Center control--

[Henk]

You often read in books that if you occupy or control the center an attack on your king can not be successful.

So I tested this by assigning a too large weight to center control.

Result was disappointing. Even if you have total control on the center you still get check mated.

Counter example: two (or of it that fails) three majors on seventh row.


Maybe only effect that it is a bit more difficult for your opponent to attack the king.

[Ras]

You often read in books that if you occupy or control the center an attack on your king can not be successful.

As a rule of thumb, this is correct - but it is not the centre control itself that is the point here. Rather it is the counter-attack in the centre, and that has to be carried out before the opponent's wing attack breaks through.

The strategic reason is that your pieces in the centre can both defend on the wing where the opponent attacks AND attack on the other wing, especially in case of opposite castling. That's why it is important to prevent centre attacks before launching a wing attack.

That leads to the conclusion that centre control is useless if no actual breakthrough is possible e.g. because the centre pawn formation is a blocked one.

Since many programs are unable to deal with blocked centres and the involved long-term planning on the wings, a workaround is to avoid opening lines that lead to blocked centres and to give the computer a penalty for positions with blocked centre.

With the example of two major pieces on the 7th rank, black obviously has waited much too long with that centre attack so that the window is closed and the game is lost.

[Henk]

My evaluation checked if center pawns were both weak and blocked. Maybe I try again and check if they are blocked only regardless if they are weak.

[Ras]

My evaluation checked if center pawns were both weak and blocked.

OK, but whom do you give a penalty? In the CT800, it is only the computer that gets a penalty for a blocked centre, regardless which side it plays.

Blocked is not only pawn on pawn on the same line, but also inability to push a pawn forward because two opposite pawns are attacking the square while no pawn is available to cover the pawn to advance.

[Henk]

My evaluation checked if center pawns were both weak and blocked.

OK, but whom do you give a penalty? In the CT800, it is only the computer that gets a penalty for a blocked centre, regardless which side it plays.

Blocked is not only pawn on pawn on the same line, but also inability to push a pawn forward because two opposite pawns are attacking the square while no pawn is available to cover the pawn to advance.

So original definition weak AND blocked was ok. Maybe I will post a game if I think it had full control over center but still lost by a checkmate combination.

[Ras]

no, weak is not the point here. A weak backward pawn is only weak if the opponent has a semi-open file to attack it. That is not the same as a pawn that cannot advance because two enemy pawns control the advance square, but no helper pawn is available to cover the advance. I mean, without a semi-open file.

Besides, the whole point of keeping the centre dynamic is to attack in the centre as soon as the opponent distracts resources from the centre to get a wing attack going. Doing that attack too late will end up with you controlling the centre while facing a mating attack.

If you evaluate centre space as a parameter, this can add up to the trouble since it might be attractive to block the centre while gaining space. The CT800 avoids this problem simply by ignoring the space parameter and going for mobility instead, and inflicting a computer penalty for a blocked centre.

A blocked centre is one where at least two pawns c-f cannot advance because they are either physically blocked or because the advance square is controlled by two enemy pawns without a helper pawn as cover.

The result is that the centre is either kept dynamic or ripped open, and in both cases, a wing attack is difficult.