Hello Chess Friends,
On this news page,
You can read my postings regarding for OTS v5 Ranking
05.05.2022:
As usually, I am pleased to announced that,
Another new Opening Suite Ranking is out there!
Btw, for anyone interested..previous OTS Results
In short, I've started to work over new project, where this time:
I analyzed the engines with new OTS v5 suite (193 positions)
Note also that after the end of competition, I hope to release it...
In other words, I think that:
My created opening suite is not faultless, why I say this...
Because in some positions, there are alternative move/s...
However, during opening suite's creation I've done my best
About picking the right best moves (bm) and after serious
Analyzing, checking comparing etc it seems via OTS v5,
The current engine ranking is not so bad indicator about
Which are the greatest engines in opening.knowledge..!)
And I hope you like and appreciate all my free efforts...
Really a lot of work is required...not so easy as it looks...
Meanwhile,
I've analyzed some of the Top engines as MP (with 12 cores) too
And to be honest, no much influence in case of analyzing as NNUE
Actually this is known even in game-play the performance is not huge
I mean that NNUE engines as MP do no gain too much for: Results
Btw, Stockfish 310720 12 cores performed much better: 102 points
E.g if comparing, Stockfish 310720 with 1 core performed 76 points
That means as NON-NNUE the well-known speed formula works !)
However, I've realized to set for all 1 core, at least in this way,
The error margin is expecting to be less than the testings as MP
Plus even with 1 core, the NNUE's performance is really so good...
Btw, it's an old, but very popular story...and just out of curiosity,
Fruit 2.1 and Rybka 1 Beta are tested too: 11 points difference
As you may know, Rybka is accused to be as clone of Fruit and
I wonder if Rybka is a clone: Why there is 11 points difference?
Note: Fruit is released on 17.05.2005 / Rybka on 04.12.2005
And same words are going for Houdini 6.03 engine as well...
For anyone missed, it has been accused as Stockfish 8 clone
If so, then I wonder again: why Houdini performed better than
Stockfish 8, Stockfish 9 and even better than Stockfish 10 ?
E.g the difference is 19 points (Houdini 603 Tact vs Stockfish 8)
In other words, maybe I am missing something...?)
Note also that I tested some of the older engine versions too...
E.g one of them is Raubfisch X39e3b, belongs to 2018 year
As we see...if NON-NNUE..Raubfish as 4 years old and still best!
Plus Number One of OTS 4 and OTS 3 opening rankings as well!
Note: Raubfish tested: Analysis_ICC ON, Null Move Heuristic OFF
However, I have to admit that for nowadays, NNUE is the KING...!
ShashChess, Fat Titz, SF PB, Stockfish, Komodo Dragon etc.are
Extremely strong and great solvers in various opening positions !!
By the way, as another good news is that (e.g after each release):
Newer Stockfish vers are becoming stronger and stronger in openings!
That means also that OTS v5 suite serves the purpose... right ?)
Stockfish - Official Releases (2016-2022)
SF Official Points
Stockfish 15 NN : 123
Stockfish 14.1 NN : 117
Stockfish 13 NN : 105
Stockfish 12 NN : 92
Stockfish 11 : 81
Stockfish 10 : 71
Stockfish 9 : 61
Stockfish 8 : 53
Stockfish 7 : 42
Ok...that's all for now...and thanks for your reading...
Best,
Sedat