Alright, the bug has finally been fixed in the dev. Any thoughts on the performance of the dev version of Element compared to the stable version (for depth 7)?

I proceeded to run 2 games against White Dove. The first game was a draw, but the second had a game-losing blunder that the stable version sees and prevents here - https://lichess.org/study/rUsfHf3a/yS33nWmg#45 . How do I deal with LMR getting in the way of tactics like this?

birdracerthree wrote:

I proceeded to run 2 games against White Dove. The first game was a draw, but the second had a game-losing blunder that the stable version sees and prevents here - https://lichess.org/study/rUsfHf3a/yS33nWmg#45 . How do I deal with LMR getting in the way of tactics like this?

check extensions, single move extenions, singular extensions, not reducing tactically suspect moves, adding more search time (only really works if you have TT and ID), making futility pruning more strict, checks in q-search, etc.

The Scratch 960 2025 has started!
https://lichess.org/broadcast/scratch-960-scratch-engine-tournament-2025/round-1/xQQKfqis#boards

Game of Kings Chess NNUE wins against Stockfish 16 level 8

GoK NNUE for the first time wins against SF16 L8 at only 25 seconds think time, and 97% accuracy, using the fast TurboWarp runtime:

https://lichess.org/RkxrDxD5#219

An Announcement
Hello, I have an announcement to make:

As you guys may be able to tell, Element has been behind its old rival of White Dove and its competitors like GoK Classic and Delta. However, I am working on a new chess engine that will change all of that.

I don’t really like Element’s code: It’s clunky, some of its code is awkward, and of course all of the floating blocks in its codebase. Of course, I am very attached to Element, but so many bad habits have been put into Element. It’s time to make the switch.

However, I am not discontinuing Element’s development. Element has its weak points, and some of them can be handled. Unfortunately, because my laptop blue screened, I was left using the “emergency” backup device I thought I would never use again. That device being a 9 year old Chromebook. I took it out, charged it, and attempted to open Element in the editor. It took 2 minutes, but it worked. I was elated. I could finally update Element. So, where have I been all of this time. Simple; the Chromebook is incredibly slow. I managed to get 7 NPS on Element on S3 depth 1. I also edit my code on S3 (the only exception was the TT, but the components were imported from TW, then implemented in S3). Despite this, I didn’t update Element a lot. It was significantly harder to edit my code on my Chromebook. However, it was enough for my next update at the time; Element v1.4.10.3 to v1.4.11. This time I simply imported the dev to the stable version. 1.4.11 prevented Element from blundering M3 by adding checkmate detection on the final ply. At the time, this was crucial, and I was extremely happy to get it finished. Much later, I discovered that my refutation moves did in fact speed of the engine during the middlegame. So, after very few block changes, 1.4.12 was out. 1.5 came later and that was a miracle. I never thought I would get it out because I was working on it in TW, so the file was on the laptop I could not access. However, I managed to find the sprite that I imported from TW to release from 1.5. To my absolute shock, it worked. This…should not have been possible. The previous TT dev had a bug where it would hang checkmate, and I could never ever figure out why it was doing that. But now, all of a sudden, the code is working perfectly and there are no bugs. I could not believe it. How did a bug that has existed for months just…disappear? Believe it or not, I managed to come up with an explanation for what happened. Long story short, the original TT dev was so old, it was before the king captures update (v1.485) and that caused complications.



So, what happened? I struggled to really gain interest in coding because of the slow device and editor. I had taken a 6 month long break from anything major for Element, but I had to take another break. After all, I was happy with my new and working transposition table. I was happy I could make changes at all with a 9 year old device.



After looking at many games from GoK, White Dove, and the rest of the engines, as well as my knowledge of the engines led me to one conclusion. These engines cannot handle aggressive play (except for Delta) and GoK NNUE). Because of this, I plan to create a new chess engine that plays extremely aggressively. However, Element will not be discontinued and will still get updates. I finally have a new device now, and I am ready to take chess engines on Scratch to a level of complexity, knowledge, and sheer strength to another level. This new engine will be stronger and faster than Element, and it will play very aggressively. I believe I have the key to unlock the next level, and all I have to do is find the correct door. It will likely take me months to find it, but I will find that door.

birdracerthree wrote:

An Announcement
Hello, I have an announcement to make:

As you guys may be able to tell, Element has been behind its old rival of White Dove and its competitors like GoK Classic and Delta. However, I am working on a new chess engine that will change all of that.

I don’t really like Element’s code: It’s clunky, some of its code is awkward, and of course all of the floating blocks in its codebase. Of course, I am very attached to Element, but so many bad habits have been put into Element. It’s time to make the switch.

However, I am not discontinuing Element’s development. Element has its weak points, and some of them can be handled. Unfortunately, because my laptop blue screened, I was left using the “emergency” backup device I thought I would never use again. That device being a 9 year old Chromebook. I took it out, charged it, and attempted to open Element in the editor. It took 2 minutes, but it worked. I was elated. I could finally update Element. So, where have I been all of this time. Simple; the Chromebook is incredibly slow. I managed to get 7 NPS on Element on S3 depth 1. I also edit my code on S3 (the only exception was the TT, but the components were imported from TW, then implemented in S3). Despite this, I didn’t update Element a lot. It was significantly harder to edit my code on my Chromebook. However, it was enough for my next update at the time; Element v1.4.10.3 to v1.4.11. This time I simply imported the dev to the stable version. 1.4.11 prevented Element from blundering M3 by adding checkmate detection on the final ply. At the time, this was crucial, and I was extremely happy to get it finished. Much later, I discovered that my refutation moves did in fact speed of the engine during the middlegame. So, after very few block changes, 1.4.12 was out. 1.5 came later and that was a miracle. I never thought I would get it out because I was working on it in TW, so the file was on the laptop I could not access. However, I managed to find the sprite that I imported from TW to release from 1.5. To my absolute shock, it worked. This…should not have been possible. The previous TT dev had a bug where it would hang checkmate, and I could never ever figure out why it was doing that. But now, all of a sudden, the code is working perfectly and there are no bugs. I could not believe it. How did a bug that has existed for months just…disappear? Believe it or not, I managed to come up with an explanation for what happened. Long story short, the original TT dev was so old, it was before the king captures update (v1.485) and that caused complications.



So, what happened? I struggled to really gain interest in coding because of the slow device and editor. I had taken a 6 month long break from anything major for Element, but I had to take another break. After all, I was happy with my new and working transposition table. I was happy I could make changes at all with a 9 year old device.



After looking at many games from GoK, White Dove, and the rest of the engines, as well as my knowledge of the engines led me to one conclusion. These engines cannot handle aggressive play (except for Delta) and GoK NNUE). Because of this, I plan to create a new chess engine that plays extremely aggressively. However, Element will not be discontinued and will still get updates. I finally have a new device now, and I am ready to take chess engines on Scratch to a level of complexity, knowledge, and sheer strength to another level. This new engine will be stronger and faster than Element, and it will play very aggressively. I believe I have the key to unlock the next level, and all I have to do is find the correct door. It will likely take me months to find it, but I will find that door.

You may try testing Patricia 4 chess bot. It is the most aggressive chess engine ever existed with 3500 rating. Hope to see your success!

birdracerthree wrote:

An Announcement
Hello, I have an announcement to make:

As you guys may be able to tell, Element has been behind its old rival of White Dove and its competitors like GoK Classic and Delta. However, I am working on a new chess engine that will change all of that.

I don’t really like Element’s code: It’s clunky, some of its code is awkward, and of course all of the floating blocks in its codebase. Of course, I am very attached to Element, but so many bad habits have been put into Element. It’s time to make the switch.

However, I am not discontinuing Element’s development. Element has its weak points, and some of them can be handled. Unfortunately, because my laptop blue screened, I was left using the “emergency” backup device I thought I would never use again. That device being a 9 year old Chromebook. I took it out, charged it, and attempted to open Element in the editor. It took 2 minutes, but it worked. I was elated. I could finally update Element. So, where have I been all of this time. Simple; the Chromebook is incredibly slow. I managed to get 7 NPS on Element on S3 depth 1. I also edit my code on S3 (the only exception was the TT, but the components were imported from TW, then implemented in S3). Despite this, I didn’t update Element a lot. It was significantly harder to edit my code on my Chromebook. However, it was enough for my next update at the time; Element v1.4.10.3 to v1.4.11. This time I simply imported the dev to the stable version. 1.4.11 prevented Element from blundering M3 by adding checkmate detection on the final ply. At the time, this was crucial, and I was extremely happy to get it finished. Much later, I discovered that my refutation moves did in fact speed of the engine during the middlegame. So, after very few block changes, 1.4.12 was out. 1.5 came later and that was a miracle. I never thought I would get it out because I was working on it in TW, so the file was on the laptop I could not access. However, I managed to find the sprite that I imported from TW to release from 1.5. To my absolute shock, it worked. This…should not have been possible. The previous TT dev had a bug where it would hang checkmate, and I could never ever figure out why it was doing that. But now, all of a sudden, the code is working perfectly and there are no bugs. I could not believe it. How did a bug that has existed for months just…disappear? Believe it or not, I managed to come up with an explanation for what happened. Long story short, the original TT dev was so old, it was before the king captures update (v1.485) and that caused complications.



So, what happened? I struggled to really gain interest in coding because of the slow device and editor. I had taken a 6 month long break from anything major for Element, but I had to take another break. After all, I was happy with my new and working transposition table. I was happy I could make changes at all with a 9 year old device.



After looking at many games from GoK, White Dove, and the rest of the engines, as well as my knowledge of the engines led me to one conclusion. These engines cannot handle aggressive play (except for Delta) and GoK NNUE). Because of this, I plan to create a new chess engine that plays extremely aggressively. However, Element will not be discontinued and will still get updates. I finally have a new device now, and I am ready to take chess engines on Scratch to a level of complexity, knowledge, and sheer strength to another level. This new engine will be stronger and faster than Element, and it will play very aggressively. I believe I have the key to unlock the next level, and all I have to do is find the correct door. It will likely take me months to find it, but I will find that door.

POV: new bird's setup is less stronger than his phone

So I'm starting to make my own chess engine on cpp, and well, I'll be happy whan it will be able, with 2 seconds think time, to beat the arnohu's one…