Your network blocks the Lichess assets!

lichess.org
Donate

I made a program to find craziest sacrifices in a row. Here is what it found

like when they end with checkmate, because it's hard to understand the point sometimes

like when they end with checkmate, because it's hard to understand the point sometimes

Is it open source? Is it available somewhere?

Is it open source? Is it available somewhere?

@TotalNoob69 said ^

Is it open source? Is it available somewhere?

Hello @TotalNoob69!
I am actually a big fan of your project, using Lichess tools constantly for the neo pieces and other features.
Right now my project is very very rough, and has quite some inconsistencies (still gives brilliants if the piece was trapped sometimes, but pretty rarely, and still gives brilliants when you sacrifice rook for 2 minor pieces (f. e. rook takes knight, bishop takes rook, and we take bishop back next move, the latter one is actually not hard to fix, but I frankly don't have any time, I'm mostly chess player ).
My main goal was creating something close to chess.com game review, because I liked it a lot, and I built working version which I use for myself.
In order to find brilliant move sequences, which I used in this article (please, take note that I only have included best ones, amongst 70, and sometimes it can give "brilliant" to the sacrifice if it was obvious, since a move being beautiful is a very human metric, unedited list is in this study - https(:)//lichess.org/study/QtgOikOD/UQWowq0Y), I've split the job into those parts:

  1. I analysed all games, using main program analyser.
  2. For each move besides checking the loss, also have checked whether it is brilliant, by a brilliance finder engine.
  3. Found whether the second move for us in the sequence is also brilliant and created pgn file which contains games with 2 or more brilliancies in a row.
    I am gonna assume that you are curious about the part that identifies whether the move is brilliant or not, feel free to message me back if you meant something else.
    Right now, it has multiple tough spots besides not being perfect at calling a move brilliant (which I think even chess.com engine has this problem), for example:
    It requires to kill the engine and restart it every move, because extra hash in stockfish leads to more mistakes in identifying brilliance of moves and clearing it via uci commands doesn't seem to help.
    It summons engine around 10 times, at the depth 10, which takes around a second, and lowering depth to 8 is possible, but may lead to less brilliances being found, so it's far from optimal.
    I plan on improving it, however it might take months for me to get back to this project. I might publish the source code or compilated file somewhere (for example, github, however I don't even think I have account there at the moment) in the next week, and will be happy if someone will use it as parts of their own projects.
    If you want to know the full unedited 70 positions list of 2 brilliant moves in a row, I created a study, which you should be able to see by a link (I need to include "()" because it posts it as a picture otherwise ) https(:)//lichess.org/study/QtgOikOD/UQWowq0Y , there are first 64 positions with 2 brilliant moves found unedited, so you could see all the picture.
    I will update you when I publish the code.
    Thank you very much for your attention to my project, hope you liked the positions.
    If you have any questions - feel free to ask, I'm glad to have feedback!
@TotalNoob69 said [^](/forum/redirect/post/y92LU4Mn) > Is it open source? Is it available somewhere? Hello @TotalNoob69! I am actually a big fan of your project, using Lichess tools constantly for the neo pieces and other features. Right now my project is very very rough, and has quite some inconsistencies (still gives brilliants if the piece was trapped sometimes, but pretty rarely, and still gives brilliants when you sacrifice rook for 2 minor pieces (f. e. rook takes knight, bishop takes rook, and we take bishop back next move, the latter one is actually not hard to fix, but I frankly don't have any time, I'm mostly chess player ). My main goal was creating something close to chess.com game review, because I liked it a lot, and I built working version which I use for myself. In order to find brilliant move sequences, which I used in this article (please, take note that I only have included best ones, amongst 70, and sometimes it can give "brilliant" to the sacrifice if it was obvious, since a move being beautiful is a very human metric, unedited list is in this study - https(:)//lichess.org/study/QtgOikOD/UQWowq0Y), I've split the job into those parts: 1. I analysed all games, using main program analyser. 2. For each move besides checking the loss, also have checked whether it is brilliant, by a brilliance finder engine. 3. Found whether the second move for us in the sequence is also brilliant and created pgn file which contains games with 2 or more brilliancies in a row. I am gonna assume that you are curious about the part that identifies whether the move is brilliant or not, feel free to message me back if you meant something else. Right now, it has multiple tough spots besides not being perfect at calling a move brilliant (which I think even chess.com engine has this problem), for example: It requires to kill the engine and restart it every move, because extra hash in stockfish leads to more mistakes in identifying brilliance of moves and clearing it via uci commands doesn't seem to help. It summons engine around 10 times, at the depth 10, which takes around a second, and lowering depth to 8 is possible, but may lead to less brilliances being found, so it's far from optimal. I plan on improving it, however it might take months for me to get back to this project. I might publish the source code or compilated file somewhere (for example, github, however I don't even think I have account there at the moment) in the next week, and will be happy if someone will use it as parts of their own projects. If you want to know the full unedited 70 positions list of 2 brilliant moves in a row, I created a study, which you should be able to see by a link (I need to include "()" because it posts it as a picture otherwise ) https(:)//lichess.org/study/QtgOikOD/UQWowq0Y , there are first 64 positions with 2 brilliant moves found unedited, so you could see all the picture. I will update you when I publish the code. Thank you very much for your attention to my project, hope you liked the positions. If you have any questions - feel free to ask, I'm glad to have feedback!

@Fire-work , cool, In the last one white sacrifice exchange on b6 to protect c4 pawn and and also force black queen to move on b6, exchange on f6 sacrificed for messed up black's pawn structure and open black king, for example black can't move 26...Kg7 because of 27. Nf5 +- !
but black have crazy brilliant move 26...Rxc4!! and white can't 27. Bxc4, because 27...Qb1 28. Bf1 Qxe4 29. Bxh6 Rc8 30. Qxf6 Qg6 31. Qh4 Qg4 and black successfully defended from white attack, but on 26...Rxc4 white have crazy brilliant move 27. h3!! +- and white have Nf5-Qg3 and Bxh6-Qxf6 and many other attacking ideas!

@Fire-work , cool, In the last one white sacrifice exchange on b6 to protect c4 pawn and and also force black queen to move on b6, exchange on f6 sacrificed for messed up black's pawn structure and open black king, for example black can't move 26...Kg7 because of 27. Nf5 +- ! but black have crazy brilliant move 26...Rxc4!! and white can't 27. Bxc4, because 27...Qb1 28. Bf1 Qxe4 29. Bxh6 Rc8 30. Qxf6 Qg6 31. Qh4 Qg4 and black successfully defended from white attack, but on 26...Rxc4 white have crazy brilliant move 27. h3!! +- and white have Nf5-Qg3 and Bxh6-Qxf6 and many other attacking ideas!

@Biggest_Floppa228 said ^

@Fire-work , cool, In the last one white sacrifice exchange on b6 to protect c4 pawn and and also force black queen to move on b6, exchange on f6 sacrificed for messed up black's pawn structure and open black king, for example black can't move 26...Kg7 because of 27. Nf5 +- !
but black have crazy brilliant move 26...Rxc4!! and white can't 27. Bxc4, because 27...Qb1 28. Bf1 Qxe4 29. Bxh6 Rc8 30. Qx...

Thanks for the clarification!
What a way to protect the pawn!! I didn't think people would do deep analysis, and I'm glad you liked those positions to that extend!
I wasn't sure whether I should put bonus positions, but thanks to your comment, now I will more confortable adding more positions there.

@Biggest_Floppa228 said [^](/forum/redirect/post/stZD5Vkp) > @Fire-work , cool, In the last one white sacrifice exchange on b6 to protect c4 pawn and and also force black queen to move on b6, exchange on f6 sacrificed for messed up black's pawn structure and open black king, for example black can't move 26...Kg7 because of 27. Nf5 +- ! > but black have crazy brilliant move 26...Rxc4!! and white can't 27. Bxc4, because 27...Qb1 28. Bf1 Qxe4 29. Bxh6 Rc8 30. Qx... Thanks for the clarification! What a way to protect the pawn!! I didn't think people would do deep analysis, and I'm glad you liked those positions to that extend! I wasn't sure whether I should put bonus positions, but thanks to your comment, now I will more confortable adding more positions there.

Hello Fire-work you got brilliant sacrifice. I know a channel who shares sacrifice games like yours if you’ll like your games to be featured on their channel you should check them at @Chesssacrificelab on YouTube. You’ve got brilliant moves and

Hello Fire-work you got brilliant sacrifice. I know a channel who shares sacrifice games like yours if you’ll like your games to be featured on their channel you should check them at @Chesssacrificelab on YouTube. You’ve got brilliant moves and