Hello friends of Stockfish!
Stockfish is a great engine for playing with engines, as well as for analyzing chess theory. I like them both. I really enjoyed playing the Engine Masters tournaments on InfinityChess, as well as the Freestyle tournament (analysis mode). In June 2022 I played Blue Marlin 15.1 at the Engine Masters tournament and I really liked the way the engine played. I also really liked Corchess for analysis.
That's why I then started to study the source code of these two engines. Although I had no idea at first, I couldn't help but want to know how these two engines differ from each other and what the difference to Stockfish dev. lies. So I compared the source codes with each other. My question was: could some of Blue Marlin's ideas be combined with Corchess' search? This would become my ideal engine! I then compared the source code from Corchess with that from Stockfish dev. compared just to find out what the difference between these engines is. I found out. Who seeks finds! After that I implemented some code from Corchess into Stockfish dev and vice versa. After that, some parameters (ideas from Blue Marlin) were changed. I had created my first Leptir Engine version!
Stockfish is evolving rapidly. The source code is constantly changing. A few days ago, 73 parameters were changed in one day. Unbelievable. Before that, there were often such extreme changes. I find it hard to believe that you can test every single possible combination of 73 parameters in such a short amount of time. Luckily, I found that just changing parameters doesn't change that much. There was a bigger change, for example implemented as "EvenDeeperSearch". Such codes change the search significantly more.
I have therefore limited myself to implementing and testing such codes. I deleted what I didn't see as beneficial. I liked some of the older codes better than the newer ones, so I replaced them. I've been doing this for about 6 months now.
"Leptir Analyzer" is the current state of it. This engine combines ideas from Blue Marlin with search technology from Corchess and my own tested best codes from Stockfish dev and Fishcooking. Yes, I can say: I managed Lepir Analyzer well, it is a fantastic tool for analyzing chess theory. The engine has a slow search with tremendous tactical power.
That's enough for me. In addition to Leptir, I will also use Corchess and Blue Marlin for my analyses! These 3 engines are my recommendation for analysis and play. Therefore I will remove all other engines from my homepage in the coming days. If you are interested in charisma, you should hurry up with the download!
There is another reason for the change. I've noticed that there are authors who only want to push their engines out of vanity or profit. Just take PlayChess as an example. There, some people brag about >3000 Elo, which they cheated by winning against friendly accounts (quasi, my engine X wins against my engine Y - hooray engine X is the king). For weeks they let themselves be celebrated at the top of the table, out of sheer vanity. An example looks like this:
There are more such kings. I have to endure such images every day when I play on PlayChess. Each of us wants to be successful, that's normal and good. I've also hit 3000 Elo several times, but I've never cheated with friends, so I'll fall again because I'm playing in automatic mode. But I won't stop criticizing such vain accounts. The people on PlayChess know what's going on, but the people outside don't. The people out there (and here in the forum) don't know how the Super-Elos came about. It's not my intention to discriminate against other engines because of my own engines. I can only say NO! Forget me, forget my engines. But don't forget to think and ask other people how the PlayChess (or elsewhere) superstar engines really tick. I can only prove that this is my honest opinion without any ulterior motives by changing my homepage.
- play against your tester account again and defeat him (Proteussf is his other test account, ProteusSF has 8 such test accounts in total on PlayChess) - you'll be over 3000 Elo again. 
