Discuss Scratch

birdracerthree
Scratcher
500+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

birdracerthree wrote:

Why doesn’t GoK see the checkmate here? White Dove sees it on ply 8.

It sees it at ply 1 in 4 seconds. Which level were you on? But maybe you ran into the ply 1 checkmate search timeout which I had reduced by your input ;-) However it should still find it later, latest at depth 8. Maybe TT reasons after limited ply 1 search.
I ran this on mobile, so the TT is probably not very good due to limited memory (after GoK played its move on Difficult, the tab refreshed automatically).

I was halfway through reviewing your opening book; it has a lot of issues. Do you know where the opening book lines come from?

Last edited by birdracerthree (May 21, 2024 11:14:28)

ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

Why doesn’t GoK see the checkmate here? White Dove sees it on ply 8.

It sees it at ply 1 in 4 seconds. Which level were you on? But maybe you ran into the ply 1 checkmate search timeout which I had reduced by your input ;-) However it should still find it later, latest at depth 8. Maybe TT reasons after limited ply 1 search.
I ran this on mobile, so the TT is probably not very good due to limited memory (after GoK played its move on Difficult, the tab refreshed automatically).

I was halfway through reviewing your opening book; it has a lot of issues. Do you know where the opening book lines come from?

Dev version finds it at depth 1 in 1.7 seconds (and if I restrict check-sequence search a bit in length, it would be 0.2 seconds, but then the mate-in-15 of two weeks ago can't be found any more); when I disable that depth 1 special search, it is found at depth 9 in 6 seconds.

Please discuss opening book topics with Grijfland, he owns it.

Last edited by ArnoHu (May 21, 2024 21:42:18)

birdracerthree
Scratcher
500+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

Why doesn’t GoK see the checkmate here? White Dove sees it on ply 8.

It sees it at ply 1 in 4 seconds. Which level were you on? But maybe you ran into the ply 1 checkmate search timeout which I had reduced by your input ;-) However it should still find it later, latest at depth 8. Maybe TT reasons after limited ply 1 search.
I ran this on mobile, so the TT is probably not very good due to limited memory (after GoK played its move on Difficult, the tab refreshed automatically).

I was halfway through reviewing your opening book; it has a lot of issues. Do you know where the opening book lines come from?

Dev version finds it at depth 1 in 1.7 seconds (and if I restrict check-sequence search a bit in length, it would be 0.2 seconds, but then the mate-in-15 of two weeks ago can't be found any more); when I disable that depth 1 special search, it is found at depth 9 in 6 seconds.

Please discuss opening book topics with Grijfland, he owns it.
I ran the position on a laptop and GoK sees the checkmate on ply 1 at 7 seconds in competition mode. The ply 1 search timed out before it sees the checkmate on Difficult.

I will definitely speak to Grijfland when I finish the review. I have looked over all of the e4 lines so far
ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

Why doesn’t GoK see the checkmate here? White Dove sees it on ply 8.

It sees it at ply 1 in 4 seconds. Which level were you on? But maybe you ran into the ply 1 checkmate search timeout which I had reduced by your input ;-) However it should still find it later, latest at depth 8. Maybe TT reasons after limited ply 1 search.
I ran this on mobile, so the TT is probably not very good due to limited memory (after GoK played its move on Difficult, the tab refreshed automatically).

I was halfway through reviewing your opening book; it has a lot of issues. Do you know where the opening book lines come from?

Dev version finds it at depth 1 in 1.7 seconds (and if I restrict check-sequence search a bit in length, it would be 0.2 seconds, but then the mate-in-15 of two weeks ago can't be found any more); when I disable that depth 1 special search, it is found at depth 9 in 6 seconds.

Please discuss opening book topics with Grijfland, he owns it.
I ran the position on a laptop and GoK sees the checkmate on ply 1 at 7 seconds in competition mode. The ply 1 search timed out before it sees the checkmate on Difficult.

I will definitely speak to Grijfland when I finish the review. I have looked over all of the e4 lines so far

The depth 1 time limit is a fraction of overall think time, that is the reason why you might need comp mode. On my system, no problem on Medium or Difficult
birdracerthree
Scratcher
500+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

Why doesn’t GoK see the checkmate here? White Dove sees it on ply 8.

It sees it at ply 1 in 4 seconds. Which level were you on? But maybe you ran into the ply 1 checkmate search timeout which I had reduced by your input ;-) However it should still find it later, latest at depth 8. Maybe TT reasons after limited ply 1 search.
I ran this on mobile, so the TT is probably not very good due to limited memory (after GoK played its move on Difficult, the tab refreshed automatically).

I was halfway through reviewing your opening book; it has a lot of issues. Do you know where the opening book lines come from?

Dev version finds it at depth 1 in 1.7 seconds (and if I restrict check-sequence search a bit in length, it would be 0.2 seconds, but then the mate-in-15 of two weeks ago can't be found any more); when I disable that depth 1 special search, it is found at depth 9 in 6 seconds.

Please discuss opening book topics with Grijfland, he owns it.
I ran the position on a laptop and GoK sees the checkmate on ply 1 at 7 seconds in competition mode. The ply 1 search timed out before it sees the checkmate on Difficult.

I will definitely speak to Grijfland when I finish the review. I have looked over all of the e4 lines so far

The depth 1 time limit is a fraction of overall think time, that is the reason why you might need comp mode. On my system, no problem on Medium or Difficult
The Dev version is faster (3.266 seconds). What changed in the dev?

Last edited by birdracerthree (May 21, 2024 22:00:15)

ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

Why doesn’t GoK see the checkmate here? White Dove sees it on ply 8.

It sees it at ply 1 in 4 seconds. Which level were you on? But maybe you ran into the ply 1 checkmate search timeout which I had reduced by your input ;-) However it should still find it later, latest at depth 8. Maybe TT reasons after limited ply 1 search.
I ran this on mobile, so the TT is probably not very good due to limited memory (after GoK played its move on Difficult, the tab refreshed automatically).

I was halfway through reviewing your opening book; it has a lot of issues. Do you know where the opening book lines come from?

Dev version finds it at depth 1 in 1.7 seconds (and if I restrict check-sequence search a bit in length, it would be 0.2 seconds, but then the mate-in-15 of two weeks ago can't be found any more); when I disable that depth 1 special search, it is found at depth 9 in 6 seconds.

Please discuss opening book topics with Grijfland, he owns it.
I ran the position on a laptop and GoK sees the checkmate on ply 1 at 7 seconds in competition mode. The ply 1 search timed out before it sees the checkmate on Difficult.

I will definitely speak to Grijfland when I finish the review. I have looked over all of the e4 lines so far

The depth 1 time limit is a fraction of overall think time, that is the reason why you might need comp mode. On my system, no problem on Medium or Difficult
The Dev version is faster (3.266 seconds). What changed in the dev?

Worked on move ordering. It is also in the main version now.
ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

birdracerthree wrote:

Why doesn’t GoK see the checkmate here? White Dove sees it on ply 8.

It sees it at ply 1 in 4 seconds. Which level were you on? But maybe you ran into the ply 1 checkmate search timeout which I had reduced by your input ;-) However it should still find it later, latest at depth 8. Maybe TT reasons after limited ply 1 search.
I ran this on mobile, so the TT is probably not very good due to limited memory (after GoK played its move on Difficult, the tab refreshed automatically).

I was halfway through reviewing your opening book; it has a lot of issues. Do you know where the opening book lines come from?

Dev version finds it at depth 1 in 1.7 seconds (and if I restrict check-sequence search a bit in length, it would be 0.2 seconds, but then the mate-in-15 of two weeks ago can't be found any more); when I disable that depth 1 special search, it is found at depth 9 in 6 seconds.

Please discuss opening book topics with Grijfland, he owns it.
I ran the position on a laptop and GoK sees the checkmate on ply 1 at 7 seconds in competition mode. The ply 1 search timed out before it sees the checkmate on Difficult.

I will definitely speak to Grijfland when I finish the review. I have looked over all of the e4 lines so far

The depth 1 time limit is a fraction of overall think time, that is the reason why you might need comp mode. On my system, no problem on Medium or Difficult
The Dev version is faster (3.266 seconds). What changed in the dev?

Worked on move ordering. It is also in the main version now.

Checkmate in 5 now detected in 1.9 seconds at depth 1 for https://lichess.org/analysis/standard/1r1q2nQ/ppp3p1/4prk1/6N1/1n1P4/8/PP1N1PPP/R5K1_w_-_-_12_23

And this mate in 4 (with a lot of queen space and not continuous quiescence sequence) in 4.1 seconds at depth 6: https://lichess.org/analysis/standard/8/2N4p/8/8/4KP2/8/3k4/7Q_w_-_-_5_59

How are other engines dealing with those?

Last edited by ArnoHu (May 23, 2024 10:43:01)

birdracerthree
Scratcher
500+ posts

Scratch Chess Engine - Game of Kings

GoK opening report

If someone quotes this, please snip it to make sure the reply doesn't take too much space. Feel free to ask questions.

e2e4 c7c5 Ng1f3 d7d6 d2d4 c5d4 Nf3d4 Ng8f6 Nb1c3 Nb8c6 Bf1e2 e7e6 Fine
e2e4 c7c5 Ng1f3 d7d6 d2d4 c5d4 c2c3 d4c3 Nb1c3 Ng8f6 Bf1c4 e7e6 OK
e2e4 c7c5 Ng1f3 e7e6 d2d4 c5d4 Nf3d4 Nb8c6 Nb1c3 a7a6 f2f3!? e6e5?! Inaccurate
e2e4 c7c5 Ng1f3 d7d6 Bf1b5 Bc8d7 a2a4 Nb8c6 O-O e7e6 OK
e2e4 c7c5 Ng1f3 Nb8c6 d2d4 c5d4 Nf3d4 Ng8f6 Nb1c3 d7d6 Perfect
e2e4 c7c5 Ng1f3 Nb8c6 Bf1b5 e7e6 O-O Ng8e7 f1e1 a7a6 Bb5f1 d7d5 d2d3 g7g6 e4d5 e6d5 OK
e2e4 c7c5 c2c3 Ng8f6 e4e5 Nf6d5 d2d4 c5d4 c3d4 d7d6 Perfect
e2e4 c7c5 Nb1c3 Nb8c6 g2g3 g7g6 Bf1g2 Bf8g7 d2d3 d7d6 Perfect
e2e4 c7c5 Nb1c3 Nb8c6 Ng1f3 d7d6 Bf1b5 Bc8d7 O-O Ng8f6 Fine
e2e4 c7c5 Bf1c4 e7e6 Ng1f3 a7a6 O-O b7b5 Bc4e2 Bc8b7 Good
e2e4 c7c5 Bf1c4 e7e6 Nb1c3 a7a6 a2a4 Bf8e7 Ng1f3 Qd8c7 O-O Nb8c6 d2d4 c5d4 Nf3d4 Ng8f6 Nd4c6 b7c6 Awkward
e2e4 c7c5 d2d4 c5d4 Qd1d4 Nb8c6 Qd4d3 Ng8f6 Nb1c3 e7e6 Ng1f3 Qd8c7 Confusing
e2e4 c7c6 d2d4 d7d5 e4e5 Bc8f5 Nb1c3 e7e6 g2g4 Bf5g6 Ng1e2 c6c5 h2h4 h7h5 Ne2f4 Nb8c6?! Inaccurate
e2e4 c7c6 d2d4 d7d5 Nb1c3 d5e4 Nc3e4 Bc8f5 Ne4g3 Bf5g6 Ng1f3 Nb8d7 Perfect
e2e4 c7c6 d2d4 d7d5 e4e5 c6c5 d4c5 e7e6 Bc1e3 Nb8d7 Perfect
e2e4 e7e5 Qd1h5 Nb8c6 Bf1c4 g7g6 Qh5f3 Ng8f6 Ng1e2 Bf8g7 d2d3 O-O
e2e4 e7e5 Qd1f3 Ng8f6 Bf1c4 Nb8c6 Ng1e2 d7d6 h2h3 Bc8e6 Bc4e6 f7e6
e2e4 e7e5 Bf1c4 Ng8f6 d2d3 Nb8c6 Ng1f3 Bf8c5 Nb1c3 d7d6 Bc1g5 h7h6 Perfect
e2e4 e7e5 Bf1c4 Nb8c6 d2d3 Ng8f6 Bc1g5!? Bf8e7 Ng1f3 O-O O-O h7h6 Bg5h4 g7g5 Bh4g3 Nf6h5?? BLUNDER
e2e4 e7e5 Nb1c3 Ng8f6 f2f4 d7d5 f4e5 Nf6e4 Ng1f3 Bf8e7 Qd1e2 Ne4c3 Great Note : Play should continue with dxc3
e2e4 e7e5 Nb1c3 Ng8f6 Bf1c4 Nf6e4 Qd1h5 Ne4d6 Bc4b3 Nb8c6 Perfect
e2e4 e7e5 Ng1f3 Nb8c6 Bf1b5 a7a6 Bb5a4 Ng8f6 O-O Bf8e7 f1e1 O-O?? BLUNDER
e2e4 e7e5 Ng1f3 Nb8c6 Bf1b5 a7a6 Bb5a4 Ng8f6 Qd1e2 b7b5 Ba4b3 Bf8e7 Great
e2e4 e7e5 Ng1f3 Nb8c6 Bf1b5 a7a6 Bb5a4 Ng8f6 d2d4 e5d4 O-O Bf8e7 Interesting
e2e4 e7e5 Ng1f3 Nb8c6 Bf1b5 a7a6 Bb5a4 Ng8f6 Nb1c3 Bf8e7 O-O b7b5 Perfect Note : Play should continue with Bb3
e2e4 e7e5 Ng1f3 Nb8c6 Bf1b5 a7a6 Bb5c6 d7c6 Nf3e5?! Qd8d4 Inaccurate Note : This is designed for the black side only. The white side should never go into this.
e2e4 e7e5 Ng1f3 Nb8c6 Bf1b5 a7a6 Bb5c6 d7c6 O-O Bc8g4 h2h3 Bg4h5 g2g4 Bh5g6 Interesting Note : Black loses a pawn, but white’s king is weaker
e2e4 e7e5 Ng1f3 Nb8c6 Bf1b5 a7a6 Bb5c6 d7c6 Nb1c3 f7f6 d2d4 e5d4 Perfect
e2e4 e7e5 Ng1f3 Nb8c6 Bf1b5 a7a6 Bb5c6 d7c6 d2d4 e5d4 Qd1d4 Qd8d4 Near Perfect Note : There is no reason why the book doesn’t capture the queen here
e2e4 e7e5 Ng1f3 Nb8c6 d2d4 e5d4 Nf3d4 Bf8c5 Bc1e3 Qd8f6 Perfect Note : play should continue with c3
e2e4 e7e5 Ng1f3 Nb8c6 d2d4 e5d4 Bf1c4 Bf8c5 c2c3 d4d3!? b2b4 Bc5b6 a2a4 a7a6 OK Note : Why d4d3!? over Ng8f6
e2e4 e7e5 Ng1f3 Nb8c6 Bf1c4 Ng8f6 Nf3g5 d7d5 e4d5 b7b5!? Bc4b5 Qd8d5 Shaky
e2e4 e7e5 Ng1f3 Nb8c6 Bf1c4 Ng8f6 Nf3g5 d7d5 e4d5 Nc6a5 Bc4b5 c7c6 d5c6 b7c6 Perfect
e2e4 e7e5 Ng1f3 Nb8c6 Nb1c3 Ng8f6 Bf1b5 Bf8b4 O-O O-O Perfect
e2e4 e7e5 f2f4 e5f4 Ng1f3 g7g5 Bf1c4?! g5g4 Nf3e5 Qd8h4 Ke1f1 Nb8c6 Very Shaky
e2e4 e7e5 f2f4 e5f4 Ng1f3 g7g5 Bf1c4?! g5g4 O-O!! g4f3 Qd1f3 Bf8c5?? g1h1 Qd8h4?? d2d4 Bc5d4 Bc1f4 f7f6 Nb1c3 d7d6 MULTIPLE BLUNDERS
e2e4 e7e5 f2f4 e5f4 Ng1f3 g7g5 h2h4 g5g4 Nf3e5 Ng8f6 d2d4 d7d6 Ne5d3 Nf6e4 Bc1f4 Bf8g7 Fine Note : King’s Gambit lines are disadvantageous to the White side
e2e4 e7e5 f2f4 e5f4 Bf1c4 Qd8h4 Ke1f1 Bf8c5? d2d4 Bc5b6 Ng1f3 Qh4h5 Bc1f4 d7d6 Nb1c3 Bc8g4 Mistake
e2e4 e7e5 f2f4 e5f4 Ng1f3 Bf8e7 Bf1c4 Ng8f6 Nb1c3?! O-O?! O-O?! Nb8c6?! d2d4?! a7a6? FILLED WITH MISTAKES
e2e4 e7e5 f2f4 e5f4 Bf1c4 Ng8f6 Nb1c3 c7c6 e4e5?! d7d5 e5f6 d5c4 f6g7 Bf8g7 Qd1e2 Bc8e6 MISTAKES, -2 after the book ends
e2e4 e7e5 f2f4 e5f4 d2d4 Qd8h4 Ke1e2 d7d6 Ng1f3 Bc8g4 Bc1f4 Bg4f3 Ke2f3 Ng8f6 Bf1d3 Nb8c6 FILLED WITH MISTAKES
e2e4 e7e5 f2f4 e5f4 Bf1c4 d7d5 Bc4d5 Ng8f6 Nb1c3 Bf8b4 Ng1f3 Bb4c3 d2c3 c7c6 Bd5b3 Very Inaccurate Note : Where are these opening lines coming from?
e2e4 e7e5 f2f4 Bf8c5 Ng1f3 d7d6 Bf1c4 Ng8f6 f4e5 d6e5 Nf3e5?? Qd8d4 Bc4f7 Ke8f8 Ne5d3 Qd4e4 BLUNDER
e2e4 e7e5 f2f4 Nb8c6?! Ng1f3 d7d6 Bf1c4 Bc8g4 O-O Nc6d4? f4e5 Nd4f3 g2f3? Bg4h3 Mistakes Note : after Nd4, there is 6. Bxf7!! After Nxf3, there is the superior move Rxf3!!
e2e4 Nb8c6 d2d4 d7d5 e4d5 Qd8d5 Ng1f3 Bc8g4 Bf1e2 O-O-O c2c4 Qd5h5 Near Perfect
e2e4 Nb8c6 Ng1f3 e7e5 Bf1b5 a7a6 Bb5a4 b7b5 Ba4b3 Ng8f6 O-O Nf6e4 f1e1 Ne4c5 Inaccurate
e2e4 Ng8f6 e4e5 Nf6d5 d2d4 d7d6 Ng1f3 Bc8g4 Bf1e2 c7c6 Perfect
e2e4 g7g6 d2d4 Bf8g7 Bf1c4 d7d6 Ng1f3 Ng8f6 Qd1e2 O-O O-O Bc8g4 Perfect
e2e4 g7g6 d2d4 Bf8g7 Ng1f3 d7d6 Bf1c4 Ng8f6 Qd1e2 O-O O-O Bc8g4 e4e5 d6e5 d4e5 Nf6d5 Perfect
e2e4 g7g6 d2d4 Bf8g7 Nb1c3?! Ng8f6 Ng1f3 O-O?! e4e5 Nf6e8 Bc1f4 d7d6 h2h3 Nb8d7 Inaccurate (don’t allow e4e5)
e2e4 e7e6 d2d4 d7d5 e4e5 c7c5 c2c3 Nb8c6 f2f4?! c5d4 c3d4 Qd8b6 Inaccurate Note : I assumed GoK played f2f4 itself. Why is this in the book?
e2e4 e7e6 d2d4 d7d5 e4e5 c7c5 c2c3 Nb8c6 Ng1f3 Qd8b6 Bf1e2 c5d4 c3d4 Ng8e7 Perfect
e2e4 e7e6 d2d4 d7d5 Nb1c3 Bf8b4 Qd1d3?!? d5e4 Qd3e4 Ng8f6 Qe4h4 Qd8d5 Inaccurate
e2e4 e7e6 d2d4 d7d5 Nb1c3 Bf8b4 e4e5 c7c5 a2a3 Bb4c3 b2c3 Ng8e7 Qd1g4 O-O Needs to be expanded Note : This opening can be used to defeat GoK in less than 30 moves. The line should continue 8. Nf3 Nbc6 9. Bd3 f5 10. exf6 Rxf6
e2e4 e7e6 d2d4 d7d5 Nb1c3 Bf8b4 e4d5 e6d5 Bc1d2 Ng8f6 a2a3 Bb4d6 Nc3b5 Bd6e7 Fine
e2e4 e7e6 Ng1f3 d7d5 e4d5 e6d5 d2d4 Ng8f6 Nb1c3 c7c6 Bf1d3 Bf8d6 Bc1g5 O-O Fine
e2e4 d7d6 d2d4 Ng8f6 Nb1c3 Nb8d7 Ng1f3 e7e5 g2g4!?! Nf6g4 Rh1g1 Ng4f6 Bf1c4 h7h6 Bc1e3 c7c6 Interesting (Shirov Gambit)
e2e4 d7d6 d2d4 Ng8f6 Nb1c3 Nb8d7 Ng1f3 e7e5 Bf1c4 h7h6 O-O c7c6 a2a4 Qd8c7 Qd1e2 Bf8e7 Perfect
e2e4 d7d6 d2d4 Ng8f6 Nb1c3 Nb8d7 Ng1f3 e7e5 Bf1e2 Bf8e7 O-O O-O f1e1 c7c6 a2a4 a7a5 Perfect
e2e4 d7d6 d2d4 Ng8f6 Nb1c3 g7g6 Bc1e3 Bf8g7 f2f3 O-O Qd1d2 e7e5 d4d5 c7c6 O-O-O c6d5 Perfect
e2e4 d7d6 d2d4 Ng8f6 Nb1c3 g7g6 Bf1c4 Bf8g7 Ng1f3 Nb8c6?! h2h3 O-O Bc1e3 Nf6e4 Bc4f7?! f8f7 Nc3e4 d6d5 Inaccurate
e2e4 d7d5 e4d5 Qd8d5 Nb1c3 Qd5a5 d2d4 c7c6 Ng1f3 Ng8f6 Perfect
d2d4 d7d5 Ng1f3 Ng8f6 Bc1g5 Nf6e4 Bg5h4 c7c5 d4c5 Nb8c6 Nb1d2 Ne4c5 c2c3 g7g6 Great
d2d4 d7d5 Ng1f3 Ng8f6 e2e3 c7c5 c2c4 c5d4 e3d4 g7g6 Nb1c3 Bf8g7 Bf1e2 O-O O-O Nb8c6 Perfect
d2d4 d7d5 Ng1f3 Nb8c6 Nb1c3 e7e6 e2e4 Ng8f6 e4e5 Nf6d7 Bc1g5 Bf8e7 h2h4 Nd7b6 Qd1d2 h7h6 Very Inaccurate Note : Please stop putting the knight in front of the c-pawn
d2d4 d7d5 Ng1f3 Nb8c6 Bc1f4 Bc8g4 e2e3 e7e6 c2c4 Bf8b4 Nb1c3 Ng8e7 h2h3 Bg4h5 Inaccurate
d2d4 d7d5 Ng1f3 Ng8f6 c2c4 c7c6 Nb1c3 e7e6 Bc1g5 Bf8e7 e2e3 O-O Perfect
d2d4 d7d5 Ng1f3 e7e6 Nb1c3 Ng8f6 e2e3 Nb8c6 Bf1d3 a7a6 O-O Bf8b4 a2a3 Bb4d6 Inaccurate
d2d4 d7d5 Ng1f3 e7e6 Nb1c3 Ng8f6 Bc1g5 Bf8e7 Bg5f6 Be7f6 e2e4 d5e4 Nc3e4 O-O
d2d4 d7d5 c2c4 e7e6 Nb1c3 Ng8f6 e2e3 a7a6 Bf1d3 d5c4 Bd3c4 Bf8b4 Awkward
d2d4 d7d5 c2c4 e7e6 Nb1c3 Ng8f6 Bc1g5 Nb8d7 e2e3 Bf8b4 Fine
d2d4 d7d5 e2e3 Ng8f6 Bf1d3 e7e6 Ng1f3 c7c5 c2c3 Nb8c6 Nb1d2 Bf8e7 O-O O-O Perfect
d2d4 d7d5 c2c4 c7c6 Nb1c3 Ng8f6 Ng1f3 e7e6 Bc1g5 Bf8e7 e2e3 O-O Bf1d3 Nb8d7 Perfect
d2d4 d7d5 c2c4 c7c6 Ng1f3 Ng8f6 Nb1c3 e7e6 Bc1g5 Bf8e7 e2e3 O-O Bf1d3 Nb8d7 Perfect
d2d4 d7d5 Bc1f4 Ng8f6 e2e3 e7e6 Ng1f3 Bf8d6 Nf3e5 Nb8d7?! Nb1d2 O-O c2c3 c7c5? Inaccuracies
d2d4 Ng8f6 c2c4 g7g6 Nb1c3 Bf8g7 Ng1f3 O-O e2e4 d7d6 Bf1e2 e7e5 d4d5 a7a5 Perfect
d2d4 Ng8f6 c2c4 g7g6 Ng1f3 Bf8g7 g2g3 O-O Nb1c3 d7d5 c4d5 Nf6d5 Perfect Note : Line should continue with Bf1g2
d2d4 Ng8f6 c2c4 g7g6 Ng1f3 Bf8g7 Nb1c3 O-O e2e4 d7d6 Bf1e2 e7e5 O-O Nb8c6 Perfect
d2d4 Ng8f6 c2c4 e7e6 Nb1c3 Bf8b4 e2e3 O-O Bf1d3 d7d5 Ng1f3 c7c5 O-O Nb8c6 Perfect
d2d4 Ng8f6 c2c4 d7d6 Ng1f3 Bc8g4 e2e3 Nb8d7 h2h3 Bg4h5 Nb1c3 e7e5 g2g4 Bh5g6 Bf1g2 c7c6?! Awkward
d2d4 Ng8f6 c2c4 g7g6 Nb1c3 Bf8g7 e2e4 d7d6 Ng1f3 O-O Bf1e2 e7e5 Perfect
d2d4 Ng8f6 Ng1f3 e7e6 c2c4 b7b6 Nb1c3 Bc8b7 e2e3 Bf8b4 Bf1e2 O-O Qd1c2 Nb8c6 Weird
d2d4 Ng8f6 Ng1f3 d7d5 c2c4 e7e6 Nb1c3 c7c6 e2e3 Nb8d7 Qd1c2 Bf8d6
d2d4 Ng8f6 c2c4 e7e6 Ng1f3 d7d5 Nb1c3 Bf8b4 Qd1a4 Nb8c6 e2e3 O-O Perfect
d2d4 Ng8f6 c2c4 e7e6 Ng1f3 b7b6 g2g3 Bc8b7 Bf1g2 Bf8e7 O-O O-O Nb1c3 Nf6e4 Perfect
d2d4 f7f5 c2c4 Ng8f6 Nb1c3 e7e6 e2e3 Bf8b4 Bf1e2 O-O Ng1f3?! Nf6e4 Weird
d2d4 f7f5 g2g3 Ng8f6 Bf1g2 g7g6 Ng1f3 Bf8g7 O-O O-O c2c4 d7d6 Nb1c3 Qd8e8 Perfect
d2d4 f7f5 Ng1f3 Ng8f6 g2g3 e7e6 c2c4 d7d5 Bf1g2 c7c6 O-O Bf8d6 b2b3 Qd8e7 Bc1b2 O-O Perfect
d2d4 e7e6 c2c4 Ng8f6 Ng1f3 b7b6 g2g3 Bc8b7 Bf1g2 Bf8e7 O-O O-O Nb1c3 Nf6e4 Perfect
d2d4 e7e6 c2c4 Ng8f6 Nb1c3 Bf8b4 Qd1c2 O-O Ng1f3 b7b6 Bc1g5 Bc8b7 O-O-O Bb4c3 Inaccurate
d2d4 e7e6 c2c4 d7d5 Nb1c3 Ng8f6 Ng1f3 c7c6 e2e3 Nb8d7 Qd1c2 Bf8d6 Bf1d3 O-O Perfect Note : Play should continue with O-O
a2a3 e7e5 e2e3 d7d5 c2c4 Ng8f6 b2b4 Bc8e6 c4c5 c7c6
a2a4 e7e5 e2e4 Ng8f6 d2d3 d7d5 e4d5 Nf6d5 g2g3 Bf8d6
a2a4 e7e5 a4a5 d7d5 h2h4 c7c5 h4h5 Nb8c6 Ra1a4 Ng8f6 Ra4h4 Qd8a5 Why
b2b3 e7e5 Bc1b2 Nb8c6 c2c4 Ng8f6 e2e3 d7d5 c4d5 Nf6d5 a2a3 Bf8d6 Fine
b2b4 e7e5 Bc1b2 d7d6 c2c4 Ng8f6 Nb1c3 Bf8e7 e2e3 O-O Ng1f3 Bc8g4 Fine
Nb1c3 d7d5 e2e4 d5d4 Nc3e2 e7e5 Fine
c2c3 e7e5 e2e4 Ng8f6 d2d3 Nb8c6 Qd1f3?! d7d6 h2h3 Bc8e6 Bf1e2 Qd8d7 Weird
c2c4 e7e5 Nb1c3 Ng8f6 g2g3 d7d5 c4d5 Nf6d5 Bf1g2 Nd5c3 b2c3 c7c6 Great
c2c4 e7e5 Ng1f3 Nb8c6 Nb1c3 Ng8f6 g2g3 Bf8c5 Bf1g2 O-O O-O d7d6 d2d3 Bc8f5?! Weird
c2c4 e7e5 g2g3 d7d5 c4d5 Ng8f6 Bf1g2 Nf6d5 d2d3 c7c6 Perfect
c2c4 e7e5 e2e4 Ng8f6 Nb1c3 Bf8c5 d2d3 d7d6 Bf1e2 Nb8c6 Weird for White, Perfect for Black
c2c4 c7c5 Nb1c3 Nb8c6 g2g3 g7g6 Bf1g2 Bf8g7 Ng1f3 Ng8f6 O-O O-O d2d4 c5d4 Nf3d4 Nc6d4 Qd1d4 d7d6 Perfect
d2d3 d7d5 g2g3 e7e5 Bf1g2 Ng8f6 Perfect
e2e3 e7e5 d2d4 e5d4 e3d4 d7d5 Ng1f3 Ng8f6 Perfect
f2f3 e7e5 e2e4 Bf8c5 Nb1c3 d7d6 Nc3d5 Nb8c6
f2f3 e7e5 e2e4 Ng8f6 d2d3 Bf8c5 g2g4 h7h6
g2g4 e7e5 Bf1g2 d7d5 c2c4 d5c4 Qd1a4 c7c6
g2g3 d7d5 Bf1g2 e7e5 c2c4 d5d4 d2d3 c7c5 f2f4 Bf8d6 f4e5 Bd6e5
h2h3 e7e5 e2e4 Ng8f6 d2d3 d7d5 Bc1g5 d5e4 Bg5f6 Qd8f6 d3e4 Bf8c5 Ng1f3 Qf6b6
h2h3 e7e5 a2a3 d7d5 e2e3 Ng8f6 b2b4 c7c5 b4b5 a7a6 a3a4 Bc8d7 Bc1b2 Bf8d6
h2h4 e7e5 e2e4 Ng8f6 d2d3 Bf8c5 Bf1e2 d7d6 c2c3 Nb8d7 a2a4 a7a5
h2h4 Ng8f6 c2c4 c7c5 Nb1c3 Nb8c6 e2e3 e7e6 d2d4 a7a6
h2h4 Ng8f6 Ng1f3 Nb8c6 h4h5 d7d5 d2d4 e7e6
Ng1f3 Ng8f6 c2c4 g7g6 g2g3 Bf8g7 Bf1g2 O-O O-O d7d6 Perfect
Ng1f3 Ng8f6 c2c4 e7e6 d2d4 d7d5 Nb1c3 Bf8e7 Bc1f4 O-O Perfect
Ng1f3 d7d5 d2d4 Ng8f6 c2c4 e7e6 Nb1c3 Bf8e7 Bc1g5 O-O e2e3 h7h6 Perfect Note : Just make sure GoK doesn’t get fishing poled here on Scratch runtime. Same for the Slav.
Ng1f3 d7d5 Nb1c3 e7e6 d2d4 Ng8f6 Bc1g5 Bf8e7 e2e3 O-O Bf1d3 h7h6
Ng1f3 d7d5 Nf3g1 e7e5 Ng1f3 e5e4 Nf3g1 Ng8f6 Nb1c3 d5d4 Nc3b1 Nb8c6 d2d3 e4d3 c2d3 Bf8d6 Why

Last edited by birdracerthree (May 24, 2024 05:52:30)

ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

Crazy game by 10sec GoK (black) vs. White Dove (P2). I guess no side deserved to win, so draw is OK. Out of GoK's 4 biggest mistake, 2 won't happen on Medium (15sec), and 1 won't happen in competition mode. https://lichess.org/qzlUxLON#104

Rematch, GoK (white) winning at 96% vs. 92% accuracy: https://lichess.org/EcJRx0Ha#133

I had to reduce some of recent fancy 15-ply-plus check/evasion-sequence checkmate puzzle solving, to regain speed and search depth, and it immediately paid off. It might be a side effect of latest chances that GoK left the shortest checkmate path once.

Last edited by ArnoHu (May 24, 2024 18:21:01)

AZURUS41
Scratcher
52 posts

Scratch Chess Engine - Game of Kings

“Note : Please stop putting the knight in front of the c-pawn”
Even I said it 1 year ago lmao
WhiteDoveOfficial
Scratcher
7 posts

Scratch Chess Engine - Game of Kings

birdracerthree wrote:

GoK opening report

If someone quotes this, please snip it to make sure the reply doesn't take too much space. Feel free to ask questions.
*snip*

Great report! Maybe I should add some more lines for WD that are not just for the French…
ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

birdracerthree wrote:

GoK opening report

If someone quotes this, please snip it to make sure the reply doesn't take too much space. Feel free to ask questions.

e2e4 c7c5 Ng1f3 d7d6 d2d4 c5d4 Nf3d4 Ng8f6 Nb1c3 Nb8c6 Bf1e2 e7e6 Fine
e2e4 c7c5 Ng1f3 d7d6 d2d4 c5d4 c2c3 d4c3 Nb1c3 Ng8f6 Bf1c4 e7e6 OK
e2e4 c7c5 Ng1f3 e7e6 d2d4 c5d4 Nf3d4 Nb8c6 Nb1c3 a7a6 f2f3!? e6e5?! Inaccurate
e2e4 c7c5 Ng1f3 d7d6 Bf1b5 Bc8d7 a2a4 Nb8c6 O-O e7e6 OK

// snip


Thanks, that must have been a lot of effort. I am sure @Grijfland will take a look at it.
ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

Crazy game by 10sec GoK (black) vs. White Dove (P2). I guess no side deserved to win, so draw is OK. Out of GoK's 4 biggest mistake, 2 won't happen on Medium (15sec), and 1 won't happen in competition mode. https://lichess.org/qzlUxLON#104

Rematch, GoK (white) winning at 96% vs. 92% accuracy: https://lichess.org/EcJRx0Ha#133

I had to reduce some of recent fancy 15-ply-plus check/evasion-sequence checkmate puzzle solving, to regain speed and search depth, and it immediately paid off. It might be a side effect of latest chances that GoK left the shortest checkmate path once.

GoK (Medium, black) vs. White Dove (P3), chapter 47 of the series “games that White Dove should have won but did not”: https://lichess.org/0ByfY90t#162 Admitted, WD had longer think time here.

Last edited by ArnoHu (May 24, 2024 19:18:38)

birdracerthree
Scratcher
500+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

ArnoHu wrote:

Crazy game by 10sec GoK (black) vs. White Dove (P2). I guess no side deserved to win, so draw is OK. Out of GoK's 4 biggest mistake, 2 won't happen on Medium (15sec), and 1 won't happen in competition mode. https://lichess.org/qzlUxLON#104

Rematch, GoK (white) winning at 96% vs. 92% accuracy: https://lichess.org/EcJRx0Ha#133

I had to reduce some of recent fancy 15-ply-plus check/evasion-sequence checkmate puzzle solving, to regain speed and search depth, and it immediately paid off. It might be a side effect of latest chances that GoK left the shortest checkmate path once.

GoK (Medium, black) vs. White Dove (P3), chapter 47 of the series “games that White Dove should have won but did not”: https://lichess.org/0ByfY90t#162 Admitted, WD had longer think time here.
What happened here? I think the check extensions are cutting into GoK’s think time.

White Dove is doing very well against GoK lately. White Dove never finishes its winning endgames.

Just so you know, Element v1.4891 in the Dev has knight underpromotion, so you can use/test that if you like
ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

ArnoHu wrote:

Crazy game by 10sec GoK (black) vs. White Dove (P2). I guess no side deserved to win, so draw is OK. Out of GoK's 4 biggest mistake, 2 won't happen on Medium (15sec), and 1 won't happen in competition mode. https://lichess.org/qzlUxLON#104

Rematch, GoK (white) winning at 96% vs. 92% accuracy: https://lichess.org/EcJRx0Ha#133

I had to reduce some of recent fancy 15-ply-plus check/evasion-sequence checkmate puzzle solving, to regain speed and search depth, and it immediately paid off. It might be a side effect of latest chances that GoK left the shortest checkmate path once.

GoK (Medium, black) vs. White Dove (P3), chapter 47 of the series “games that White Dove should have won but did not”: https://lichess.org/0ByfY90t#162 Admitted, WD had longer think time here.

Game #4 was a clear win for GoK (white): https://lichess.org/aEAekwv9#85

Series result:
GoK        3.0
White Dove 1.0
ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

birdracerthree wrote:

ArnoHu wrote:

ArnoHu wrote:

Crazy game by 10sec GoK (black) vs. White Dove (P2). I guess no side deserved to win, so draw is OK. Out of GoK's 4 biggest mistake, 2 won't happen on Medium (15sec), and 1 won't happen in competition mode. https://lichess.org/qzlUxLON#104

Rematch, GoK (white) winning at 96% vs. 92% accuracy: https://lichess.org/EcJRx0Ha#133

I had to reduce some of recent fancy 15-ply-plus check/evasion-sequence checkmate puzzle solving, to regain speed and search depth, and it immediately paid off. It might be a side effect of latest chances that GoK left the shortest checkmate path once.

GoK (Medium, black) vs. White Dove (P3), chapter 47 of the series “games that White Dove should have won but did not”: https://lichess.org/0ByfY90t#162 Admitted, WD had longer think time here.
What happened here? I think the check extensions are cutting into GoK’s think time.

White Dove is doing very well against GoK lately. White Dove never finishes its winning endgames.

Just so you know, Element v1.4891 in the Dev has knight underpromotion, so you can use/test that if you like

I just reduced check extensions, think time / search depth is OK. Blunders' effect was far beyond search horizon, some games simply develop negatively…

Yes, I read about underpromotion, but I don't have special test scenarios, and it is not very likely to happen during normal games.

Last edited by ArnoHu (May 24, 2024 20:28:06)

birdracerthree
Scratcher
500+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

birdracerthree wrote:

ArnoHu wrote:

ArnoHu wrote:

Crazy game by 10sec GoK (black) vs. White Dove (P2). I guess no side deserved to win, so draw is OK. Out of GoK's 4 biggest mistake, 2 won't happen on Medium (15sec), and 1 won't happen in competition mode. https://lichess.org/qzlUxLON#104

Rematch, GoK (white) winning at 96% vs. 92% accuracy: https://lichess.org/EcJRx0Ha#133

I had to reduce some of recent fancy 15-ply-plus check/evasion-sequence checkmate puzzle solving, to regain speed and search depth, and it immediately paid off. It might be a side effect of latest chances that GoK left the shortest checkmate path once.

GoK (Medium, black) vs. White Dove (P3), chapter 47 of the series “games that White Dove should have won but did not”: https://lichess.org/0ByfY90t#162 Admitted, WD had longer think time here.
What happened here? I think the check extensions are cutting into GoK’s think time.

White Dove is doing very well against GoK lately. White Dove never finishes its winning endgames.

Just so you know, Element v1.4891 in the Dev has knight underpromotion, so you can use/test that if you like

I just reduced check extensions, think time / search depth is OK. Blunders' effect was far beyond search horizon, some games simply develop negatively…

Yes, I read about underpromotion, but I don't have special test scenarios, and it is not very likely to happen during normal games.
I have a few positions in the study, including this one from a GoK-WD game (WD blundered because it doesn’t have underpromotion). There was an M3 position that required a knight underpromotion, but I have lost the FEN.
Grijfland
Scratcher
100+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

birdracerthree wrote:

GoK opening report

If someone quotes this, please snip it to make sure the reply doesn't take too much space. Feel free to ask questions.

e2e4 c7c5 Ng1f3 d7d6 d2d4 c5d4 Nf3d4 Ng8f6 Nb1c3 Nb8c6 Bf1e2 e7e6 Fine
e2e4 c7c5 Ng1f3 d7d6 d2d4 c5d4 c2c3 d4c3 Nb1c3 Ng8f6 Bf1c4 e7e6 OK
e2e4 c7c5 Ng1f3 e7e6 d2d4 c5d4 Nf3d4 Nb8c6 Nb1c3 a7a6 f2f3!? e6e5?! Inaccurate
e2e4 c7c5 Ng1f3 d7d6 Bf1b5 Bc8d7 a2a4 Nb8c6 O-O e7e6 OK

// snip


Thanks, that must have been a lot of effort. I am sure @Grijfland will take a look at it.

Absolutely, it was always planned, but it never happened.
Now just have to find some time for it.
ArnoHu
Scratcher
1000+ posts

Scratch Chess Engine - Game of Kings

GoK (Medium, white) vs. Element (6+8), GoK wins at 96% vs. 91% accuracy; https://lichess.org/3KkoTvH8#131
birdracerthree
Scratcher
500+ posts

Scratch Chess Engine - Game of Kings

ArnoHu wrote:

GoK (Medium, white) vs. Element (6+8), GoK wins at 96% vs. 91% accuracy; https://lichess.org/3KkoTvH8#131
I was not able to replicate a5 (Nd4) on Stable or Dev versions of Element. I should add the #OfMoves into the FEN, but the move is the same after manually changing it. This could be due to some sort of FEN inconsistency (they are common in Element).

Last edited by birdracerthree (May 25, 2024 01:37:23)

Powered by DjangoBB