Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Tweak pawn connections #2629

Closed

Conversation

locutus2
Copy link
Member

Use less bonus for blocked connected pawns so closed positions are less worth.

STC:
LLR: 2.96 (-2.94,2.94) {-0.50,1.50}
Total: 60004 W: 11904 L: 11619 D: 36481
Ptnml(0-2): 1066, 7083, 13535, 7136, 1182
https://tests.stockfishchess.org/tests/view/5e941a8063d105aebbab23e3

LTC:
LLR: 2.95 (-2.94,2.94) {0.25,1.75}
Total: 36606 W: 4831 L: 4556 D: 27219
Ptnml(0-2): 252, 3353, 10872, 3520, 306
https://tests.stockfishchess.org/tests/view/5e9444b963d105aebbab2427

Bench: 4961260

Use less bonus for blocked connected pawns so closed positions are less worth.

STC:
LLR: 2.96 (-2.94,2.94) {-0.50,1.50}
Total: 60004 W: 11904 L: 11619 D: 36481
Ptnml(0-2): 1066, 7083, 13535, 7136, 1182
https://tests.stockfishchess.org/tests/view/5e941a8063d105aebbab23e3

LTC:
LLR: 2.95 (-2.94,2.94) {0.25,1.75}
Total: 36606 W: 4831 L: 4556 D: 27219
Ptnml(0-2): 252, 3353, 10872, 3520, 306
https://tests.stockfishchess.org/tests/view/5e9444b963d105aebbab2427

Bench: 4961260
@Rocky640
Copy link

For a supported white pawn on e5 against a supported black pawn on e6, (a french structure)
this is a White penalty of S(11,7) compared to current master.

This result is rather unexpected

@locutus2
Copy link
Member Author

@Rocky640
Yes. In general this should disencourage closed positions.

@locutus2
Copy link
Member Author

locutus2 commented Apr 13, 2020

This penalty have a broad effect.

Most important is to avoid the lockdown in the center if it not berings you an advantage,
Perhaps only counting center pawns is better.

Further i recently tried something similiar where only connected pawns which directs not to the center was a penalty given. I should try this now only with blocked pawns.

EDIT:
I have later tried but only with the full penalty.

@31m059
Copy link

31m059 commented Apr 13, 2020

@Rocky640 Also, as @Vizvezdenec mentioned on the test itself, it makes sense to consider this patch as a form of compensation for a recent commit db59696, which increases the space weight with the number of blocked pawns. To my understanding, that should favor White in most French Advance (3. e5) structures. This helps offset that a bit.

@vondele vondele closed this in 5c58f67 Apr 14, 2020
@vondele
Copy link
Member

vondele commented Apr 14, 2020

Thanks!

@xoto10
Copy link
Contributor

xoto10 commented Apr 15, 2020

Regarding the advance French structure, @Lolligerhans made a good comment recently. Although the blocked pawns on d4/5 and e5/6 define the basic structure, perhaps we can gain by making sf play better when given this structure, instead of just trying to avoid it? e.g. I'm not good at chess, but even I can see that having blocked pawns on h5/6 as well as the ones mentioned above looks really bad, see game 28 in the current TCEC superfinal.

I'm not sure what that would mean we should look for, but I've already seen people looking at things like pieces trapped on 1st rank defending another pawn/piece, I've tried looking for >1 attacking pawn, and so on. Perhaps we need some combination of 2 or 3 factors, pawns unable to move, low space / mobility / trapped piece(s) / xxx ?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

5 participants