Assorted tweaks 2

[Lyudmil Tsvetkov]

I will post another short thread with some eval suggestions, even if Jose calls me a troll again.

I can take that risk, as long as there is even a single person reading this thread.

[Roger Brown]

I will post another short thread with some eval suggestions, even if Jose calls me a troll again.

I can take that risk, as long as there is even a single person reading this thread.

Hello Lyudmil,

You already have a post on the front page entitled Assorted tweaks.

Please append these ideas to that thread in order to keep the board tidy and not to have a multiplicity of threads on a common or similar topic.

Thank you.

Later.

[velmarin]

I will post another short thread with some eval suggestions, even if Jose calls me a troll again.

I can take that risk, as long as there is even a single person reading this thread.

Although already has answered the moderator in a perfect way.
I follow him from the beginning, I am older than you in the Forum, seen to have even translated his work privately.

But you turned in an elitist way, at this time has been moved to a safe bet, the number one, Stockfish, This is very easy for anyone..
That's not bad.
But that does not entitle you to insult or disparage the work of other amateur chess, computer programming or simply play.
Follow being impartial, is prettier.

[Lyudmil Tsvetkov]

I noticed this just now.

It has been there all the time, subconsciously I knew it existed, but I noticed it just now.

This is also the important missing link to correctly evaluating long chains.

Well, this is only for authors who have chain/defended pawns in their eval, considered at least in terms of ranks.

Engines who consider this, SF for example, would give bonus to any chain/defended pawn in respect of the rank where it is.
That is OK, you can certainly also add a file distinction, even better.

However, there is one more element to assessing defended pawns that is of crucial importance. And that will be whether the pawn is defended by a more central pawn or a less central pawn.

Try guessing if pawns defended by more central pawns are worth a bigger bonus or vice-versa?

Of course, pawns defended by less central pawns will be worth a higher bonus, as this supposes efficiency of resources: as you assess and score only the defended pawn, it is better when you are able to defend such a pawn with a less valuable pawn, rather than with a more valuable one, i.e. with a pawn on a less central file than the file where the defended pawn is.

In such a case, a more valuable pawn on a more central file will be able to assume other, possibly more important functions.

Centralisation will be defined in the following way: d and e files more central than c and f files, which in turn are more central than b and g files, which in turn are more central than a and h files.

Thus, you will have a less central pawn defending a more central pawn, when pawn on file a defends pawn on files b, or pawn on file b defends pawn on files c, or c defends d, or h defends g, or g defends f, or f defends e.

You will have a more central pawn defending a less central pawn, when pawn on file b defends pawn on file a, or pawn on file c defends pawn on file b, or d defends c, or g defends h, or f defends g, or e defends f.

Whenever a less central pawn defends a more central pawn, you might give some 5cps bonus over the available bonus points.
The opposite case might just be left unattended, so you give neither a penalty, nor a bonus there.

In this way, your engine will prefer to defend its pawns with less central pawns, when possible.

[d]6k1/3p4/2p5/8/8/2P5/1P6/6K1 w - - 0 1
you might think that c3 and c6 defended pawns above are one and the same pawn with flipped colours, deserving the same bonus in terms of ranks and files, a defended pawn on the 3rd rank, file c, but no - c3 is a more valuable defended pawn than c6, as c3 is defended by a less central pawn, b2, while c6 is defended by a more central pawn, d7.

In this way, c3 gets the additional 5cps bonus, while c6 gets nothing.

[d]6k1/5p2/2pp4/1p2p3/1P2P3/P4PP1/8/6K1 w - - 0 1
you might think that b4 and b5 are one and the same pawn with flipped colours, but no - b4 is more valuable and deserves some overbonus, as it is defended by a pawn on less central file, a file, while b5 is defended by a pawn on a more central file, c file; similarly, you might think that e4 and e5 are one and the same pawn, but no - e4 is more valuable, as it is defended by a pawn on less central file, f3, while e5 is defended by a pawn on the d file, d6, that is as central as the e file. (in the case of e pawn defended by d pawn or d pawn defended by e pawn, as there is not a less central and more central files, but both files are equally central, you might just not give the bonus)

You can not imagine how important this rule is, especially as defended pawns are a very frequent phenomenon.

Any thoughts, comments on this?

[Lyudmil Tsvetkov]

I will post another short thread with some eval suggestions, even if Jose calls me a troll again.

I can take that risk, as long as there is even a single person reading this thread.

Hello Lyudmil,

You already have a post on the front page entitled Assorted tweaks.

Please append these ideas to that thread in order to keep the board tidy and not to have a multiplicity of threads on a common or similar topic.

Thank you.

Later.

Hello Roger.

Please be so kind to allow me to post here.

I have many threads with eval suggestions, but I try to keep the maximum number of subthreads to 5, otherwise it gets a real mess and no one would read it. I hope you understand me.

The last thread, called assorted tweaks, has eval suggestions, but is unrelated. That is why I opened assorted tweaks 2, to keep it tidy and clean.

Assorted tweaks 1 and 2 is like Terminator 1 and Terminator 2.

Please note that it is probably easier for me not to post at all, I might instead team up with some good programmer and keep my ideas private to try create a very strong engine, but for the time being I prefer that those are public domain.

Later.

[Lyudmil Tsvetkov]

Again, this is an extremely important eval feature, associated with chain/defended pawns. By not scoring this right, you miss an awful lot of chain evaluation.

For chain/defended pawns, in sharp distinction to levers, it makes very much sense to consider if the respective pawn is defended by a more central or less central pawn, as defended pawns are more or less a stable, medium to long-term eval feature.

Levers, on the other hand, are mostly ephimeral, do not last much on the board, though of course there are exceptions even here, so considering more central or less central levers is strictly dependant on tactics, which means unreliable.

With chain pawns you not only can, but could I say should if possible, consider more centrally or less centrally defended pawns.

In a tuned engine, it is not clear how the rule should work, but it is a genuine chess rule, so maybe after a lot of efforts a way out could be found.

[Roger Brown]

I will post another short thread with some eval suggestions, even if Jose calls me a troll again.

I can take that risk, as long as there is even a single person reading this thread.

Hello Lyudmil,

You already have a post on the front page entitled Assorted tweaks.

Please append these ideas to that thread in order to keep the board tidy and not to have a multiplicity of threads on a common or similar topic.

Thank you.

Later.

Hello Roger.

Please be so kind to allow me to post here.

I have many threads with eval suggestions, but I try to keep the maximum number of subthreads to 5, otherwise it gets a real mess and no one would read it. I hope you understand me.

The last thread, called assorted tweaks, has eval suggestions, but is unrelated. That is why I opened assorted tweaks 2, to keep it tidy and clean.

Assorted tweaks 1 and 2 is like Terminator 1 and Terminator 2.

Please note that it is probably easier for me not to post at all, I might instead team up with some good programmer and keep my ideas private to try create a very strong engine, but for the time being I prefer that those are public domain.

Later.

Hello Lyudmil,

I do not know what to make of your implicit "threat" to cease from posting and to keep your ideas private, but I am insisting that you comply.

Should you decide to stop posting then that would be unfortunate, but you must admit that you have been given unusual latitude in terms of threads started.

Should this thread be extended any further, then I may have to lock it.

I am hoping that you will understand my perspective.

Your move...

Later.

[Lyudmil Tsvetkov]

I will post another short thread with some eval suggestions, even if Jose calls me a troll again.

I can take that risk, as long as there is even a single person reading this thread.

Hello Lyudmil,

You already have a post on the front page entitled Assorted tweaks.

Please append these ideas to that thread in order to keep the board tidy and not to have a multiplicity of threads on a common or similar topic.

Thank you.

Later.

Hello Roger.

Please be so kind to allow me to post here.

I have many threads with eval suggestions, but I try to keep the maximum number of subthreads to 5, otherwise it gets a real mess and no one would read it. I hope you understand me.

The last thread, called assorted tweaks, has eval suggestions, but is unrelated. That is why I opened assorted tweaks 2, to keep it tidy and clean.

Assorted tweaks 1 and 2 is like Terminator 1 and Terminator 2.

Please note that it is probably easier for me not to post at all, I might instead team up with some good programmer and keep my ideas private to try create a very strong engine, but for the time being I prefer that those are public domain.

Later.

Hello Lyudmil,

I do not know what to make of your implicit "threat" to cease from posting and to keep your ideas private, but I am insisting that you comply.

Should you decide to stop posting then that would be unfortunate, but you must admit that you have been given unusual latitude in terms of threads started.

Should this thread be extended any further, then I may have to lock it.

I am hoping that you will understand my perspective.

Your move...

Later.

I completely do not understand you, Roger.

If there is a requirement for posting a limited number of threads per day, or per month, or per year, I would gladly comply.
Is there such a requirement, written somewhere?

Or, for posting a limited number of messages?
I see my posted messages number is not particularly high in respect to the leading posters, so what is wrong here?

Or maybe my threads are undesirable, because I post specific topics and develop them with diagrams, etc., rather than just quick comments, emotional expressions, etc.?

With what is this thread, with specific scientific content, worse than a thread with fake announcements, as you have multiple, and yet not locked, or highly dubious polls, also not locked?

If almost all leading and actively developed engines nowadays have benefited at least a tiny little bit, but in some cases more, from the ideas I have posted in the last 2 or 3 years, why should my threads be locked?

Also, I could still post the contents of this thread under the first thread, but I am very much afraid this is technically impossible now, as the thread is already live and kicking with much posted content.

I do not know what to say, maybe someone who reads my threads might support me, but if you have decided to lock it...

Again, the other thread will become ugly, if I post overtime 5 more suggestions, you yourself will not like it, and therefore I opened this new thread.

I can try to post more messages into a single thread starting from this one, as long as this is humanly possible and does not get ugly.

[Lyudmil Tsvetkov]

OK, I am not posting until Roger announces his decision.

I have always been ethical towards people, unfortunately, not always people have been ethical towards me.

I wanted to post one more subthread this evening, but that might wait.

[Roger Brown]

OK, I am not posting until Roger announces his decision.

I have always been ethical towards people, unfortunately, not always people have been ethical towards me.

I wanted to post one more subthread this evening, but that might wait.

Hello Lyudmil,

I already announced my decision.

What I am going to do now is effect it.

Your comment about ethical behaviour is difficult to parse but I am not going to concern myself overmuch about it.

As you insist on misunderstanding what I am saying then let me be clear:

I am not restricting the number of times that you may post, the number of messages, per month or any other basis.

It is ridiculous to take the argument to that extreme. You were asked, quite politely, to append these ideas to another thread, started by you, still at the front of the forum.

You have refused.

That is your decision.

Quite frankly, you seem to think rather more highly of your threads than you ought. I would not know what diagrams, charts etc. you specifically post in them, as I do not read them, save in response to a complaint and even then, not the entire thing.

Your comment about the multitude of other posts with fake announcements (locked/moved) and highly dubious polls (locked/moved) is taken. I assure you that I will assidiously search for these and deal with them as the members point them out.

This thread is now locked.

Later.