True Combat, Winning the Won Game Part One

Before I start with this new five part series on "Winning the won game," I want to thank the many people who emailed me saying they liked "True Combat" and asking when the next column was coming. I could never give a clear answer -- but at this moment I can simply reply, "Now!"

I'm going to do something a little different in this series. Instead of focusing on my games as usual, this time I'm going to show games played by my wife, Liz. For those of you who don't know, I got married early this year. We were just blessed with our first child, Nikolay Samuel Taylor, on October 14.

Liz and I met when she signed up for chess lessons (now she gets them for free!). She was about 1500 then, now she's up to 1738. She had one huge problem when she started taking lessons, and even today has trouble with this: namely, the subject of this series, winning the won game. Time and again Liz would get a decisive advantage -- clear, absolutely winning positions -- and then somehow draw or even lose!

Of course my wife is not alone -- the well-known quote says it all: "The hardest thing in chess is to win the won game." I might not be the first to admit it, but I once managed to draw a game against Enrico Sevillano when I was a piece up for nothing!

However, I think the games of my wife might be even more instructive on this subject. I'm sure everyone can relate to her misadventures!

Taylor, Liz (1593) - Malecki, Alex (1485)
Center Game [C22]
National Open/Las Vegas, 6/16/2006

Before showing the moves of this game, let me give you the story: My wife obtains the advantage out of the opening. By move ten she is clearly better against her unprepared opponent. By move thirteen her advantage has reached a highly dangerous level. On move sixteen she wins a clear pawn. One can objectively evaluate her position on move 25 as absolutely winning: she has reached an endgame with material and positional advantage --and the material consists of an extra, protected passed pawn (the strongest of all pawns) which is on the sixth rank!

From move 25 to move 60 she maintains a winning position -- while missing numerous clear wins -- and finally agrees to a draw on move 60, while still having a winning position.

I'm only going to lightly annotate the first twenty-five moves -- it's the last thirty-five that concern me here. How could she not convert such an overwhelming advantage? Why did she take a draw when she could still win? Why did she miss different types of clear wins?

Now let's take a look at the game, and see if we can answer these questions.

1.e4 e5 2.d4 exd4 3.Qxd4 Nc6 4.Qe3 Bb4+

The usual move order is 4...Nf6 5.Bd2 Bb4 6.Nc3 0 0 7.0-0-0 Re8 8.Qg3 Rxe4 9.a3! Bd6 10.f4 when White has compensation for the pawn, as in Kulaots,K-Piarnpuu,L/Tallinn 1998, where White finally won with a kingside attack.

5.Nc3 d6

An early ...d6 in the Center Game is usually a sign that Black doesn't know the opening. He probably was afraid to play the book 5...Nf6 because of 6.e5, but that would be a bad mistake for White after the evident 6...0-0! when White has serious problems on the e-file. Instead White would continue with the normal 6.Bd2, transposing to the Kulaots game above.

6.Bd2 Nf6 7.0-0-0 Ng4

Driving the Queen where it wants to go!

8.Qg3 Bc5

More loss of time!

9.Nh3 Rg8

White already has a practically winning position. Liz's next move is fine, from a computer standpoint -- Fritz thinks it's just about as good as my following suggestion -- but from a human perspective, my 10.Nd5! is a lot simpler. After 10...Be6 11.f3 Nge5 (e3 is covered, and 11...Nf6 loses to 12.Bg5) 12.Nhf4 Black has no play and practically no moves, as 12...Qd7 loses instantly to 13.Nf6+.

10.f3 Ne3

This would not be possible with a Knight on d5, but on the other hand this "attack" just leads to a favorable exchange for White.

11.Re1 Nxf1 12.Rhxf1

White has everything developed while Black's King is stuck in the center.

12...Be6 13.Nd5 f6

13...Qd7 14.Nf6+ wins for White.

14.Nhf4 Bf7 15.Qh3!

White wins a pawn, as 15...h6 obviously loses to 16.Ne6.

15...Qd7 16.Qxh7 0-0-0 17.Qf5!

Correctly going for exchanges while material ahead.

17...Qxf5

Better is 17...Kb8 -- now White gets the e6 square to go with her material.

18.exf5 Rh8 19.h3 Ne5 20.Ne6 Bxe6 21.fxe6 Rde8 22.f4 Nc4 23.f5

A protected passed pawn on the sixth!

23...Nxd2?

Black's only hope was 23...Bd4 24.b3 Nxd2 25.Kxd2 Be5, blocking the e-file as occurs later in the game, though White should of course still win.

24.Kxd2 Rh4 25.Rf4!

Correctly offering a further exchange. As I've said to Liz -- and to my students many, many times, "When you're ahead in material, exchange, exchange, exchange!"

I should clarify this by adding that I mean exchange pieces, not pawns. One tries to make the mathematical ratio slide in your favor. For example, let's say you have a Queen and pawn ending: you have Queen + 2 pawns, your opponent has Queen + 1 pawn. By the generally accepted chess point value, you have 11 points to 10 -- a slight advantage. However, if you exchange Queens, you have a 2 to 1 advantage, double your opponent's material, and should usually win.

Therefore Black avoids the exchange of Rooks here, but this allows White her first clear forced win.

25...Rh5

Twenty-five moves have been played. White has an absolutely winning position. Let s count up Liz's advantages: A powerful Knight in the center of the board; a powerful rook developed to the open fourth rank; another powerful rook behind her passed pawn; and last, but far from least, an extra, protected passed pawn on the sixth rank, just two squares from queening.

You might think such a position would win itself -- but it doesn't work that way. Chess games need to be won. You have to take active measures to transform an advantage, however great, into resignation or checkmate.

Before I cover what Liz should have done now, let's take a lesson from the master of winning won games, Mr. Alexander Alekhine.

Here is one of his famous wins, vs. Tartakover at Vienna 1922 -- Alekhine is White.
1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 dxe4 5.Bxf6 gxf6 6.Nxe4 f5 7.Nc3 Bg7 8.Nf3 0-0 9.Qd2 c5 10.dxc5 Qa5 11.Nb5 Qxd2+ 12.Nxd2 Na6 13.c3 Nxc5 14.Nb3 Nxb3 15.axb3 a6 16.Nd6 Rb8 17.b4 Rd8 18.0-0-0 Be5 19.Nxc8 Rxd1+ 20.Kxd1 Rxc8 21.Kc2 Kf8 22.Be2 Rc7 23.Ra1 Ke7 24.h3 f4 25.Kb3 Rd7 26.Ra5 Bc7 27.Rh5 Rd2 28.Bf3 b6 29.Rxh7 Rxf2 30.Bh5 Rxg2 31.Rxf7+ Kd8 32.Bg4 e5 33.Rd7+ Kc8 34.Rd2+ Rxg4 35.hxg4 f3

White has a clearly winning position, the exchange up for no pawns, and he even has an outside passed pawn. Black does have connected passed pawns that look like they might be dangerous, but Alekhine cuts through the clutter and finds the clearest and best win.

36.Rd5!!

He explains his reasoning very clearly, which I will paraphrase as follows: Black's only trump is his connected passed pawns; these pawns are inoffensive if they can be forced onto the color of the black Bishop, as then the white King can blockade them; if the White Rook gets behind the pawns without loss of time, White also wins easily.

Now the logic of 36.Rd5 is clear: if Black moves the e-pawn as in the game, White gets behind the pawns and wins with 37.Rf5, while if Black pushes with 36...f2, White plays 37.Rd1 e4 38.Kc2 Bf4 (38...e3 39.Kd3) 39.Rf1 e3 40.Kd1 and White wins easily now that the pawns are neutralized.

Tartakover tried pushing the e-pawn and lost as follows: 36...e4 37.Rf5 Bg3 38.g5 Kd7 39.g6 Ke6 40.g7 Kxf5 41.g8=Q Bf4 42.Qf7+ Kg4 43.Qg6+ Bg5 44.Qxe4+ Kg3 45.Qg6 Kg4 46.Qxb6, 1-0.

Short and sweet! Alekhine's logic was as faultless as his moves, and the technical part of the game looks easy. The key phrase above is "looks easy."

Let's go back to our game. Liz -- just like Alekhine -- has an absolutely winning position, material up. What she needs is a clear logical plan to victory. By not finding that plan, but waiting for the game to "win itself" she became discouraged and finally dropped a half point.

Alekhine would have summarily dispatched any foe unwary enough to have fallen into this dreadful position. There is a clear path to victory here. What is it?

26.Rf3?!

Not this!! It's important to note that neither this nor Liz's next 34 moves are huge blunders. She makes some mistakes, yes, but she still has a winning position even when she agrees to the draw.

What she doesn't do is make a plan to win the game. She makes individual moves, but doesn't find a winning idea of several moves together. Following Alekhine, she should have, right here, figured out the win. This is how I would put it:

WIN ONE

White has a passed pawn on the sixth, but the board is too cluttered to queen it in the near future. Advancing it to e7 would only make it weak. So first White needs to prepare the ground. We already know that piece exchanges favor White -- what can we exchange? Ah, Knight for Bishop can be forced, leaving only rooks.

Next, rooks need open files to be effective, and further, it's always helpful if the enemy has two weaknesses, so we can switch from one to another. Black already has one weak point at g7 -- it would be nice to have one more target. With this in mind, all that is required is a three-move calculation to force the win.

Note that the thoughts, the ideas as stated above, take a lot longer than the calculation. The moves are simple, forcing, and easy to calculate -- but clarifying the position in your mind, figuring out what you want and how to get it, is the hard part.

To recap, White wants to exchange Knight for Bishop in such a way that Black gets another weakness besides the one on g7. White accomplishes this simply and easily as follows: 26.b4! Bb6 (forced, for if 26...c6 27.bxc5 cxd5 28.cxd6+-) 27.c4! (Black's pawn weakness will appear at d6, as White's pawn storm will either now or on the next move force c6 or, even worse, ...a6. In both cases White will then take on b6, leaving Black a backward or isolated d-pawn on the open file, which the now advanced c-pawn will keep under restraint) 27 c6 (If Black won't play this right away, and makes a waiting move -- say 27...Rh7 -- White forces it anyway by playing 28.a4 threatening to win a piece) 28.Nxb6+! (the third move: White forces the desired exchange and weakness at d6) 28...axb6. White now wins mechanically with natural moves, for example 29.Re3 Kd8 30.Rg3 Re7 31.Rd3 Kc7 32.g4 Rh8 33.Rfd4 Rd8 34.Kc3 (It's obvious Black has no useful moves; White finishes with a straight forward kingside pawn advance, while still eyeing d6) 34...Ree8 35.h4 Rh8 36.g5 Rh5 37.Rg4 fxg5 38.Rxg5 Rxh4 (38...Rxg5 39.hxg5 Re8 40.Re3 d5 41.f6 etc.) 39.Rxg7+ Kc8 40.e7 Re8 41.f6 Re4 42.f7 wins a rook.

While this variation may seem long, there's actually nothing to it -- if you see the first three moves. And to see the first three moves, you have to "cut through the clutter" as Alekhine did, and find the simple clear win: exchange pieces and create a weakness.

If Liz had reached the position given in the note above after 28.Nb6+, she would have won easily -- but she didn't see the idea. And I would add, the important thing is to look for this idea. Don't just say, "I have a crushing position" but rather ask, "How can I turn my crushing position into a forced win?"

26...c6 27.g4 Rh6 28.Nf4 Bd4 29.c3 Be5

Black offered a draw in this dead lost position -- more on this tactic later.

30.Ng6 a6

Black s only chance to prolong (though not save) the game is the Exchange sacrifice 30...Rxg6 31.fxg6 Rxe6. We will see a similar position later in the game. Since Black avoids this slight counter chance, White now has --

WIN TWO

When Liz first showed me the game, I missed the earlier win with 26.b4 (which requires some thought to see) but I spotted this one right away. I interrupted her moves by interjecting, "Just take the Bishop and it's all over!" This is true. I was really interested to discover why Liz did not capture the Bishop; in fact, she barely considered it.

In essence, she took a rule too seriously.

It's very important to know that in chess, all rules are approximate. For example, it's "wrong" to develop your Queen early in the game, but in this very game, and the Center Game in general (played by such powerful GMs as Morosevich and Kulaots) White's Queen is the first piece developed! Liz knew that contradiction, but here she was led astray by another -- that a Bishop is bad if it is blocked by pawns on its own color. Here she reasoned that her Knight was stronger than the Black Bishop, blocked by pawns at d6 and f6, so she didn't want to take it.

However, one should consider the overall position. Here the Bishop does some valuable duty by blocking the e-file, while White's Knight can be taken at any time, which means losing the star pawn at e6! So best is to exchange right now, which follows the other rule we've already mentioned, namely, "exchange pieces when material ahead."

Here we have two rules in conflict (bad Bishop vs. exchanges) so one must analyze and find the best move.

In fact, White has a forced win, and here the variations are just two moves long. I would describe Win #2 like this: Exchanges are good for White, and once again it is possible to exchange minor pieces. Then White will either get another open file for his rook, or connected passed pawns -- in either case the win will be routine.

Here is the two-move calculation: White should play 31.Nxe5! dxe5 (forced, for if 31...fxe5 32.g5! [two moves!] and White wins by creating connected passed pawns, for example 32...Rhh8 33.h4 Kd8 34.h5 Rxh5 [else 35.h6] 35.f6 Rxg5 36.e7+ Kd7 37.f7 wins a rook) 32.g5! (two moves!) and White wins by prying open the g-file, e.g. 32...fxg5 (else White takes on f6 and penetrates down the g-file) 33.Rxe5 Rf6 34.Rg3 g6 35.fxg6 Rxg6 36.h4 g4 37.h5 Rgxe6 38.Rxe6 Rxe6 39.Rxg4 Rh6 40.Rh4 with a routine outside passed pawn win with the white Rook behind her passed pawn.

I wouldn't ask Liz, or myself for that matter, to calculate to move 40 here. All you have to do is see that taking on e5 can be followed by 32.g5! regardless of how Black captures, and the rest is easy.

While I think Liz missed Win #1 by not stopping to figure out the winning method, here she missed Win #2 by thinking too much about her "good Knight" on the sixth and her opponent's "bad Bishop" -- instead of viewing the big picture, of which Knight and Bishop are just small parts.

31.Ree3 Kc7 32.Rd3

For the next several moves, White should always take on e5, reaching Win #2, while Black should always take on g6, giving practical drawing chances.

32...a5 33.Kc2 b6 34.Rf2 c5 35.Kb3

Win #1, which has been hanging around for a while, now offers itself in its simplest form. Correct is 35.Nxe5! fxe5 (35...dxe5 36.Rd7+ Kc6 37.Rxg7 Rxh3 38.g5 fxg5 39.f6 Rxe6 40.f7 Rh8 41.Rg8+-) 36.Rfd2 Rd8 37.g5 Rhh8 38.e7 Rd7 39.f6 gxf6 40.gxf6 Re8 41.f7 wins a rook.

This variation is so easy to calculate (all you really have to see is that Black must take on e5 with the f-pawn to know the kingside pawn avalanche will be decisive) that it would be impossible to figure out why Liz missed it -- except that she still had the mental block about capturing the "bad Bishop" -- and so never considered the now obvious winning move!

The moral is, remember chess is changeable! Don't stick to just one idea. Furthermore, one should always analyze forcing moves like captures.

One notes that all Liz did over the last several moves was shuffle her rooks. This cannot be done indefinitely! Either you will draw by repetition, or as here, your opponent will finally come up with a good move!

35...Kc6 36.Rfd2 Rxg6!

Black should have taken this Knight ages ago, but now he finally sees this is his best practical chance. Of course he's still lost though!

37.fxg6 Rxe6 38.c4 Re8

Black has made an exchange sacrifice that offers some practical chances -- or to be strictly accurate, offers some practical chances only if White misses the correct next move. Since Liz does do that, the win starts becoming much more difficult.

As it is, the win is still relatively easy, which means it is time for --

WIN THREE

How do we win endings? In most cases, ending are won by queening a pawn, or threatening to queen a pawn and obtaining decisive material advantage.

Because of Black's Exchange sacrifice, White no longer has a passed pawn, but she does have a very important potential passed pawn, due to the two to one pawn majority on the extreme right flank. Therefore a simple advance is decisive: Win #3 is 39.h4!, and if 39...Rh8 40.h5. Then White secures her pawn majority, and ties down the black Rook (which must always monitor the h6 break). One doesn't need specific variations to see that after 39.h4 Black is in a desperate, passive position, while White can win as she pleases. Ideas are Ka4, a3 and b4 to break through on the queenside; Rf3-f5, and g5, to break through on the kingside; and some type of sac on e5, where the other white Rook breaks through to the seventh rank.

In any case, after h4-h5, White secures the full material advantage of the Exchange, and should win with no problems.

39.a4?

Whereas before Liz missed her own two-move threats, now she misses her opponent's two-move threat! This loses a pawn, so the material advantage drops to Rook vs. Bishop + pawn, still a decisive advantage in this position, but the win is much more difficult.

Why is a Rook better than a Bishop? Simply put, the Rook can control more squares: it can cover squares of both colors, and when the Rook has open files, the Bishop is no match for the rampaging Rook!

The big "if" above is "open files" -- but White will soon get a chance to open a crucial line. To begin with, Liz maneuvers correctly, centralizing her King, and sets up the decisive break. However, while she is so maneuvering, she runs into another problem, often encountered in the weekend Swiss -- namely, Topalov style counterplay!

At this point Liz's opponent, who had already offered one draw, began offering her draws on practically every move, thus breaking into her concentration. Not only is this unsportsmanlike, it is also illegal. Article 12.5 of The Official Rules of Chess states: "It is forbidden to distract or annoy the opponent in any manner whatsoever; this includes the persistent offer of a draw."

Liz did not know this rule at the time, but as I said then, one can always go to the director if your opponent is annoying you. This was a clear case of illegal annoyance, and so she should have (and will if she runs into this again!) stopped the clock and gone to the TD.

Instead she suffered in silence, and then didn't have the psychic will to break through when the last forced win beckoned.

39...Rh8 40.Kc2 Rh6 41.Rf3 Rxg6 42.Kd3 Rh6 43.Ke4 g5 44.Kf5 Rh7 45.Rfd3 Rh8 46.b3 Rh7 47.Rd1 Rh8 48.Rh1 Rh6

Everything is in order, which means it s time for --

WIN FOUR

White needs an open file to break through with her Rooks, and there is only one candidate. Therefore, correct and winning is 49.h4!

The variations are clear but rather complex: 49.h4! gxh4 (49...Rxh4 50.Rxh4 gxh4 51.Rh3 Kd7 52.Rxh4 Kc6 53.Rh7 Bd4 54.Rf7 Be5 55.Rxf6 Bxf6 56.Kxf6 d5 57.cxd5+ Kxd5 58.g5 and queens with check) 50.Ke6 Kc7 51.Rdh3 Bg3 52.Kf7 Rh8 53.Rf1 Be5 54.Rf5 Kc6 55.Rh5 Rd8 56.Rd3 Bd4 57.Rxh4 and White will win with a well timed Exchange sacrifice on f6, followed by the advance of the g-pawn.

This is master level play, and I don't blame Liz for not wanting to risk the temporary pawn sacrifice -- but objectively she could still win here. The draw offers coming thick and fast were also a factor -- she was starting to believe her opponent's nonsense and so lost faith that the position really was winning.

Unable to convince herself to break, she commenced to shuffle again.

49.Ke4 Rh7 50.Rf3 Rh8 51.Kd3 Rh7 52.Ke2 Rh8 53.Kf1 Rh7 54.Kg2 Rh4 55.Rf2 Kd7 56.Kf3 Ke7 57.Ke4 Ke6 58.Rff1 Rh6 59.Rf3 Bd4, 1/2-1/2.

Black should have played 59...Rh4 -- instead the Bishop move allows the above-mentioned break, or in other words, Win #4 is still there! White should now play 60.h4! gxh4 (60...Rxh4 61.Rxh4 gxh4 62.Rf5 Be5 63.Rh5 Bg3 64.Rh7+- idea is Kf3 and Rb7 to attack the Black weakness at b6) 61.Rfh3 Rg6 (61...Bf2 62.Kf3 +-) 62.Rxh4 Rg5 63.Kf3 Re5 64.Rh8 Re3+ 65.Kf4 Rd3 (65...Rxb3?? 66.R1h7 d5 67.Re8+ Kd6 68.Rd8+ Kc6 69.cxd5 mate) 66.Re8+ Kf7 67.Rb8 Rxb3 68.Rh7+ Kg6 69.Rbb7 Be3+ 70.Ke4 Bh6 71.Rh8 and the threat of 72.Rg8+ is decisive.

Instead of this accurate win, Liz finally gave in to her opponent's begging for a draw.

What can we learn from this? First, when you realize your position is winning, don't stop there: figure out how to actually win it! Look for a clear plan, which in endings usually means playing for a new Queen. Try to view the totality of the position, rather than just elements, like "better minor piece."

Finally, one could learn a lot by working through Alekhine's crystal clear winning procedures!