Assorted tweaks

[Lyudmil Tsvetkov]

Well, this is a very well known rule, but I do not know if all engines score it.

In any case, it makes no less sense than king in the eg attacking enemy pawns.

Any king adjacent to an own passer in the eg will get some 10cps bonus for being able to support the advance of the passer, and sometimes also for finding shelter behind such a pawn.

Adjacent will mean of course king attacking the own passer.

[d]8/8/8/6kp/2PK1p2/8/8/8 w - - 0 1

Kd4 is due the bonus above for attacking own c4 passer, and also black Kg5 for attacking both f4 and h5 passers. So the black king gets larger bonus on that.

Trivial as you might think this rule is, it should be very efficient, not less than king attacking enemy pawns.

[Lyudmil Tsvetkov]

A defended passer with blocked defending pawn is any defended passer whose defending pawn is blocked by an enemy pawn.

[d]6k1/8/6Pp/1pP4P/1P6/8/8/6K1 w - - 0 1
c5 and g6 are defended passers with blocked defending pawns

c5 is defended by own b4, which is blocked by b5, while g6 is defended by own h5, blocked by h6

Any such passers are due some 5cps overbonus - small, but fully deserved.
The bonus is due both for the mg and the eg.

One reasonable condition is to give the bonus only to more advanced passers on ranks 5,6 and 7.

Why is the bonus due?

Well, in this case the defending pawn being blocked by an enemy pawn means that the passer itself is more secure, as the defending pawn will be more difficult to attack and destroy.

The bonus is much smaller than in the case of spearhead passers, when you can give straight away 20cps, but the condition is much more frequent.

I would be very surprised if you get nothing from this, though, when it comes about tuning, I am surprised of absolutely nothing.

[Lyudmil Tsvetkov]

Well, controlling the square in front of doubled isolated pawn is always due some small bonus, maybe some 2-3cps, as a pawn controlling this square stops in their movement forward not only the doubled isolated pawn, but also the pawn that goes behind it.

In this way, one pawn stops 2 enemy pawns.

[d]6k1/p7/8/p3p3/4p3/1P4P1/5P2/6K1 w - - 0 1
the bonus goes to b3 pawn, controlling the square in front of a5, and f2 pawn, controlling the square in front of e4.

This is a good rule, although the bonus is small.

[Lyudmil Tsvetkov]

What is cohesive control in front of weak pawns?

Control of the square in front of enemy weak - backward or isolated - pawns, when the pawn controlling it is a part of a larger group, either a duo pawn, or a chain/defended pawn. In this way, not only the pawn that controls the square in front of the enemy weak pawn makes it weak, but the entire larger group; this only increases the weakness of the pawn.

In all such cases, a small, but well-deserved bonus is due, maybe some 2-3cps.

[d]6k1/2p5/1p2p3/1P3p2/P4P2/6PP/1P6/6K1 w - - 0 1
b5 chain pawn gets the bonus here, as it makes c7 backward, while being defended, and also f4 chain pawn gets the bonus, as it is defended and makes e6 backward

[d]6k1/8/1p6/3p4/PP6/4PP2/6P1/6K1 w - - 0 1
a4 duo pawn, controlling the square in front of b6 isolated pawn and e3 duo pawn controlling the square in front of d5 isolated pawn get the bonus here.

same bonus would also be due if the respective white pawns were chain/defended instead of duos.

Again, this is for the fact that the whole group influences the enemy weakness.

I think this is a very good chess rule, although the bonus is small.

[fern]

Pal, have you written a chess program?
If not, you should.
So many ideas.

Fern

[Henk]

Pal, have you written a chess program?
If not, you should.
So many ideas.

Fern

He gets a bonus of 50 cps if his new chess program beats Skipper and 2-3 cps when it beats Pos.

[Lyudmil Tsvetkov]

I see Roger still has not opened my new thread.

I understand people to be stingy, but to be stingy about a couple of Kbytes?

Well, blocked pawn on the 6th rank has been discussed amply in the past, I am sure at least some programmers have something along these lines in their code, although it is not very clear where and how exactly to apply the bonus, as there are blocked pawns on the 6th rank in the mg, blocked pawns on the 6th rank in the eg, pawns that are defended and pawns that are not defended, pawns that are storming pawns and pawns that are not.

My current impression is that, as long as a very neat way to address the issue is not found for the mg, so that the bonus there does not clash with storming pawns bonus, possible wedges bonus, etc., it might make very much sense to consider such blocked pawns on the 6th rank only for the eg, where the bonus could hardly clash with anything else.

Of course, the eg bonus for a blocked pawn on the 6th rank will be mainly due for the fact that such a pawn is a dangerous future passer.

The bonus might be some 20-30cps, so not the biggest, but a nice bonus really.

[d]6k1/p4rp1/P2p3p/4p3/4P3/3P2P1/7P/1R4K1 w - - 0 1
obviously, the white a6 pawn is very close to promotion, if a7 falls. That makes an awful lot of difference in the current position.

So, basically, I think giving any blocked pawn on the 6th rank the bonus is a very good rule; obviuosly, it will fail sometimes, but all chess rules sometimes fail.

[Lyudmil Tsvetkov]

You might try this idea, although I do not know if it is working: give some penalty, maybe 10cps, for any own passer that has an enemy pawn on the same file on the rank just behind it.

The penalty will be due of course for the fact that such passers are impossible to support from behind with the heavy pieces, meaning they are much more vulnerable.
Sometimes search might see nothing troublesome with such a passer, but the danger is still there, as frequently the relevant lines are quite deep.

[d]6k1/8/3Pp3/3p4/8/3PP3/8/6K1 w - - 0 1
d6 passer has an enemy pawn, d5, on the same file, on the rank behind it

d6 is certainly derived from the own forces, and could easily fall prey with other pieces on the board, as it can not be protected from behind with heavy pieces along the d file

Search will not always see that, especially in the mg, as the associated lines are often deep.

[d]2b3k1/1pP3p1/2p1n2p/8/8/5PB1/5NPP/6K1 w - - 0 1
c7 passer has an enemy pawn, c6, on the same file on the rank just behind the passer

c7 is also derived from the own forces, and could easily fall.

Usually, this rule will concern advanced passers on the 5th, 6th and 7th ranks.

I think the rule is meanigful, but who knows if working.

[Lyudmil Tsvetkov]

Man, how I would like to ask Stefan to push a second patch on this theme with a smaller bonus, maybe 20-25 instead of 50cps.

Usually, I do not mess up with programmers at all, but when someone pushes a patch on the framework, I simply go crazy...

It is a pity that no one visits this thread, as otherwise I could have asked someone to ask Stefan for a second try, at least to have something to follow on the framework...

[Lyudmil Tsvetkov]

Lyudmil, if you happen to read this, could you please ask Stefan to run one more attempt with smaller, maybe 20-25cps bonus instead of 50?

Also, what happened to king psqt, did you manage to try it out locally?