Control of center

[Lyudmil Tsvetkov]

Now, everyone knows that control of center is one of the most important terms, and all programs have an eval for pawns placed centrally or pawns controlling squares of the center. But it is quite probable that not many go further than that and evaluate also control of center by minor pieces. At least my experience is that even many of the top engines do not fully understand this factor. And consequently, miss potential good moves and valuable points.

Many will say that minor piece psqt more or less does the same thing, but I do not agree here. Usually, minors controlling the central d4,e4,e5 and d 5 squares are placed either on c3 and f3 for knights, or d3, e3 for bishops, which means that those squares get lower values than when the same pieces are placed more centrally on d4,e4,d5 and e5. So that for me psqt does not solve the problem of minor piece control over the center. It should be done separately.

I would give some 10cps bonus for each minor piece controlling one of the 4 most central squares, 5 cps when it is an x-ray control by a bishop. (i.e. when there is one own or enemy pawn or piece between the bishop and the attacked central square)

Some diagrams.

[d]2b3k1/1p1n4/4p1n1/4P3/3P4/2N2N2/6B1/6K1 w - - 0 58
White is much better here in terms of minor piece control of center. Nc3 controls/attacks e4 and d5, Nf3 controls d4 and e5, Bg2 controls on an x-ray e4 and d5. This will score for white 50cps total in terms of minor piece control of center - 40cps for the knights and 10cps for the bishop. Black will score just 20cps for the two knights controlling the e5 square. Thus, white has an advantage of 30cps in terms of this factor. On many occasions this might decide the battle.

[d]6k1/1p6/2b1p3/4Pn2/1n1PN3/8/4N3/1B4K1 w - - 0 58
Black is much better here in terms of minor piece control of center. Nf5 controls the d4 square, Nb4 controls d5, Bc6 controls d5 and e4. This scores 40cps bonus total for black. For white, Ne2 controls d4 and Bb1 controls e4. This scores only 20cps for white and thus black is better here in terms of this factor by 20cps.

[d]6k1/1pn2bb1/1n2pp2/8/3PP3/1BN2N2/8/6K1 w - - 0 58
White scores better here, but black also has its assets. For white, Nc5 controls e4 and d5, Nf3 controls d4 and e5, Bb3 controls d5, this makes 50cps bonus in total. For black, Nb6 and Nc7 control d5, Bf7 controls d5 on an x-ray (5cps half bonus) and Bg7 controls e5 and d4 on an x-ray (another 10cps). This makes 35cps bonus in total. So that white leads by 15cps, but the advantage would be artificially and wrongly bigger if you do not consider the two black bishops controlling squares on an x-ray.

Of course, minor piece control of center should be considered only for the middlegame, as in endgame it is pretty much irrelevant. In the opening and middlegame however the importance of this factor is really considerable: much more considerable in any case than many modern engines would think.

For the time being I will leave this at that. I will not be posting any example games, as, in whatever game you look, you will notice the effect of this factor. Its frequence is so tremendous, that it would simply be a sin not to take it into account in one way or another. Of course, much tuning is required for the term to function properly, especially in engines that have focussed their center evals around pawn control, but I think it is doable.

How many of you implement minor piece control of center?
What is your experience, any measurable elo gain from that?

[Hood]

Hi,
do you consider that centre is moving with pieces movement?
Rgds Hood

[Lyudmil Tsvetkov]

Hi,
do you consider that centre is moving with pieces movement?
Rgds Hood

Hi Krzych, (man, those Polish names are difficult to write )

You mean if with game advancing, the e4 and d4 squares move to f5 and g6? It might be the case, but I somehow doubt it.

The main and the extended center are of course a fixed notion, and they are always due bonus in the middlegame, but if I understand you correctly, you mean that there is something more important, more central than the center itself. Of course, there is, and that is the positions of both kings; their squares, their shelters and the enemy king attack upon them are the main and most real center of the game. That is why king safety is so important, I would say by far the most important term.

But after king safety, control of the fixed center is the second most important term. So that the main center, the squares where both kings are located, is moving with the advance of the game and the changing positions of the respective kings. Sometimes they stay on the same square for a long period of time, but frequently they also move around. So that there are 2 main moving centers in chess: the squares where both kings are located.

I do not know if I understood you correcly, maybe you meant something else instead...

[Lyudmil Tsvetkov]

Below a game of mine against SF 4, where it is very evident how SF underestimated control of center, minor piece control included. Now, latest SF probably plays better here, I did not play enough games on the theme with it, but this game just as an illustration.

[pgn][PlyCount "56"]
[Event "Blitz 2m+2s"]
[Site "Sofia"]
[Date "2013.09.22"]
[White "Stockfish 4 64 SSE4.2"]
[Black "Tsvetkov, Lyudmil"]
[Result "0-1"]
[ECO "B51"]
[TimeControl "120+2"]
[Annotator "Tsvetkov,Lyudmil"]
[MLNrOfMoves "28"]
[MLFlags "000100"]

{1024MB, Dell XPS 4Cores} 1. e4 {[%emt 0:00:05]} 1... e5 {[%emt 0:00:04]} 2.
Nf3 {[%emt 0:00:05]} 2... d6 {[%emt 0:00:03]} 3. Nc3 {[%emt 0:00:06]} 3... c5
{[%emt 0:00:02]} 4. Bb5+ {[%emt 0:00:05]} 4... Nc6 {[%emt 0:00:02]} 5. O-O
{[%emt 0:00:03]} 5... g6 {[%emt 0:00:04]} 6. d3 {[%emt 0:00:04]} 6... Bg7
{[%emt 0:00:03]} 7. Nd5 {[%emt 0:00: 06]} 7... Nf6 {[%emt 0:00:07]} 8. Nxf6+
{[%emt 0:00:00]} 8... Qxf6 {[%emt 0:00:10]} 9. Bg5 {[%emt 0:00:05]} 9... Qe6
{[%emt 0:00:02]} 10. Bc4 {[%emt 0:00:22]} 10... Qd7 {[%emt 0:00:02]} 11. c3
{[%emt 0:00:02]} 11... h6 {[%emt 0:00:02]} 12. Be3 {[%emt 0:00:10]} 12... O-O
{[%emt 0:00:04]} 13. a3 {[%emt 0:00:02]} 13... Kh8 {[%emt 0:00:03]} 14. b4
{[%emt 0:00:07]} 14... Qe7 {[%emt 0:00:13]} 15. Rb1 {[%emt 0:00:06]} 15... f5
{[%emt 0:00: 04]} 16. bxc5 {[%emt 0:00:04]} 16... dxc5 {[%emt 0:00:02]} 17. Bd5
{[%emt 0:00:02]} 17... f4 {[%emt 0:00:05]} 18. Bd2 {[%emt 0:00:07]} 18... g5
{[%emt 0:00:03]} 19. h3 {[%emt 0:00:14]} 19... h5 {[%emt 0:00:37]} 20. Nd4
{[%emt 0:00:04]} 20... g4 {[%emt 0:00:13]} 21. Nxc6 {[%emt 0:00:04]} 21... Qg5
{[%emt 0:00:19]} 22. hxg4 {[%emt 0:00:07]} 22... hxg4 {[%emt 0:00:01]} 23. g3
{[%emt 0:00:04]} 23... Qh5 {[%emt 0:00:15]} 24. Ne7 {[%emt 0: 00:04]} 24... f3
{[%emt 0:00:20]} 25. Bc1 {[%emt 0:00:07]} 25... Rf6 {[%emt 0:00:24]} 26. Nf5
{[%emt 0:00:05]} 26... Bxf5 {[%emt 0:00:11]} 27. exf5 {[%emt 0:00:00]} 27...
Rxf5 {[%emt 0:00:10]} 28. Be3 {[%emt 0:00:07]} 28... Qh3 {[%emt 0:00:06]} 0-1
[/pgn]

[d]r1bqk1nr/pp3pbp/2np2p1/1BpNp3/4P3/3P1N2/PPP2PPP/R1BQ1RK1 b kq - 0 7
Look at this. Center pawns are 1 each, e4 for e5, but look at the additional control of the d4 and e5 squares by black. White controls d5 by e4, but only this one. Black controls d4 by 2 pawns, e5 and c5, by Nc6 and Bg7 on an x-ray. (that is why I say x-ray control is important). The e5 square is controlled by black Nc6 and Bg7, and only by white Nf3. So that overall the black control of center, both in terms of pawns and minor pieces, is much much better. SF somehow underestimates that. When you control the center, you control most of the board.

[d]r1b2r1k/pp2q1b1/2n4p/2pBp1p1/4Pp2/P1PP1N2/3B1PPP/1R1Q1RK1 w - - 0 19
Look at that: white can not budge with d4 because of the abundant black control over that square. Bg7 plays an important role here.

So that minor piece control of center is important for several reasons: supporting the own pawns in the center, stopping the enemy pawns, as well as favouring possible pawn and piece exchanges upon those squares. Capturing in the center of the board is always welcome.

[Hood]

That what I meant was that it is not proper in many positions adding evals for controlling e4,e5,d4,d5 squares. The centre is there, only at a start of a game. Then it is moving....
Exaggregated example. Lets assume that a b pawns left the board and pieces are concentrated on the kings side. The centre is for sure not at starting position. Eval shall be added for controlling the squares around consolidation point of the position.
Rgds H.