ... Kxc6
[He was forced to take or lose at least the exchange.]
21 Qf3+ d5 22 Re5 Qf7 23 cxd5+ Nxd5 24 Qe4
[Here I started thinking that I wasn't looking so good, but Black's next move allowed me to exchange some pieces and actually trap his king later on.]
24 ... Bf6 25 Rdxd5 Rxd5 26 Rxe8 Bxb2+ 27 Kxb2
[Here I knew I had the game won, thanks to a nice little waiting move later in 29 Ka3.]
27 ... Kb6 28 Rb8+ Kc6 29 Ka3 gxf4 30 Rd8 a6 31 Qc4 Rd7 32 Qxa6+ Kd5 33 gxf4 Ke4
34 Re8+ Kd4 35 Qe2 Kd5 36 Re5+ 1-0
1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 f3 0-0 6 Be3 Nc6 7 Nge2 a6 8 Qd2 Rb8 9 g4 b5
[In this variation of the Saemisch, white mobilizes for a quick kingside attack.]
10 h4 h5 11 Ng3 Bxg4!?
[11 ... hxg4? 12 h5! is strong.]
12 fxg4 Nxg4 13 cxb5 axb5 14 Bxb5 Nxe3 15 Bxc6 Bxd4! 16 Nb5 Bc5 17 Nf1 Nc4 18 Qd5 Ne5 19 0-0-0 e6?!
[Better is 19 ... Qc8 with the unstoppable threat of ... e6.]
[20 Qxc5 dxc5 21 Rxd8 Rfxd8 22 Na7 Rd6 wins for Black.]
20 ... dxe5 21 Rxd8 Rfxd8 22 Nc3 Rd3!
[Black must play actively.]
23 Ba4 Bd4 24 Kc2 Rf3 25 Nd2 Rf4 26 Bb3
[White mobilizes his passed a-pawn.]
26 ... Bf2 27 Nc4 Bxh4 28 Nxe5 Rf2+ 29 Kb1 Bg3 30 Nd3 Rf3 31 Bc4 h4 32 a4 g5 33 a5 g4 34 a6 h3 35 a7 Ra8 36 Nb5 h2 37 Nd4 Rxd3
[The sacrifices continue.]
38 Bxd3 Rxa7 39 e5 Ra5 40 Nc6 Rc5 41 Be4 Bxe5 42 b4 Rc4 43 Bg2 Bd6 44 Kb2 Rf4 0-1
1 e4 e5 2 Nf3 Nf6 3 Nxe5
[3 d4 is the main alternative.]
3 ... d6 4 Nf3 Nxe4 5 d4 d5 6 Bd3 Be7 7 0-0 Nc6 8 c4 Nb4
[This was a new variation to me, but the Informants had several games to examine. Usual is 8 ... Be6.]
[Much of the strategies employed in playing the Petroff involves the light-squared bishop. It is White's most valuable minor, so its preservation is essential.]
9 ... 0-0 10 Nc3 Be6 11 Be3
[Where to place this bishop is a typical problem - the two bishops are a definite asset in this type of game, but posting them effectively can be a problem. The square e3 is not great, but 11 Bf4 only encourages 11 ... Bd6.]
11 ... Nxc3
[This is theory, but I can't help thanking Black for removing his aggressive knight, shoring up my d-pawn, and chasing away the other intruder.]
12 bxc3 Nc6 13 Nd2?!
[White was trying not to tip his hand. Better was 13 cxd5 Bxd5 14 Nd2. I agonized trying to determine what was the best continuation here, as there were few games at this point in theory, and none offered a clear-cut strategy. The point was for 14 cxd5 Bxd5 15 c4 Be6 16 d5 with initiative and possible material gain.]
13 ... dxc4! 14 Nxc4 b5
[Black intends to keep White from establishing the hanging pawns by taking away the c4 square.]
[Everything else was retreat in disarray. The square e3 would be a great post for the knight, except White couldn't decide where to use his dark bishop.]
15 ... Nxe5 16 dxe5 c6 17 f4
[Played more to shore up the e-pawn. White must watch for garbage on the a7-g1 diagonal.]
17 ... Qa5 18 Qc2
Position after 18. Qc2
18 ... Qa3!?
[This was not considered, but Black appears to be not only focusing on white's c-pawn, but also the rest of the third rank. Threatened is 19 ... b4 as well as ... Bc5 maneuvers. Expected was placing a rook on d8, as white has to care for his a-pawn.]
[The dark-squared bishop is shielded by this one, and White begins to focus on a kingside assault rather than wrestling over control of the d-file.]
19 ... Kh8!?
[A very tricky move. On 19 ... f5! 20 exf6 (though White would have a passed e-pawn if he did not capture, the b1-h7 diagonal would be permanently shut) Bxf6 is probably equal.]
20 f5 Bc5?!
[I'm not convinced this is the correct continuation for Black. But 20 ... Bd5 21 Rf2! Bc5 22 Bxc5 Qxc5 23 c4! is better for White due to the weak c6.]
21 Qf2 Bxe3 22 Qxe3 Bd5
[22 ... Bd7? 23 f6 Rg8 24 Bxh7 Kxh7 25 Qd3+ followed by Qxd7 and Qh3; 22 ... Bc4!?]
[Though hazy, with three pieces at the ready (Qe3, Bd3, Rf1) and a battering ram (Pf6), there must be a decisive attack here someplace.]
23 ... g6 24 Rae1
[Black's position is hanging by a thread, but White's king is subject to interpolating harassment that seems to buy tempi and defense for Black. Otherwise 24 Qh6 Rg8 25 Rf4 wins easily - 26 Qxh7+! is mate in two. However, 25 ... Qc5+ 26 Kh1 Qe3! pins the rook and allows ... Qh6 in the event of a sacrifice on h7; or 25 ... Qf8 may allow a saving exchange. After 24 Rf4 Qb2! forces retreat. White has one of those mixed blessing positions: he has a great middlegame, but his pawn structure (except for the pawn at f6) is worse. Thus, White must avoid going into an ending unless forced, though exchanging bishops would be a big advantage for White. The text move is necessary preparation. It overprotects the e-pawn and removes the rook from possible danger when the other forces lift to the third or fourth rank, while the queen serves a purpose by controlling the a7-g1 diagonal until the rook reaches h4. The alternative 25 Rc1 would leave the rook vulnerable at critical moments. The move 25 Kh1 leaves the back rank still vulnerable, and ... Bxg2 is with check if Black needs some sort of distraction from impending mate. Now White menaces 25 Rf4 and if 25 ... Qb2 or 25 ... Qxa2, Rg4 followed by 27 Qh6.]
24 ... Qxc3!
[Black has a tempo, but his rooks are out of play. Interestingly, 24 ... Qxa2 had me more concerned, but this capture also accomplishes two important things: the queen pressures the e-pawn, and has the Re1 in sight.]
[As intimidating as White's attack is, his piece coordination is horrible because the queen can't attack and defend simultaneously, and Black has an array of dangerous zwischenzugs and checks to keep White off balance. The move 25 Rf4 was strongly considered, with 25 ... Bc4 26 Rh4! and now ... Qxd3 27 Qh6 or 26 ... Bxd3 27 Rc1! which allows 28 Qh6 winning outright. However, 25 ... Qa5 threatening 26 ... Qb6 and 26 ... Qxa2 discourage such ideas. Shunting the king out of range slightly I hope will not tip White's attacking plan (the check along the a7-g1 diagonal is removed), but doing so leaves White potentially vulnerable to back rank mates and sudden ... Bxg2+ counter-ploys. White's really annoying problem is if the kingside attack gets shut down, Black has no inherent weaknesses (other than his king position) to exploit in search of compensation for the pawn. The move 25 Rf2 leaves the Re1 vulnerable, but may be the critical and better move.]
25 ... Rad8
[A very straightforward move with a straightforward threat: 26 ... Bxg2+. Black is countering White's initiative nicely.]
[This is forced. White has to break the pin without submitting to a queen exchange.]
26 ... Rg8 27 Bb1
[Better is 27 Bf1 or 27 Be4 but both are unplayable. The move 27 Re3 is countered with 27 ... Bxg2+, again.]
27 ... Qb2
[Only moves can be strong moves. Black keeps hold of the a1-h8 diagonal, and now makes motions about the seventh rank, beginning with a mate threat.]
[The only other choices were 28 Qh3 and 28 Rg1. The latter paralyzes the Re1 because if Re3, ... Bxg2 (again!) picks up the Bb1. It is absolutely frustrating how Black has so few options, but the ones available are 100% effective and not difficult to find.]
28 ... Be6
[This is what discouraged me from playing 28 Qh3. Black undoubtedly has the idea to place something on d2, what I'm not sure.]
[Finally, a rook lift (after five moves have gone by), but Black can pin White in so many ways. The move 29 Rf4 runs into 29 ... Rd4 30 Be4 Qd2 31 Qh4 Bd5 and White must acquiesce to exchanges. White's back rank is more vulnerable than Black's.]
29 ... Rd2
[Who is the hunter, and who is the hunted? The move 29 ... Qd2 is a blatant attempt to exchange queens.]
[White sees a grand opportunity, but needs for Black to play along.]<
Position after 30. Rfe1
30 ... Rf2!
[The text was not considered. It vacates d2 for the queen which would force exchanges, but I was quick to realize all of the e6 swindles are nipped in the bud (the rook is behind the f-pawn) and Black has the perfidious 31 ... Rf5! which would win the e-pawn and undermine the pawn at f6. Note 32 Bxf5 gxf5 opens up the g-file decisively as White must give up the queen to stave off mate.]
[Avoiding future ... Rf5 enticements, but because of the g2 square, White still can't get something over to the h-file.]
31 ... Bd5
[Continuing to bully on White's Achilles, the king position. This will encourage exchanges.]
[About the only move that offers something. The move 32 Rg1 doesn't quite do it.]
32 ... Qxa2?
[More to the point over this ostensibly greedy move seems to be 32 ... Bxe4. Despite the pin/counterpin, the Be4 is still very dangerous, like Rambo in the jailhouse.]
[White is one tempo and one oversight away from saving this. The less obvious 33 Qh3 would only encourage the "one saving move" (33 ... Be6) with tempo.]
33 ... Qd2!
[This isn't the first time I missed an obvious and strong move that I had seen many times before in similar variations. At the time I thought 33 ... Be6 was the only move that didn't drop the game - 33 ... Rf5 (intending ... Rh5) 34 Qxh7+! - most other moves allow it as well. It seemed that White would collect the c6-pawn and the d5-square, but Black still had a strategically won game with the passed connected pawns. This puts all matters to rest. Now if 34 Qxh7+ Kxh7 35 Rh3+ Qh6 wins. White will drop material unless he submits to the passive 34 Qg3 or 34 Qh3 (34 Qg5? Bxe4! wins), else 34 ... Qxe3 wins a rook, e.g. 34 Bxd5 Qxe3 35 Rxe3 Rf1#, or experience 34 ... Rxg2. I really thought of just resigning here.]
[34 ... Qg3 would be a total admission of failure. White is "inspired" if 34 ... Be6 35 g4!? as a hope and a gasp. White is hoping for an oversight; there isn't much else to play on for.]
34 ... Bxe4
[Black isn't playing along with anything. In the meantime I just cemented my 6.5th point in this section, my best result in the Semis, but now it's even doubtful I'll even tie for first.]
35 Rxe4 Rd8
[Bringing up the reserves.]
[Now Black gets to play with an extra piece as white's back rank problems are becoming painfully obvious.]
36 ... Kg8
["Run Away!" he says. Black also recognizes that his passers are major trumps, but "... if I can't mobilize them, I'm sunk." Excuse me, aren't you the one with the initiative?]
[The faux attack continues. The rook is actually safer here than if it remains on e4. The alternative was 37 Qa3 but Black can counter this with 37 ... Qd3. If the Black queen can get on a light diagonal (ideally get to f3), then ... Rxg2! can force more unwanted exchanges, win material, or end in mate.]
37 ... Qd3
[Pity, White could find drawing hopes with 37 ... h5 38 Rxh5! and proceed to expose and harass the king for as long as possible, e.g. 37 ... h5 38 Rxh5 gxh5 39 Qg3+ Kf8 40 Qg7+ Ke8 41 e6 (which may win back the rook, but not regain the lost pawns) or 41 Qg8+ Kd7 42 Qxf7+ Kc8 43 Qxa7 and then make the most of his connected passers. Black instead offers up a pawn for the exchange of queens (38 Rxh7 Qxh3 39 Rxh3 etc.), but with the rook on the seventh, control of the only open file, three connected passes pawns, and White's shaky e-pawn, that ending offers even less hope than this position.]
[Better to go down with pennants up. There is no endgame to play for. How did Frank Marshall come up with so many of his brilliantly cheesy swindles, and could he have the same success in postal?]
38 ... Rf1
["OK, time to harass your king," writes Black, as if he hasn't been doing that indirectly throughout the middlegame. In sum, White can't avoid the exchange of one set of rooks and once the Achilles king is flushed out, Black will easily force a queen exchange. Game over. Flawless defense. A rule of thumb is that a sac'd pawn should be worth three tempi, but White couldn't buy one extra for two pawns for the attack.]