The Booster section continues to grow, as it once again was bigger than any before. A total of 51 players, and then two re-entries and one houseman, came to compete in Lansing. The 4-day schedule was the smallest group with only nine players. The many young players prefer the faster time controls. There were 25 players and a re-entry in the 3-day schedule. And then 17 players with another re-entry in the 2-day schedule. A special note about the 2-day was that one players rated under 700 actually made the cut in round one, showing the attractiveness of this schedule to scholastic players.

Rounds 1-4

Most of the games from this section were picked for their instructional value.

Black finds some nice tactics to win a pawn, but then let White come back to build up a kingside attack:

Sheldon Mandell (1099)
Nishant Jain (975)
2-day, Round 2
Queen's Pawn: Zukertort (London), D02

Notes by Tim McGrew

1.d4 d5 2.Nf3 Nf6 3.Bf4 c5 4.c3 cxd4 5.cxd4 Nc6 6.e3 e6

6...Qb6 is reasonable here, underscoring the fact that White's early development of the darksquare bishop has left the queenside weak. For example, 7.Qc2 Bf5! 8.Qxf5? Qxb2-+

7.a3 Qa5+ 8.Qd2 Bb4! 9.Nc3 Ne4! 10.Qc1 Bxc3+ 11.bxc3 Nxc3 12.Bd3

12.Qd2 would prevent the coming discovered check, but White probably didn't want to trade queens now that he's a pawn down.

12...Ne4+ 13.Nd2 Nxd2 14.Qxd2 Qxd2+ 15.Kxd2

Off they come anyway. Now White has to find some comfort in the fact that Black's remaining bishop is passive.

15...0–0 16.g4 e5?!

This bid for freedom is superficially logical, but it lands Black in some difficulties.

17.dxe5 Bxg4 18.Rhg1 h5 19.Be2 g6?

19...Bxe2 20.Kxe2 Rfe8 leaves Black with the upper hand.

20.Bxg4 hxg4 21.Rxg4 Rae8 22.Rb1 f5??

This blunder makes White's task unexpectedly easy.

23.Rxg6+ Kf7 24.Rf6+ Kg8 25.Rg1+ Kh7 26.Rh6# 1–0

Here is a quick endgame:

Ed Mandell (1371)
Kurtis Droge-Germain (1332)
3/4-day, Round 3

Notes by Tim McGrew

Black to Move

41...Rxe4+!

An excellent simplifying combination. It's the shots like this that turn long, grind-me-down endings into walk-away victories.

42.Rxe4 Nxe4 43.Kxe4 Ke6 44.Kd4 c5+!

This pawn is just a decoy.

45.Kxc5 Ke5 46.h5 g5!

Black correctly keeps on his g-pawn. Exchanging would be a horrible error, throwing away half a point.

47.Kc6 Kf4 48.Kd6 Kxg4 0–1

A nice display of technique by Black.

Michael Carter Jr. (1136)
Chris Schmidt (1322)
3/4-day, Round 4
English: Anti-Benoni (Geller), A33

Notes by Edward Laurin

Whoa?! Remember that one? Michael Carter Jr. is one of Monroe's Elite, young and rising stars. Even though he doesn't score a point in this game, I must note that his opponent is also a very talented young man from Kearsley, keep an eye on the tactics: These two are going to be seeing a lot of their games on magazines this upcoming scholastic year, I think.

1.d4 Nf6 2.c4 e6 3.Nc3 c5 4.Nf3 cxd4 5.Nxd4 Nc6 6.Nxc6

White has a nice centralized position here. A good rule of thumb for youngsters (anyone really) is to not trade your pieces so willingly unless you see no other good moves or if the trade does something that benefits you, like brings your opponents king out too early. 6.Bg5 gets another piece out and pins a knight.

6...bxc6 7.Bg5

See? White makes this move anyway.

7...Bb4 8.Qc2 h6 9.Bxf6

See above note.

9...Qxf6

See what has happened with all the trading? White takes his pieces, which are out and about, trades them, and allows Black to get more pieces out. You want to stay ahead in development, and if trading helps your opponent get his pieces out don't do it.

10.Rc1 d5

We see another drawback to White's trades. Black gets more support in the center. Another good lesson: You want to control the center and if by trading pieces you help your opponent to do that don't trade!

11.e3 Ba6?

Oops. Black hurries to get his last piece into the game. But, here comes another good lesson. When you're getting your pieces out you want to try and put them on squares where they can't be easily attacked (especially if they aren't defended!) One of the keys to finding tactics in chess is looking for the under protected pieces (or the undefended pieces). Try using that idea to find White's next winning move:

White to Move

12.Qa4!

Bam! White sees it. The c6-pawn, the bishop on a6 and the bishop on b4 are all undefended and White attacks all three!

12...Bxc3+ 13.Rxc3 Bb7 14.cxd5

14.Rb3 wins quicker, although it is much harder to see. White's plan in the game works just as well! 14...0–0–0 (14...Qe7 15.cxd5 exd5 16.Ba6 Bxa6 17.Qxc6++-) 15.Qxa7+-

14...exd5 15.Ba6

Once again White is attacking two undefended pieces!

15...Bxa6 16.Rxc6

Another nice tactical fork by White.

16...Qxb2 17.Rc2+?

Ouch! White gets over anxious to win a queen with the discovered attack but misses Black's response. After 17.Rxa6+ White will be able to take the a-pawn and go on to a nice win.

17...Bb5!

Black defends with style. Despite the pressure of White's attack, he remains in a calm nature. This is another good lesson: Never give up. Even when it seems like your opponent is going to win a bunch of material play your best moves: You never know when they'll make a mistake.

18.Rxb2 Bxa4 19.Rd2 Bc6 20.0–0 0–0 21.Rc1 Bb7 22.Rc7 Rab8 23.h3 Rfc8 24.Rdc2 h5 25.Rxc8+ Rxc8 26.Rb2 Bc6 27.Rb3 g6 28.Ra3 Rc7 29.Rc3 Rc8 30.Rc5 Bb7 31.Ra5 a6 32.Ra4 Rd8 33.Kf1 0–1

And Black went on to win.

After the merger of all three sections:

4.0: Hodge, Droge-Germain, Van Horn
3.5: Zaremba
3.0: 10 players

Round 5

Here are the pairings of the top four players:

1. Hodge 1 Droge-Germain
2. Van Horn 0 Zaremba

Otho Payne (937)
Michael Carter Jr. (1136)
Caro-Kann: Advance, B12

Notes by Edward Laurin I chose this, another of Michael's games, because of the sacrifice in it. It is important for players to understand some of the thematic sacrifices in chess as they reoccur quite frequently every year. This particular one is called the “Greek Gift.”

1.e4 c6 2.e5 d5 3.d4 e6 4.Nf3 Nd7 5.Nc3 Bb4 6.Bd2 Ne7 7.Bd3 0–0

Player's of these “French/Caro-Kann” positions must be well aware of the “Greek Gift” Sacrifice. White to play and win!

White to Move

8.Bxh7+!

This move works because Black has no pieces on his kingside or in front of the king to protect him, while White has his darksquare bishop, his knight (and soon his queen) all over there!

8...Kxh7 9.Ng5+ Kg8

9...Kg6: This odd move is Black's only hope! But, White still easily wins: 10.h4. This is the usual idea behind this opening. 10...Qb6 11.Qg4. And this is the other. 11...Nf6, trying to chase away the pesky queen. 12.h5+ Kh6 13.Nxe6+ Kh7 (13...g5 14.hxg6#) 14.Qxg7#. This just gives the reader an idea of how to deal with this sacrifice if the king tries to escape by going to g6.

10.Qh5 Re8

This is forced otherwise White will checkmate Black soon.

11.Qh7+

This works. But White can win quicker with: 11.Qxf7+ Kh8 12.Nxe6 and White attacks both the queen and threatens checkmate on g7.

11...Kf8 12.Qh5

Nope! White already started the attack he must keep going. 12.Qh8+ Ng8 13.Nh7+ Ke7 14.Qxg7. Now Black cannot stop White from playing Bg5+, winning even more material. This is a normal position after the Greek Gift and it would do well for every player to know it so you can avoid it!

12...Ng6

Nice. This is the only move that wins for Black.

13.Nh7+ Kg8 14.Bg5

White's attack is gone now.

14...f6?

Black cracks under pressure. Defending is a very difficult thing to do at all levels of chess. However, a good rule of thumb is to never push a pawn in front of your king unless you absolutely have to.

15.Qxg6

This recovers the piece, but White had a winning combo with: 15.Nxf6+ Kf7 (15...gxf6 16.Qxg6+ Kh8 17.exf6 and Black will have to sacrifice material to stop the checkmate on g7. Note that Rg8 doesn't work because of Qh6#.) 16.Nxe8+-

15...fxg5 16.Nxg5 Nf8 17.Qf7+ Kh8 18.Qh5+ Kg8 19.Qf7+ Kh8 20.Qh5+ ½–½

White's best is to go for the three-move repetition here, as there is no way to recover the sacrificed piece. This was a very interesting game full of tactics, but the point isn't to memorize or be able to see those ideas. The point is that with move 8, White enters into a well known position and it benefits the chess player that knows these common sacrifices both for attacking and defensive purposes.

White wins a couple of pawns, but Black has a good counterattack going when he misses a tactic:

Mervin Draper (1052)
Joel Schwiebert (572)
Slav, D10

Notes by Tony Palmer

1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.Bg5 Bf5

Sometimes Black chooses 4...Ne4 to avoid doubled pawns on f6; 5.Nxe4 dxe4 and now White must avoid 6.e3?? Qa5+ and 7...Qxg5.

5.e3

5.Qb3 b6 6.Bxf6 dxc4 7.Qxc4 gxf6 8.e3 b5 9.Qb3 a5 10.a3 Bg7 (Beck-Volkers, Nijmegen 2001) with chances for both sides.

5...e6 6.Nf3 Be7 7.Qb3 0–0?

Drops the pawn on b7. 7...Qb6 8.Qxb6 axb6 and Black is fine.

8.Qxb7 Nbd7 9.Qxc6 Qb6 10.Qxb6 Nxb6 11.b3 dxc4 12.bxc4 Nxc4?!

An interesting idea, trying to regain the piece on the c-file, but this shouldn't work.

13.Bxc4 Rac8 14.Bxe6?!

14.Nd2 Bb4 15.Rc1 and White is a clear piece up.

14...fxe6 15.Nd1 Bb4+ 16.Kf1

16.Ke2 Rc2+ forces 17.Kf1 anyway.

16...Bd3+ 17.Kg1 Ne4 18.h3?

18.Bh4 saves the bishop while defending f2.

18...Nxg5

18...Rxf3! 19.gxf3 Nxg5 wins two pieces for a rook.

19.Nxg5 Be2 20.Nxe6 Rf6 21.Nc5 Rg6 22.Ne4 Bf3

Immune because of the pin.

23.Ng3 Rc2?

But the pin is now broken.

24.gxf3 1–0

While the top two meet, can the others catch up?:

5.0: Hodge
4.5 Zaremba
4.0: E. Mandell, Aguilar, Vorhauer, Droge-Germain, Schmidt, Van Horn, Watkins

Round 6

1. Zaremba 1 Hodge
2. Vorhauer ½ E. Mandell
3. Aguilar ½ Schmidt
4. Droge-Germain 1 Van Horn
5. Browne (3.5) 1 Watkins

Black walks into a shot:

Ted Gregg (1211)
Rishi Narayan (656)
Bird, A02

Notes by Tony Palmer

1.f4

Bird's Opening, a solid alternative to 1.e4 and 1.d4. Black often replies 1...d5 to gain a foothold in the center.

1...e6 2.e4 d6

Solid but passive. 2...d5 offers better chances.

3.d4 Be7 4.Nf3 Nf6 5.Nc3 Nc6

See how White's control of the center plus his huge advantage in space gives him an easy game.

6.Bb5 Bd7 7.0–0 0–0 8.a3!?

8.h3 would be better, then 9.Be3 to complete development.

8...a6 9.Bxc6 Bxc6 10.d5 exd5 11.exd5 Bb5 12.Re1 c6 13.dxc6

13.a4? Bc4 14.b3 Bxd5 and Black wins a pawn.

13...Bxc6 14.Nd4 d5 15.b4 Ne4 16.Bb2 Qd6

16...Qd7 is somewhat better.

17.g3

17.Nf5 Qe6 18.Nxd5! Bxd5 19.Qxd5! Qxd5 20.Nxe7+ Kh8 21.Nxd5 wins a piece. 17.Nf5 Qe6 18.Nxd5 Rad8 19.Ndxe7+ Kh8 20.Bxg7#

17...Rfe8 18.Nxc6 Qxc6?

18...bxc6 was forced.

19.Nxd5 Rad8? 20.Nxe7+! 1–0

20...Rxe7 21.Qxd8+ and White is up a rook.

White goes on a hunt for Black’s king:

Sheldon Mandell (1099)
Kevin Hoyt (Unr)
Queen's Pawn: Zukertort (Chigorin), D02

Notes by Tim McGrew

1.d4 Nc6 2.Nf3 d5 3.Bf4 Nf6 4.h3 Bf5 5.e3 Nb4 6.Na3 e6 7.c3 Nc6 8.Bb5 Qd7?

Breaking a pin with your queen is like blocking a punch with your face. It just doesn't work out the way you had hoped. Black would more or less equalize with 8...Bd6 or 8...Bxa3 9.bxa3 0–0.

9.Ne5! Qe7?

Black is lost, but 9...Qc8 would at least keep the queen in contact with the rook.

10.Nxc6 bxc6 11.Bxc6+ Kd8 12.Bxa8

We could draw the curtain here, but the coming king hunt has some nice touches.

12...Ne4 13.Qb3 Qh4 14.Qb8+ Ke7 15.Qxc7+ Kf6 16.Be5+ Kg6 17.g3! Nxg3 18.Bxg3 Qe4 19.Rg1! Bxa3 20.bxa3

Here 20.Bf4+ is very convincing, e.g. 20...Kh5 21.Qxf7+ Bg6 22.Qxg7 with the threat of 23.Qh6#.

20...Rxa8 21.Bh4+ Kh5 22.Rg5+ Kh6 23.Qf4 Qh1+ 24.Rg1+!

Very pretty!

24...g5 25.Qxg5# 1–0

Michael Carter Jr. (1136)
Joseph Tang (677)
Queen's Gambit: Chigorin, D07

Notes by Tim McGrew

In this tussle between two promising juniors, Michael comes out on top.

1.d4 d5 2.c4 Nc6

If this were home preparation it would be one thing, but to venture the Chigorin without preparation can be dangerous.

3.Nc3

Perfectly playable, though 3.Nf3 is a bit more popular these days.

3...Nf6

A few years ago I experimented in some online games with 3...e5!?, generally with good results. White can keep a small plus with accurate play, but even in that variation Black's position is reasonable. And White has several ways to lose the advantage altogether. Witness this miniature won by one of my students: 4.e3 (4.cxd5 Nxd4 5.e3 Nf5 is a more testing line.) 4...exd4 5.exd4 Nf6 6.Bg5 Be6 7.cxd5 Bxd5 8.Nxd5? Qxd5 Black is already clearly better thanks to White's isolated pawn and lack of kingside development. But now White compounds his problems. 9.Bxf6?? Bb4+! Surprise! 10.Ke2 Qe4# IlPadrino-RichardH, ICC 1998. In any event, this sub-variation of the Chigorin can pose White some interesting problems. If you're searching for a Black weapon against the Queen's Gambit, give it a look. But don't forget to prepare something against 3.Nf3!

4.Bg5 Ne4 5.Nxe4 dxe4 6.e3 a6 7.f3 exf3 8.Nxf3 h6 9.Bf4 Nb4?

This one-horse attack is premature and easily rebuffed. Black should be thinking about getting some more pieces off of the back rank, and 9...e6 is a good way to start.

10.Qa4+ Nc6?

This loses a piece outright. Black is in a bit of a jam, but the tricky 10...Qd7 11.Qxb4 e5!? might be the best way to minimize the damage.

11.d5 Bd7 12.dxc6 Bxc6 13.Qd1 e6 14.Be2

If material were even, White probably wouldn't try 14.Qxd8+ because it develops a Black rook. But here, with a piece in his pocket, White could swap without any hesitation.

14...Bb4+ 15.Nd2 0–0 16.a3 Bxd2+

By exchanging pieces, Black helps White to move closer to a winning ending.

17.Qxd2 Qxd2+

Here again, it would be preferable for Black to keep queens on in the hope of stirring up trouble.

18.Kxd2 Rad8+ 19.Ke1 Rd7 20.Rd1!

White has the right idea.

20...Rxd1+ 21.Kxd1 Bxg2

This doesn't really win a pawn since White can get it back at c7 or h6.

22.Rg1 Rd8+ 23.Ke1 Bc6 24.Bxc7 Rd7 25.Be5! f6?

Probably an impulsive move. 25...g6 was necessary.

26.Bxf6 Ba4 27.Bd4 Rd6?

Oops!

28.Bg4?

Counter-oops! 28.Rxg7+ would mop things up in a hurry.

28...Kf7 29.Rf1+

There is still time to nab the g-pawn with 29.Bh5+ etc.

29...Ke7 30.Bxg7 Rd1+??

This blunder brings matters to a speedy conclusion.

31.Bxd1 Bxd1 32.Kxd1 Kd6 33.Bxh6 Ke5 34.h4 Kd6 35.h5 Ke7 36.Bf8+ 1–0

Zaremba is in the driver’s seat:

5.5: Zaremba
5.0: Hodge, Droge-Germain

Round 7

The champion(s) will come from the top two boards:

1. Zaremba Droge-Germain
2. E. Mandell (4.5) Hodge

Starting with board two, Hodge picked up an exchange just as the scoresheet became undecipherable. Board one quickly went to a rook and knight endgame. Droge-Germain had the advantage with a couple of solid center pawns, but they became overextended and Zaremba won a piece as Droge-Germain tried to defend the pawns.

Anatoli Zaremba is the Champion. A half-point behind in second is Larry Hodge. Taking third on tiebreaks is Javier Aguilar over Sam Vorhauer.

White misses an opportunity:

Brent Kowalski (1174)
Sheldon Mandell (1099)
Queen's Pawn: Mason, D00

Notes by Tony Palmer

1.d4 d5 2.Bf4!?

Not a mistake but definitely unconventional. One of the oldest maxims in chess is “knights before bishops” because the knights have obvious best squares for development (f3/c3 and f6/c6) whereas the bishops have many fine choices, depending on the opening set-up. White does better with 2.Nf3 or entering the Queen's Gambit with 2.c4.

2...Bf5!?

Same comment. 2...Nf6 is good, waiting to develop the bishop for now.

3.Nf3

3.Nc3 e6 4.a3 Bd6 5.Bg3 c5 6.dxc5 Bxc5 7.e3 Nd7 8.Nf3 Ne7 9.Be2 0–0 10.0–0 Ng6 (Mason-Fritz, Wiesbaden 1880) and Black had a good game.

3...Nc6?!

But this is less helpful. For queen pawn openings, it's important to allow the c-pawn to advance. Black has a far better set-up with ...c6 and ...Nbd7 than blocking the c-pawn with 3...Nc6?!. Better would be 3...Nf6 or 3...e6 intending 4...c5 with a good game.

4.a3 e6 5.e3 a6 6.c4 Bd6 7.Bxd6 cxd6 8.Nc3

This is now more like a conventional Queen's Gambit Declined, where the exchange of darksquare bishops helps White.

8...dxc4 9.Bxc4 e5 10.0–0 Qf6 11.dxe5 Nxe5 12.Nxe5 dxe5 13.Nd5 Qc6 14.Rc1!

White is developing with a threat = excellent strategy.

14...Qd6

14...Qd7 should also lose to the winning tactic. Do you see it?

White to Move

15.Ba2

15.Nc7+! wins a rook, or else the queen for two pieces 15...Qxc7 16.Bb5+! Qd7 17.Bxd7+ Bxd7 18.Rc7 Rd8 19.Qd6 with a won game for White.

15...Rc8?!

Much better is 15...Rd8.

16.Qf3

16.Nc7+! wins the queen again 16...Qxc7 17.Rxc7 Rxc7 18.Qd6 and the rest is mop-up.

16...Rxc1 17.Rxc1 Be6 18.Nc7+

Still good, but the golden opportunity was lost.

18...Ke7 19.Nxe6 fxe6 20.Qxb7+ Kf6 21.h3!

Wisely protecting the back rank before advancing the rook to c6.

21...Nh6 22.Rc6 Rb8

This loses two pawns, although Black's position is hard to hold in any case. 22...Qd1+ 23.Kh2 Re8 24.Qxa6 Qd7 25.Rd6 leaves Black in a bind.

23.Rxd6 Rxb7 24.Rxe6+ Kf5 25.g4+ Ke4 26.b4 Nf7 27.Rxa6

White has won three pawns and stands to win the endgame. Black's next quickens the end.

27...Kf3?? 28.Bd5+ e4 29.Bxb7 Ng5 30.Ra5 1–0

And White closed out the endgame.

Vishnu Rajendran (759)
Forrest Reddick (1061)
King's Gambit Accepted: Knight's Gambit, C34

Notes by Tim McGrew

Sometimes it is instructive to see how a winning position gets turned into a losing one. White gets a beautiful game in the opening, but then he suffers from one-dimensional thinking.

1.e4 e5 2.f4

What is the point of the King's Gambit? I suppose some people think that it is a kingside attack, but that is not the primary positional aim. The positional idea behind the gambit is to establish a strong center. White offers the f-pawn as a decoy to lure Black's e-pawn away from the control of d4.

2...exf4 3.Nf3

Stopping any obnoxious checks at h4.

3...Bc5?

From a static perspective this looks great: Black's bishop takes up the longest open diagonal on the board. But the bishop suffers from exposure here -- it will be kicked by d4, after which that diagonal won't look nearly so impressive.

4.Bc4

This developing move is not bad, but White would do well to play d4 immediately.

4...Nh6?!

Why this square? Probably Black fears that after ...Nf6, his knight could be chased by e5. There is an interesting sort of pattern recognition at work here -- Black sees this pattern but overlooks the similar pattern that White puts into practice with his next move.

5.d4

About time! Now Black must retreat.

5...Bd6

Black wants to hold onto his extra pawn -- his only consolation for the fact that he has no center.

6.0–0

Perfectly reasonable development. White has a clear advantage at this point.

6...Ng4

This non-developing move is more or less pointless. But it draws out a strange reaction from White.

7.h4?

With this move White throws away nearly all of his advantage. But why is it weak? And why did White play it? Part of the answer to the first question is that the move weakens the kingside and wastes time. But another part of the answer lies in the nature of a lead in development: of all of the sorts of opening advantages, time is the most easily squandered, and there is usually nothing left to recover after a bad move or two like this. White played h4, I think, because he saw two things: the weak pawn at f7, and the undefended knight on g4. He wanted to play Ng5 to hit both spots at once. Of course Black could retreat with ...Nh6 covering f7, but then White could reply with Bxf4, looking very active. But it would have been better to play 7.e5 at once and grab the f-pawn without creating any weaknesses or wasting any moves. This is a forcing move: Black must lose time again, and White will have a magnificent advantage in space to add to his advantage in time. Black's knight will probably end up retreating to h6, but then White can add pawn structure to his assets, e.g. 7...Be7 8.Bxf4 0–0 9.h3 Nh6 10.Bxh6 gxh6 11.Nc3 and with advantages in space, time, and pawn structure White has a winning game.

7...c5?

Black misses White's idea. It is important to keep White's knight out of g5 -- otherwise even a slow plan can work. 7...h6 is not very impressive, but at least it prevents Ng5.

8.Ng5!

Now White seizes the initiative.

8...Nh6 9.Qh5

I would still vote for 9.Bxf4 Bxf4 10.Rxf4 with a clear advantage for White.

9...g6?

It would be preferable to castle.

10.Qxh6

Now White is clearly winning.

10...cxd4 11.Qg7! Rf8

The moment of truth: White's advantage is overwhelming, both in position and in force. This is the time to look carefully at all forcing moves.

12.e5?!

A step in the wrong direction. White sees that he can knock Black's bishop around, but there is something better here. 12.Bxf7+! is the knockout blow. It isn't too hard to calculate things out to mate here after 12...Ke7 13.Bxg6+ Rf7 14.Qxf7#. But even if White found this difficult to visualize, the capture on f7 should have seemed natural on general principles.

12...Bc5

Black is really in an intolerable situation so there is no sense criticizing his moves.

13.Rxf4?

Missing a second shot at the f-pawn. 13.Bxf7+ Ke7 14.Qf6#.

13...d6

This still doesn't stop the mate!

14.Nxh7

Strike three. White is still winning, though. 14.Bxf7+ Kd7 15.Be8+! Kxe8 16.Rxf8#.

14...d5

I wonder whether Black set this up on purpose...

15.Qxf8+?? Bxf8

Oh agony! And yet, unbelievably, White should still be winning here.

16.Bb5+??

This throws away the win. 16.Nf6+! Ke7 17.Nxd5+ Kd7 18.Rxf7+ Kc6 19.Rf6+ Kc5 20.Nd2!! (White threatens 21.b4#.) 20...a5 21.b3! Now the threats of 22.Ba3# or 22.Ne4# cost Black his queen. An amazing line!

16...Bd7

Black plays the most natural move, but this boxes in the king. 16...Nc6 would force White to take a perpetual check: 17.Nf6+ Ke7 18.Ng8+ Ke8 19.Nf6+=.

17.Bxd7+??

This automatic capture finally lands White in a lost position. Unbelievably, 17.Nf6+ still works: 17...Ke7 18.Nxd5+ Ke6 (18...Ke8? 19.Nf6+ Ke7 20.Bxd7 Nxd7 21.Rxd4! wins material since 21...Qb6? drops the queen to 22.Nd5+ etc.) 19.Bc4! Kxe5 20.Rxf7! and White threatens 21.Bf4+ Ke6 22.Nc7#.

17...Nxd7 18.Rxd4 Qb6 19.c3 0–1