Play Like a Grandmaster (Kotov)

Play Like a Grandmaster by Alexander Kotov

Why should we be concerned at all with a chess program that plays a positional game of chess? If we play a strong tactical game (like the current traditional chess programs), shouldn't that be enough?

Grandmaster Kotov describes the 2 fundamental types of positions that one is likely to see in a high-level chess game:

p.14 There are two main types of position[s], and resulting from that, two different kinds of struggle. In the one case we get a constant clash of pieces mixing it in tricky patterns, with tactical blows, traps, sometimes unexpected and shattering moves. In the other case it is quite different. The respective armies stand at a distance from each other, the battles are restricted to reconnaissance and minor sorties into the enemy position. The thrust of the attacking side is prepared slowly with the aid of piece regrouping and 'insignificant' pawn advances. We may call positions of the first type combinative-tactical, of the second type manoeuvring-strategical.

Each type of position requires a specific type of strategy. Note that there are consequences for approaching the 2 fundamental types of positions with the wrong strategy:

p.15 If you make plans in sharp tactical positions, you can easily fall into a trap that figures in the calculations you failed to make. Vice versa, if you are going to calculate variations in positions where you should be thinking about general planning, you will waste precious time and will not get the right orientation. So let us commit firmly to memory the fact that the mind of a grandmaster is principally occupied, in combinative-tactical positions, with the calculation of variations; in manoeuvring-strategical positions, with the formulation of general plans and considerations.

A computer that plays a strong tactical game of chess (and weak positional game of chess) is likely to face opponents that will avoid tactical positions.

In a recent tournament of correspondence games between correspondence Grandmaster Arno Nickel and the hardware-based supercomputer Hydra, Nickel steered the games into openings that required good positional moves from the computer. The machine, known for its tactical abilities, was unable to capitalize on the many 'adequate' positional moves it made, and it either lost or drew the games of the tournament with GM Nickel.

GM Nickel attempted to repeat this strategy in a recent correspondence game against the members of chessgames.com. Using a combination of computers and human reasoning, the chessgames members were able to find deep positional moves and were able to defeat the correspondence grandmaster. The World vs. Arno Nickel 2006

In Kotov's mind, the "attack" and the art of attack forms the basis for play. Therefore it is no surprise that attacking is the centerpiece of middlegame play.

p.22 For greater clarity we shall try to express the concepts of middlegame theory in short exactly formulated points.

1. In chess only the attacker wins.

2. The right to attack is enjoyed by that player who has the better position.

3. The side with the advantage has not only the right but also the duty to attack, otherwise he runs the risk of losing his advantage.

4. The defender must be prepared to defend, and to make concessions.

5. The means of attack in chess are twofold, combinative and strategical.

6. The attack must be directed at the opponent's weakest spot.

Once again, is the ability to play positional chess an option, or is it something that is required in order to win at high level competitive games?

p.24 The player who wishes to improve, who wants to win in competitive play, must develop his ability to assess positions, and on that basis to work out plans for what comes next... every plan is intimately liked with assessment. The analysis and assessment enables the player to find the weak point in the enemy camp, after which his thoughts will naturally be directed towards exploiting this weakness with the aid of a deeply thought out plan campaign with takes account of the slightest nuances of the given position.

What exactly is meant by 'positional play'? What exactly are we looking for?

p.25 The task of the positional player is systematically to accumulate slight advantages and try to convert temporary advantages into permanent ones, otherwise the player with the better position runs the risk of losing it.

Additionally, from Aron Nimzowitsch's My System (2007, Quality Chess edition):

p.178 A typical and widespread misconception on the part of the amateur we are speaking about starts from the idea that every single move must immediately achieve something; this means that he is always looking for moves which threaten something or which parry a threat by his opponent, and in doing so he totally neglects all other possible moves, such as waiting moves, or moves to reposition his pieces, etc. We must really insist that this way of looking at things is quite wrong. On the whole, positional moves are neither threats nor defensive moves; the way I see things, it is much more a question of moves which are intended, in the wider sense, to consolidate our position and therefore it is necessary to bring our own pieces into contact with the squares which are strategically important for us or for our opponent (see later under the headings 'the struggle against freeing moves by your opponent' and 'overprotection').

From Kotov we can derive the requirement that a computer chess program, in order to succeed at the highest levels of competitive play, must be able to perform strongly both in positions requiring tactical moves and in positions requiring positional play.