New Column - Queenless Positions

I am starting a new column now, as I promised before. The subject of the column will be positions in which the queens have been traded. I don't say "endgames", because I don't think all positions without queens can be called as such - there are "queenless middlegames", a neglected part of chess, especially for amateur players. Naturally we will also be covering actual endgames, since chess positions tend to simplify.

When does the middlegame end and the endgame begin? This is a tough question to answer. It is not merely a question of semantics. The basic difference between a middlegame and an endgame is that the goal differs. Of course the final goal of chess is always checkmate - but in the middlegame and endgame you have different methods of getting there - different "secondary goals", if you will.

In the middlegame the two main goals are: 1) checkmate by a direct attack; and 2) obtaining a material advantage. Naturally, to reach those goals there are many other tertiary goals - the creation of weaknesses, the activation of the pieces, gaining space, centralization of the pieces, and so on. On the other hand, in the endgame you cannot usually win by a direct mating attack. In the endgame the main goal is to win a pawn and promote it to a queen.

Obviously there are exceptions to both of these - in the middlegame you can sometimes queen a pawn, and in the endgame there are direct mating attacks by small numbers of pieces. But basically the method of play is somewhat different, and this is what differentiates the middlegame from the endgame, in my opinion - not some arbitrary count of how many pieces are left on the board.

So in this column we will be studying this transition from middlegame to endgame, and the various methods of play when there are no queens on the board. Some of the examples will feature my own games (since naturally I know them best) and some will feature the games of other players. I will cover both positions which are technical in nature (i.e. one side has gained a decisive advantage and must convert it) and positions which are not. In fact, I think drawing a line between these two is a problem for some players (including myself) and should not really be done. You obviously have to be aware when you have an advantage, but seeing the position as "technical" is probably not a good mindset.

We will be covering various themes each week, in pretty much random order.

Bishop against Knight, with Weaknesses

The battle between bishop and knight is fundamental to endgames and queenless middlegames. The cats and dogs of chess are often traded for each other, and when this happens much of the battle revolves around the different propensities of each of these animals.

Evaluating a position where there is a bishop against a knight is not usually a simple task of looking at whether the bishop is bad or good and whether the position is open or not. In fact there are many factors which determine which piece is superior, and the process of evaluation really has to just be internalized. You need to study and play many games in order to get a good feel for the evaluation of positions with this imbalance. When another factor is added - pawn weaknesses - and you need to evaluate the position as a whole, it becomes even more difficult.

A basic example is the game between Chekhover and Emanuel Lasker, from the Moscow 1935 tournament. Very soon after the opening, a basic endgame was reached with Black having a bishop against a knight as well as some doubled pawns. In the presence of the knight, the doubled pawns really could be weak - indeed, the f7 pawn later fell. How do you evaluate the following position:

Lasker, who went straight for this ending, evaluated it spot-on: Black has the advantage. But it is not easy to make this assessment so quickly. You need to evaluate which is more important - the superiority of the bishop in the open board, or the damage to Black's pawn structure. You cannot really make the evaluation on sight alone - you have to assess whether the position can really open up, whether the black king will be able to invade.

In particular, if White could place his knight on some square from which - along with the pawns - it establishes a barrier on the entire queenside, then the position would be at least equal. This square is d3 - if the knight could reach d3 before the black king gets to c5, then White would be fine. But due to one tempo (and the position of the pawn on a3, which means that an immediate Nc1 is not possible due to ...Bb2) White is not able to set up this barrier to the black king in time. As a result some pawns are traded, White is left with a weak a-pawn, and the game opens up, after which the bishop becomes a powerful force.

Now we will see a somewhat more complicated endgame that I played earlier in the year. The first was from January. After 15 moves the following position arose in my game against GM Sergey Erenburg:

The opening had went well from me - Black has a slight advantage. The bishop on d4 will have a great diagonal whether it ends up back on g7 or on b6. Black has more center pawns and it will be awkward for White to defend e4. Most importantly, Black can create strong pressure on the c- and b-files.

One factor which could possibly be in White's favor is the queenside majority and Black's isolated a-pawn. The 2 to 1 majority is not an advantage in itself - in fact, these pawns are likely to come under pressure - but is a potential advantage, which Black should not forget about. The game continued as follows:

At move 18 I had to make a critical choice. Allow Bxf6, doubling my pawns and isolating the d-pawn, but leaving me with a much superior minor piece (similar to the Lasker example - obviously the black pawns would be weaker here, but the presence of rooks makes the stronger bishop count for more), or play 18...e5, creating a backward d-pawn.

It wasn't an easy choice for me - although it probably should have been - and various calculations and psychological factors played a part. Unfortunately I chose the wrong path, although the game ended reasonably anyway.

First of all, I was fairly sure that playing 18...0-0 (allowing 19.Bxf6) should lead to a draw. In doing so, I misevaluated the rook endgame which arrives 8 moves later in which I have an extra pawn. I understood that I would not stand better, but figured that it would be a fairly simple draw. As it turned out, it was far more difficult than I thought.

Thus, factors relating to the ultimate result of the game came into play, and almost certainly I would not choose that path against a lower-rated player who I would want to defeat with Black. But it was very tempting to force an easy draw as Black against a strong GM. The temptation to play this way induced me to imagine some tactical ghosts after the more complicated 18...e5.

After 18...e5, White has the chance to make some threats against d6 such as by Rad1 and Bb4. There is also a hint of some tactical possibilities i.e. Nxd6+ Rxd6 Bxe5, or, after ...Ke7, Nxd6 Rxd6 Bb4. In some lines, Black will need to take the e4 pawn, and then comes Rfe1 and there could be some tactics on the two open files. But basically, if White has no forcing way out, Black will chase the knight to a3 (perhaps even prefaced by ...Bc5) and stand very well, with great centralization and continuing pressure on f2 and e4, as well as the open lines on the queenside. As it turns out, White has no special tactical possibilities, so 18...e5 should have been chosen. Here is how the game continued:

Imbalances between bishop and knight are central to queenless positions, and every amateur should learn how to evaluate the resulting positions in order to make exchanges which will allow your minor piece to dominate.